Search is shifting. It is no longer just about fighting for the top spot on a page of blue links. It is about being the single source of truth that a machine trusts enough to repeat to a patient.
In healthcare and pharma, where the stakes are high and the regulations are tight, standard content does not cut it anymore. We have to build for two new realities: AEO (answering the user directly) and GEO (ensuring AI models cite you as the authority).
Traditional SEO was about keywords and backlinks. It still matters, but it is just the baseline. Today, we have to look at two specific ways people and the machines they use find information:
Think of AEO (Answer Engine Optimization) as "Zero-Click" optimization. The term "zero-click" was popularized by Rand Fishkin at SparkToro back in 2019, when his research showed that less than half of Google searches resulted in a click to any website. The idea is simple: a user gets their answer directly on the search results page and never clicks through to your site. That number has only grown since then. AEO takes that reality and runs with it. When a patient asks, "Is Metformin safe for my kidneys?" they do not want to read a long blog post. They want a clear, factual answer at the top of the screen. AEO is the art of structuring your data so a voice assistant or a "Featured Snippet" can grab it instantly.
Generative engines like ChatGPT or Google’s AI Overviews do not just find links; they synthesize information. Research from institutions like Princeton and Georgia Tech shows that "Generative Engine Optimization" can boost a brand's visibility in AI responses by up to 40% when content is properly structured for citation.
| The Goal | AEO (Answer Engines) | GEO (Generative Engines) |
|---|---|---|
| Main Objective | Answer a specific question. | Be the trusted source for an AI summary. |
| Focus | FAQs, snippets, and voice. | Entity relationships and citations. |
| Content Style | Clear Q&A and bulleted lists. | Fact-heavy, authoritative, and cited. |
| Healthcare Priority | High (for patient education). | Critical (for brand authority). |
Google treats healthcare as a YMYL (Your Money or Your Life) topic. According to Google’s latest Search Quality Rater Guidelines, the bar for accuracy and trust is significantly higher here than in any other industry.
If you are in pharma or clinical care, you are not just competing on content; you are competing on trust. AI models are inherently cautious. They will not cite you if they cannot verify who you are, who wrote your content, and whether a medical professional actually reviewed it. This concept is known as E-E-A-T (Experience, Expertise, Authoritativeness, and Trustworthiness), which is Google's framework for evaluating whether content deserves to rank. In healthcare, every one of those four signals carries extra weight.
If you want a machine to understand your page without guessing, you use Schema markup, a standardized vocabulary of code that you add to your website so search engines and AI models can read your content the way a database reads a spreadsheet, not the way a human reads a paragraph. Think of it as a nutrition label for your content. Instead of hoping Google figures out that your page is about a specific drug or condition, Schema tells it directly:
It is a valid fear. While "zero-click" searches are rising, the traffic that does click through is often much higher quality. In 2026, we are seeing that users who click through from an AI citation have already vetted your brand via the AI summary. They are looking for specific, deep-dive information from a source they already trust.
The metrics are shifting. Instead of checking if you are in the top spot for a specific drug side effect, you should look at your Citation Share. This involves auditing how often AI models mention your brand as a source. If the AI is repeating your data, you are winning.
It is for everyone. While Google pioneered it, models like Claude, GPT-4, and Perplexity all use structured data to verify facts. Think of Schema as a universal translator that helps any machine understand that your product treats a specific condition.
For healthcare and pharma, the answer is yes. AI systems check for expertise by looking at the credentials of the author and reviewer. Content that is not tied to a verified medical professional is increasingly being ignored by generative engines as unreliable.
The biggest risk is not just being hard to find. It is being misrepresented. If you do not provide the structured data and clear facts for your brand, AI engines will try to guess based on third-party info or old data. For pharma companies, having an AI hallucinate your drug’s safety profile is a massive regulatory liability.
Traditional SEO gets you found. AEO makes you helpful. GEO makes you the authority. In an industry built on trust, you need all three.--- # American Academy of Family Physicians: Why I Chose Google Over Adobe ($100K+/Year in Savings) URL: https://brianroseman.com/blog/why-i-chose-google-over-adobe-enterprise-analytics Category: Analytics Published: 2026-01-02 After joining AAFP to lead digital transformation, I made the strategic decision to migrate from Adobe Analytics to Google Analytics 4. Here is why the Google tech stack won and what I learned along the way. ## The Challenge: Legacy Analytics Holding Back Growth When I joined the American Academy of Family Physicians (AAFP) to help lead digital transformation, one of the first major decisions I faced was evaluating our analytics infrastructure. The organization was running Adobe Analytics, a powerful enterprise solution, but one that came with significant overhead and, frankly, was collecting dust. Here is the reality: Adobe Analytics is a Lamborghini sitting in a garage with flat tires. Powerful? Absolutely. But if nobody knows how to drive it, you are just paying for an expensive parking spot. ## What The Audit Revealed We performed a deep-dive web analytics audit across 100+ dimensions, including a head-to-head comparison of Adobe Analytics vs Google Analytics. What we found was sobering: - Basic event tracking with no ongoing maintenance - Tag manager being used as "patches" for data layer issues - Goals not working properly - Inconsistent UTM tagging with manual, error-prone processes - Tribal knowledge loss due to turnover - Reports done manually, one at a time - No formal data governance or data dictionary - User segments too broad to be actionable The assessment ranked us at the bottom of the analytics maturity curve. We were stuck in "Capture and Store" when we needed to be at "Act" with visualization, segmentation, prediction, and personalization. ## The Plot Twist Here is the interesting part: the agency initially recommended sticking with Adobe Analytics. They said we would "quickly outgrow" Google Analytics in 1-2 years. I disagreed. After evaluating both platforms against our actual needs, team capabilities, and budget constraints, I made the call to migrate to GA4 anyway. Sometimes the right answer is not the fanciest tool. It is the tool your team will actually use. ## Why Organizations Consider Switching from Adobe Analytics **Cost Comparison** - **Adobe Analytics** requires significant licensing fees, often $100,000+ annually for enterprise deployments - **Google Analytics 4** is free for most organizations, with GA360 available for enterprise needs at lower cost than Adobe - We saved **$50K+ annually** by making the switch **Integration Capabilities** - **Adobe** works best within the Adobe ecosystem (Adobe Target, Adobe Campaign, etc.) - **Google** seamlessly connects with Google Ads, Search Console, BigQuery, Looker Studio, and Tag Manager **Learning Curve and Talent Pool** - **Adobe** requires specialized training and certified analysts are harder to find - **Google** skills are more common, making hiring and training easier - This matters more than people realize. What good is a powerful tool if your team cannot use it without a PhD in Adobe? **Data Ownership and Flexibility** - **Google Analytics 4** with BigQuery export provides raw data access and complete ownership - Custom analysis becomes possible without platform limitations ## The Google Tech Stack I Implemented My recommended stack for digital transformation included: 1. **Google Analytics 4** for web and app analytics 2. **Google Tag Manager** for tag deployment and management 3. **Google Search Console** for SEO performance data 4. **Looker Studio** for custom dashboards and reporting 5. **BigQuery** for raw data storage and advanced analysis 6. **PowerBI** for enterprise dashboards our dev team could own 7. **Windsor.AI** for paid media data aggregation (similar to Supermetrics) ## The Multi-Phase Rollout We did not just flip a switch. This was a multi-phase implementation starting in late 2021: **Phase I:** Build the business case, audit existing Adobe implementation, document gaps **Phase II (December):** Data layers and implementation, BI platform configuration, API integrations. We used Windsor.AI to pull data from Google Ads, Bing, Facebook, LinkedIn, and other advertising platforms into our unified data warehouse. **Phase III (January-March):** Self-service dashboards across channels: paid media, SEO, content performance, membership, conferences, and more We integrated 12+ data sources into a unified warehouse: GA4, Firebase, Google Search Console, Google Ads, Bing, Sprout Social, Facebook, LinkedIn, Twitter, YouTube, and more. **Want the full story?** [Read the complete case study here](/case-studies/enterprise-analytics-google-adobe-migration). ## Key Benefits After Migration **Unified Data View** GA4's event-based model provides better cross-platform tracking than Adobe's session-based approach. **Real-Time Insights** Faster data processing means quicker decision-making for marketing campaigns. **Reduced Total Cost of Ownership** Licensing savings redirected to actual marketing spend and team development. That $50K we saved? It went back into initiatives that moved the needle. **Future-Proof Architecture** GA4 was built for a privacy-first, cookieless future with machine learning built in. **Self-Service Culture** This was the biggest win. Stakeholders could finally pull their own reports instead of waiting in a queue. We shifted from "request data" to "access data." That cultural shift is worth more than any tool. ## Lessons Learned from Enterprise Analytics Migration 1. **Start with clear objectives** before choosing platforms 2. **Map your current implementation** to understand what you will lose or gain 3. **Plan for parallel tracking** during transition periods 4. **Train your team early** on the new platform 5. **Document everything** for institutional knowledge 6. **Build for self-service** from day one. If stakeholders cannot access their own data, you are still a bottleneck 7. **Trust your judgment** even when the experts recommend something different ## Is Adobe Analytics to GA4 Migration Right for You? Consider migrating if: - Your Adobe licensing costs exceed the value delivered - You want tighter integration with Google advertising products - Your team struggles to find Adobe-certified analysts - You need more flexible data export options - You are not Bank of America. If you are a mom and pop shop selling pizzas, GA4 gives you everything you need without the enterprise price tag Stay with Adobe if: - You are deeply invested in the Adobe Marketing Cloud ecosystem - You have specialized Adobe implementations that would be costly to recreate - Your team has strong Adobe expertise you do not want to lose - You actually use the platform to its full potential (be honest with yourself here) ## Bottom Line Digital transformation is not about chasing the newest tools. It is about choosing the right tools for your organization's goals, budget, and team capabilities. For AAFP, the Google tech stack delivered better value, easier integration, and a clearer path forward. The agency said we would outgrow GA4 in 1-2 years. Three years later, we are still running strong, $150K+ richer, and our team actually uses the data to make decisions. The best analytics platform is the one your team will actually use to make better decisions. Everything else is just expensive shelfware. --- # The Complete Guide to AIO Content Templates URL: https://brianroseman.com/blog/complete-guide-aio-content-templates Category: Content Strategy Published: 2026-01-01 AI Overviews and answer engines are changing how people discover content. This guide breaks down 23 page types that consistently earn visibility in AI-powered search experiences. AI Overviews and answer engines are changing how people discover content. A growing share of searches now end with a summary and a handful of sources, not a click to a traditional results page. If you are seeing organic traffic flatten or decline even when rankings look fine, you are not imagining it. The intent is still there. The click behavior is different. This guide breaks down 23 page types that consistently earn visibility in AI-powered search experiences because they are easy to quote, easy to verify, and easy to understand. ## Key Takeaway These 23 content templates are designed to increase the chances your pages are referenced inside AI-generated answers. The patterns are consistent across tools: lead with the answer, keep structure predictable, state who the content is for, and show evidence of experience. ## Why AIO Templates Matter AI-powered systems do not read the way humans read. They scan for: **A clear definition or direct answer** **A structured explanation they can reuse** **Named entities and relationships** **Timeframe and scope** **Confidence signals like author expertise, citations, and transparency** If your answer is buried three paragraphs deep, wrapped in a story, or written in a vague marketing tone, the system often moves on. ## The Core Principles Behind Every AIO Template These five rules apply to every page type in this guide. ## 1. Answer First, Explain After Your first 2 to 5 sentences should stand on their own. Bad: "In today's changing landscape…" Good: "A GA4 event is a tracked interaction such as a form submit or video play. You should track events that indicate intent and conversion progress." ## 2. Opening Summary Block Is Non-Negotiable After your H1, include a 2 to 3 sentence summary that makes sense even if the reader sees only that block. ## 3. Be Explicit About Subject, Audience, and Timeframe AI systems need clarity. Tell them what this page covers, who it is for, and whether it applies today, this year, or to a specific scenario. ## 4. One Page, One Job Mixed-intent pages confuse people and machines. Do not try to make one page be a service page, a blog post, an FAQ, and a landing page all at the same time. ## 5. Structure Beats Fluff Short paragraphs, headings, tables, lists, and examples outperform long narrative blocks. ## Content Templates Overview The guide covers 23 page types across three categories: **Content Templates (7 Types):** Blog posts, news articles, glossary pages, hub pages, resource libraries, comparison pages, and FAQ pages. **Website Templates (6 Types):** Homepage, about page, services page, policy pages, author pages, and methodology pages. **Marketing Templates (10 Types):** Landing pages, case studies, product pages, testimonials, and more. Each template includes the exact recommended structure, what the page is best used for, a realistic example topic, common mistakes to avoid, and a citation-ready sample snippet idea. ## Making Your Content Citation-Ready The goal is simple: make each page safe to quote without needing extra context. When AI systems can clearly identify your expertise, extract your key points, and verify your claims, your content becomes a trusted source they reference again and again. --- # SEO When Clicks Disappear: Earning Visibility in AI-Powered Search URL: https://brianroseman.com/blog/seo-visibility-ai-powered-search Category: SEO Published: 2026-01-01 If it feels like your organic traffic is slowly slipping away, you are not imagining it. Search experiences are becoming more conversational and more selective. Systems choose sources they trust and reuse them across many answers. If it feels like your organic traffic is slowly slipping away, you are not imagining it. Many teams are seeing the same pattern. Rankings look stable. Search demand appears steady. Yet clicks are down, and performance feels harder to explain. The disconnect is unsettling, especially for organizations that have invested years building strong SEO foundations. What is happening is not a collapse of search. It is a shift in how answers are delivered. Modern search experiences increasingly resolve questions directly. Users are shown summaries, explanations, and follow-up context without needing to visit multiple sites. The information still comes from somewhere, but the interaction no longer guarantees a click. This creates a new challenge for brands. Visibility now matters just as much inside the answer as it does on the results page. The good news is that this change does not require starting over. The strongest SEO strategies still apply. They simply need to evolve to match how modern search systems decide what to reference, trust, and reuse. Below are four ways to do exactly that. ## From Ranking Pages to Being Recognized For years, SEO focused on optimizing pages. The goal was clear. Create the best page for a query, earn the ranking, and capture the traffic. That approach still has value, but it no longer tells the full story. Search systems increasingly operate on entities rather than pages. They try to understand brands, products, services, and people as distinct, reusable sources of knowledge. When an answer is generated, the system pulls from entities it recognizes and trusts. This explains why two pages with similar content can perform very differently. One might rank well but never be referenced in summaries. The other might receive fewer clicks but appear consistently as a cited or implied source. The shift is subtle, but it changes what success looks like. Search demand is not disappearing. Attention is being redistributed. ## 1. Make Experience Visible, Not Assumed Expertise has always mattered in search, but modern systems look harder for proof. It is no longer enough to explain how something works. Content performs best when it demonstrates that the author or organization has actually done the work. Experience shows up through: - Clear authorship - Relevant background - Specific examples - Evidence of outcomes This applies across industries, from ecommerce to healthcare to enterprise software. **Example:** A consulting firm publishes an article titled "How to Improve Marketing Attribution." It explains common models, tools, and challenges. The piece is accurate but generic. It performs moderately well and attracts some traffic. Later, the firm publishes a second article titled "What Broke Our Attribution and How We Fixed It." This version includes the role of the author, screenshots of reporting changes, mistakes they made along the way, and how results changed after the fix. The article is longer and more specific. The second article is referenced more often in summaries and follow-up answers, even when it ranks slightly lower. The system can clearly see experience rather than interpretation. The lesson is simple. Content that shows real work is easier to trust and reuse. ## 2. Write So Systems Can Read You Clearly Design trends often favor complexity. Interactive elements, animations, and heavy client-side rendering can make pages look impressive. But many modern crawlers struggle with complexity. While major search engines can handle JavaScript reasonably well, many answer systems rely on simpler parsing methods. They look for clear structure, predictable formatting, and complete statements. Content that performs well is usually: - Fully rendered in HTML - Structured with clear headings - Written in plain language - Free of unnecessary layout distractions **Example:** A product company builds its documentation inside a dynamic interface. Everything loads after the page renders. Users love the experience, but references in summaries are inconsistent. The team tests a simplified version of the same content. It uses standard headings, clear definitions, and static HTML. Nothing about the information changes. Over time, that version becomes far more visible in generated answers. The insight here is not that design is bad. It is that clarity beats cleverness when machines are deciding what to reuse. ## 3. Strengthen the Signals Most People Never See Some of the most important signals for modern search visibility exist behind the scenes. Structured data, metadata, alt text, and transcripts help systems understand what something is, who created it, and how it relates to other information. These signals do not directly drive traffic. They improve confidence. **Example:** A professional services firm adds structured information across its site. Author profiles are connected to content. Services are clearly defined. FAQs are marked up consistently. Video pages include transcripts. Nothing changes visually. Traffic does not spike. But over several months, the firm begins appearing more frequently in answer summaries related to its specialty. Sales teams report prospects mentioning the brand before the first call. This is what presence looks like when it works. Quiet, consistent, and cumulative. ## 4. Measure What Matters When Clicks Decline Traffic used to be the clearest signal of success. More visits meant more opportunity. That relationship is weakening. As answers are resolved earlier, visibility often happens without a visit. Measuring only sessions can make strong performance look like failure. Teams that adapt shift their focus to: - Conversions influenced by organic content - Brand recognition in search experiences - Changes in buyer readiness - Downstream impact on sales cycles **Example:** A B2B company sees organic traffic decline year over year. At the same time, demo requests remain steady and close rates improve. Sales conversations are shorter. Prospects arrive with more context. The marketing team correlates these trends and realizes that content is doing more work earlier. Fewer clicks are needed because fewer questions remain unanswered. The conclusion is not that SEO is broken. It is that success moved upstream. ## Why This Shift Matters Now This evolution is not temporary. Search experiences are becoming more conversational and more selective. Systems choose sources they trust and reuse them across many answers. That trust compounds over time. Brands that focus only on rankings will feel like something is being taken from them. Brands that focus on clarity, credibility, and consistency will quietly gain influence. This is not about gaming a new system. It is about aligning with how information is evaluated. ## What Strong Teams Are Doing Differently Organizations that adapt successfully tend to share a few habits. They: - Write content based on real questions they hear internally - Show experience instead of summarizing theory - Simplify structure so information is easy to interpret - Invest in clarity behind the scenes - Measure success through outcomes, not vanity metrics None of this is flashy. All of it works. ## What This Means for You Search isn't disappearing. It's becoming more selective. Users get answers faster. Systems pick sources more carefully. Brands that stay clear, credible, and consistent earn the visibility that matters. The organizations that figure this out won't necessarily see traffic spike. They'll see something better: recognition at the exact moment someone needs an answer. --- # What Actually Matters in Digital Marketing for 2026 URL: https://brianroseman.com/blog/digital-marketing-priorities-2026 Category: Strategy Published: 2026-01-01 Marketing in 2026 rewards teams that focus on clarity, consistency, and trust. The most successful organizations are not louder. They are sharper, more disciplined, and better aligned with how people actually make decisions. By 2026, digital marketing is no longer about running more campaigns. It is about building systems that create consistent results, earn trust, and hold up under scrutiny from leadership teams. The best teams are not chasing trends. They are tightening fundamentals, improving execution, and making smarter decisions with better data. Below are the twenty initiatives that matter most, along with real examples of how they show up in strong organizations. ## 1. Content Built Around Real Questions The most effective content in 2026 starts with real questions people ask internally and externally. Instead of building pages around keywords, teams are mapping content to conversations they hear from customers, sales teams, and support teams. **Example:** A company replaces five thin blog posts with one in-depth guide answering "How does attribution actually work in our industry?" The page includes a summary, definitions, common misconceptions, examples from real campaigns, and follow-up questions. Traffic grows steadily, but more importantly, sales teams start sharing it directly with prospects. ## 2. First-Party Data as the Source of Truth Strong teams know exactly where their data comes from and how it connects. They invest time in cleaning CRM records, defining events clearly, and making sure web, email, and paid data align. **Example:** A marketing team discovers their lead counts look strong, but sales conversion is weak. After auditing first-party data, they realize half their "leads" never completed meaningful actions. They update event definitions and scoring rules. Lead volume drops, but close rates increase. ## 3. Faster Content Production Without Cutting Corners Speed matters, but not at the expense of credibility. Teams that publish consistently tend to document processes, reuse formats, and standardize quality checks. **Example:** Instead of writing every article from scratch, a team creates a repeatable structure for case studies. Writers focus on insights and outcomes instead of formatting. Publishing cadence doubles without sacrificing quality. ## 4. Experience-Based Content That Shows the Work People trust content that shows evidence. This means explaining what was tried, what worked, and what changed as a result. **Example:** A growth team writes a post explaining how they reduced paid search waste. They include screenshots of dashboards, before-and-after metrics, and lessons learned. The article attracts fewer casual readers but more qualified inbound leads. ## 5. Measurement That Connects to Revenue Marketing teams gain credibility when they report on outcomes leadership cares about. This includes pipeline contribution, retention, and lifetime value. **Example:** Instead of reporting monthly traffic growth, a team shows how a content series influenced demo requests and renewals. The marketing budget conversation shifts from cost to investment. ## 6. Automation That Removes Busywork Automation works best when it supports people rather than replacing judgment. Teams use it to handle repetitive tasks so humans can focus on strategy. **Example:** Weekly performance reports are automatically generated and summarized. Team meetings focus on decisions and next steps rather than reviewing charts. ## 7. Search That Includes Visual and Video Content People consume information in different ways. Search reflects that. Pages that combine written explanations with visuals tend to keep attention longer. **Example:** A technical article includes diagrams and short explainer videos. Time on page increases, bounce rate drops, and the content is shared more frequently. ## 8. Short-Form Video as a Core Channel Short video works because it lowers the effort required to learn something. Teams use it to explain ideas quickly and build familiarity. **Example:** A product leader records weekly one-minute videos answering common customer questions. These clips are shared on social, embedded in emails, and reused on landing pages. ## 9. Long-Form Video for Trust and Depth Longer video formats help people understand complex ideas and build confidence. They are especially effective for education and onboarding. **Example:** A company records a forty-minute walkthrough explaining how their platform works in practice. Sales teams use it during late-stage conversations, reducing repetitive demos. ## 10. Paid Media That Learns From Organic Performance Paid media works better when it reflects what already resonates organically. Teams watch which topics and messages perform best before scaling them. **Example:** A blog post consistently drives high engagement. The team repurposes its core message into paid ads. Cost per conversion drops because the message is already proven. ## 11. Content Designed to Drive Action Strong content guides readers toward the next step naturally. This does not mean aggressive selling. It means clarity. **Example:** A guide ends with "If this matches your situation, here is how we typically help." Conversion rates improve because the call to action feels relevant. ## 12. Scaling Content Without Losing Quality Publishing at scale requires discipline. Teams define standards and review regularly to avoid dilution. **Example:** A company expands location pages but includes real examples, testimonials, and local context on each one. Pages feel useful instead of generic. ## 13. Trust Signals Across the Entire Experience Trust is built through consistency and transparency. This includes clear ownership and honest positioning. **Example:** Author bios include real experience. Content is updated regularly. Contact information is easy to find. Prospects mention feeling more confident before sales calls. ## 14. Deeper Coverage of Fewer Topics Depth builds authority faster than volume. Teams choose fewer topics and cover them thoroughly. **Example:** Instead of publishing broadly, a team focuses on three core themes and builds guides, FAQs, videos, and case studies around each. Organic growth becomes more predictable. ## 15. Looking Ahead, Not Just Back Strong teams use past performance to inform future decisions. They model scenarios and plan proactively. **Example:** Marketing leaders forecast pipeline impact before reallocating budget. Fewer surprises show up at the end of the quarter. ## 16. Continuous Conversion Optimization Small improvements add up. Teams review behavior regularly and test changes consistently. **Example:** A simple form redesign reduces friction and increases completions by ten percent without increasing traffic. ## 17. Lifecycle Marketing That Feels Connected Messages should evolve based on behavior. Disconnected communication creates confusion. **Example:** A customer who downloads a guide receives follow-up content aligned to that topic instead of generic newsletters. Engagement improves. ## 18. Repurposing That Extends Value Good ideas should work harder. Teams intentionally reuse strong content across channels. **Example:** A webinar becomes blog posts, short videos, internal training, and sales enablement. One effort supports multiple teams. ## 19. Running Marketing Like a Product High-performing teams plan, test, and iterate. They prioritize work based on impact. **Example:** Marketing roadmaps are reviewed quarterly. Experiments are documented. Learnings inform future decisions. ## 20. Clear Storytelling With Data Numbers matter, but stories drive decisions. Leaders connect results to meaning. **Example:** Instead of showing charts, a report explains what changed, why it matters, and what action is recommended. Executives engage instead of skim. Marketing in 2026 rewards clarity, consistency, and trust. The most successful teams aren't louder. They're sharper, more disciplined, and better aligned with how people actually decide. --- # INSIGHTS # Travel Email Marketing: Seasonal Campaigns That Fill Off-Peak Inventory URL: https://brianroseman.com/insights/travel-email-marketing-seasonal-campaigns Published: 2026-05-09 Most travel brands market heavily during peak demand and go quiet when they actually need bookings. Off-peak email campaigns with the right segmentation and timing fill the inventory that would otherwise sit empty. **Summary:** Travel email marketing is fundamentally a seasonality problem. Most travel brands are great at marketing during peak demand and mediocre or worse at generating bookings during shoulder and off-peak periods. This post covers the campaign structure, audience segmentation, and timing strategy that fills inventory during the seasons when you actually need it filled. **The Seasonality Trap Most Travel Brands Are In** Travel demand has natural peaks that vary by destination: summer for beach destinations, winter for ski resorts, spring and fall for city breaks, holiday periods for family travel. Most travel brands run their heaviest marketing during these peaks and pull back during slow periods, which is almost exactly backwards from an efficiency standpoint. During peak demand, customers will find you. They're actively searching, your destination is top of mind, and you're competing with every other travel brand for the same pool of motivated buyers. Marketing spend during peak periods often produces diminishing returns because every competitor is also spending heavily. During shoulder and off-peak periods, customers need a reason to consider traveling. That reason can come from your email list, from a well-timed promotion, or from content that reframes the slow season as an advantage. The brands that win the off-peak game have built a marketing calendar that works year-round rather than spiking in one season. **Your Email List Is a Seasonality Hedge** An email list of past guests or previous bookers is the most valuable asset a travel brand has for managing seasonality. These are people who have already experienced your destination or product and liked it enough to return or share it with others. Converting them again costs a fraction of acquiring a new customer. The travel brands that manage seasonality well treat their email list as an always-on asset rather than a broadcast channel for peak promotions. They send relevant, targeted content year-round that keeps the destination in consideration even when the subscriber isn't actively planning travel. A beach resort that only emails "Book your summer vacation now!" misses the subscriber who would have considered a March trip if they'd been reminded at the right moment. The same resort sending a "Spring break options still available" email in late January captures planning intent that the subscriber didn't know they had. **Segmentation That Makes Off-Peak Campaigns Work** Generic off-peak promotions to your entire list underperform because different segments of your list have different travel patterns and price sensitivity. The segmentation that makes off-peak email campaigns work: Past travel timing: When did they book and travel before? A subscriber who booked a trip in October is far more likely to book another October trip than a subscriber who always travels in July. Use past booking data to build timing-based segments and target off-peak campaigns at the subscribers with demonstrated willingness to travel in that period. Geographic proximity: Subscribers within driving distance of a destination are far easier to convert on short-notice campaigns than subscribers who need to book flights. A "this weekend" offer to subscribers within 150 miles of your destination can fill rooms that would otherwise sit empty without requiring the six-week planning window that fly-in guests need. Household composition: Family travelers have a fundamentally different booking calendar than couples or solo travelers. Kids' school schedules create hard peaks and valleys that families cannot easily shift. Couples and solo travelers are far more flexible. Off-peak campaigns targeted at non-family segments where flexibility is higher convert at better rates. Interest-based segments: If you capture interests at sign-up or can infer them from browsing behavior, segment campaigns by activity type. A hiking-focused promotion for a mountain destination reaches the right audience for a fall foliage campaign. A spa and wellness promotion reaches a different segment who might be less interested in the hiking itinerary. **The Off-Peak Campaign Structures That Work** Not all off-peak campaigns take the same form. The structures that consistently perform: Flash sales with genuine scarcity: A 48-72 hour window with specific inventory available at a specific discount. The scarcity has to be real. If your flash sale inventory is always available at that price, subscribers learn that and stop acting urgently. Real flash sales pull forward decisions from subscribers who were already considering a trip. Shoulder season positioning: Some destinations have a compelling case for the shoulder season that isn't being made explicitly. "September in [destination] has the same weather as July with 40% fewer crowds and better availability at top restaurants" is a genuine value proposition that repositions a traditionally slow period as a benefit, not a compromise. Value-add bundles instead of discounts: Adding a restaurant credit, activity voucher, room upgrade, or late checkout to a booking maintains rate integrity while adding perceived value. For premium travel brands, this approach is often preferable to price discounts that undermine positioning. Last-minute availability: Genuinely last-minute availability (7-14 days out) with a clear, honest message ("we have rooms open this [specific dates], here's a reason to come now") performs well with flexible, nearby subscribers who can act quickly. **Campaign Timing and Planning** The timing mistake most travel brands make with email is building campaigns around when they need bookings rather than when subscribers are making decisions. For summer travel, decisions are being made in January through March. An email campaign sent in May for summer availability is catching most buyers after they've already committed elsewhere. For holiday travel, decisions happen in August through October. By November, most holiday travel is booked. For shoulder and off-peak, the planning window is shorter (because the urgency is lower) but the campaign still needs to reach subscribers 4-8 weeks before the travel period begins, not during it. Build your email calendar backward from the travel period. A September shoulder season campaign needs to start in mid-July. A January off-peak push needs to start the day after Christmas, when people are sitting at home feeling restless and ready to plan their next escape. The [travel attribution post](/insights/travel-marketing-attribution-47-day-booking-journey) covers how the 47-day average booking window in travel affects channel strategy, which connects directly to this planning calendar logic. **Subject Lines for Travel Email** Travel email subject lines fall into predictable traps: urgency language that's been overused ("Last chance!" "Don't miss this!"), generic destination names without context, and discount-first framing that immediately trains subscribers to expect deals. The subject lines that perform consistently in travel email: Specificity about what's available: "3 nights in [specific property type] from $[specific price]" outperforms "Great deals this fall." The specific offer is scannable and immediately relevant or irrelevant to the subscriber. Seasonal framing that reframes slow season as advantage: "Why September might be the best time to visit [destination]" is a curiosity-driver that doesn't lead with a price. Time and place specificity: "[Destination] in [specific month]: what to expect" performs well because it's search-like and immediately relevant to anyone planning travel to that destination in that period. Personal tone for re-engagement: "You've been to [destination] before. Here's what's changed." This works for past guest segments because it acknowledges the relationship. **Measurement for Travel Email Campaigns** Open rate and click rate are insufficient measures for travel email performance. The metrics that matter are booking conversion rate (what percentage of recipients placed a booking) and revenue per email sent (total booking revenue divided by list size for that send). For off-peak campaigns specifically, benchmark against your historical off-peak period without campaigns. The question is not whether these campaigns beat your summer revenue, it's whether they beat what you would have done in the off-peak period without running them. Attribution is complicated in travel because bookings may come through direct links in email, through the website after a subscriber closes the email and searches later, or through phone calls that can't be tracked to the email. Use a combination of tracked links, UTM parameters, and booking code attribution (unique codes included in emails that guests use at checkout) to build a cleaner picture. **Key Takeaways** - Marketing during peak season captures demand that exists. Off-peak email campaigns create demand that wouldn't otherwise materialize. - Segment your list by past travel timing, geographic proximity, household composition, and interest type before building off-peak campaigns. - Campaign planning needs to happen 4-8 weeks before the travel period, not during it. Build your email calendar backward from the travel dates. - Flash sales work when the scarcity is real. Value-add bundles are often preferable to discounts for premium travel brands. - Subject lines with specific offers, dates, and pricing outperform generic urgency language like "Last chance" and "Don't miss." - Measure off-peak campaigns against your historical off-peak baseline, not against peak season performance. --- # Why Most Real Estate Agent Blogs Fail (And What Works Instead) URL: https://brianroseman.com/insights/real-estate-agent-content-marketing Published: 2026-05-06 Agent blogs fail because every agent writes the same posts. The ones that build real search presence use neighborhood-specific data, honest local perspective, and topics no national site can compete on. **Summary:** Most real estate agent blogs fail for a reason that has nothing to do with writing quality. They fail because every agent writes about the same topics in the same way, producing a content catalog that Google cannot distinguish from ten thousand competitors. This post covers the content strategy, topic selection, and distribution approach that builds genuine search presence and client trust for real estate agents and small brokerages. **Why Most Agent Blogs Produce Nothing** Pick any real estate agent blog from any market and you'll find some combination of these same posts: "5 Tips for First-Time Home Buyers," "How to Stage Your Home for Sale," "Questions to Ask Your Realtor," "Spring Real Estate Market Update." These posts exist in such abundance on the internet that Google has no reason to send a single visitor to an individual agent's version of them. This isn't a problem with real estate content broadly. It's a problem with non-specific content. The agents who build real search traffic have made a choice that most agents avoid: they write about their specific market, in their specific price range, for their specific buyer or seller profile, in enough detail that no other agent has produced the same content. The shift is from "tips for home buyers" to "buying a home in Leawood in 2026: what the data actually shows about which neighborhoods have multiple offers." The first post is forgettable. The second is useful to exactly the person who is considering buying in Leawood and has enough specific detail to rank for local searches that nobody else is targeting. **The Neighborhood Content Strategy** The single most underexplored content opportunity for most real estate agents is neighborhood-level content that goes far beyond the standard "great schools, friendly neighbors, close to downtown" boilerplate. What agents who rank in local real estate search have in common: they publish neighborhood content that includes actual data. Not "prices are up," but "median sale price in Prairie Village was $485,000 in Q1 2026, up from $461,000 in Q1 2025, with an average of 18 days on market and 94% of homes selling above asking." These numbers come from your MLS. You have access to them. Most agents don't publish them because it takes time, but the agents who do own the local search results for high-intent neighborhood queries. Neighborhood content should also include things that aren't on Zillow. The commute time to the major employers in the market. Which streets have significant traffic noise. Which blocks have the best walkability. Which neighborhoods are seeing teardowns and new construction, which affects resale value differently than established stock. This kind of local knowledge is impossible for national real estate sites to replicate. It's your competitive advantage. Update neighborhood market reports quarterly at minimum. Stale data (a market report from 2024 sitting on your site in 2026) signals to potential clients that your content isn't maintained, which isn't the trust signal you're trying to build. **The Search Topics That Drive Real Estate Traffic** Beyond neighborhood content, the search topics that actually drive traffic and leads for real estate agents are ones where the person searching has a specific, local question: "How much are property taxes in [city/county]?" This is searched by buyers trying to understand total cost of ownership. It's a completely answerable, local question that most agent sites don't address at all. "[Neighborhood] homes for sale under [price]" searches drive traffic that most agents try to capture through IDX integration rather than content, but IDX pages are often thin and rank poorly. A hybrid approach, an editorial piece about what you can buy in a specific neighborhood at a specific price point right now, with links to live IDX listings, captures both content ranking and IDX utility. "Is [neighborhood] a good place to live?" This is a top-of-funnel search from people who haven't decided where to buy yet. An honest, specific answer ("yes, if you prioritize X, but not if Y is important to you") positions you as a trusted advisor rather than a salesperson. "How long does it take to close on a house in [state]?" Process questions like this get searched heavily by first-time buyers who are anxious about the timeline. These posts build trust and often show up in AI overviews because they're answerable, specific, and local. **Building a Content Calendar That Stays Consistent** The most common content failure for agents isn't quality, it's inconsistency. Two posts in January, nothing for three months, then a burst in spring market season, then silence. This pattern produces none of the compounding benefits that consistent content creates. A realistic content calendar for a solo agent or a small team looks like this: One substantive piece per month (1,500+ words): neighborhood market update, "what you can buy for $X in [market]" piece, or a deep-dive on a specific buying or selling situation relevant to your clients. Monthly market report: A structured update on your core market with current data. This can follow a consistent format month to month, which makes it faster to produce. Occasional event-driven content: When interest rates move significantly, when a major employer announces a relocation, when a new development breaks ground in your market. These pieces earn links and shares because they're timely and locally specific. This pace is sustainable and produces a meaningful body of content over 12-24 months. Trying to produce five posts per week leads to thin, generic content that doesn't rank. **Distribution for Real Estate Content** Real estate content distribution works differently from B2B content. The primary channels: Organic search is the long-term play. Neighborhood content with specific data can rank for local real estate queries and drive inbound traffic for months or years after publication. Email to your past client list. Your sphere of influence is the most valuable marketing asset a real estate agent has. A monthly email with your market update content keeps you top of mind with past clients who will refer you when their friends and family are looking to buy or sell. This doesn't need to be a polished newsletter. A short personal note with the data and a link to your full report works fine. Local Facebook groups. Many markets have active neighborhood Facebook groups where residents share local information. Market updates and neighborhood content shared authentically (not promotionally) in these groups often get significant organic reach. The rules vary by group; check before posting. Instagram and video content. Neighborhood walkthroughs, "what this house sold for vs. listing price" real-time commentary, and market data visualizations work well on Instagram and short-form video. This requires a different format than written content but builds a different kind of trust, one that's more personal and visible. For deeper insight into how local search connects to real estate lead generation, the [real estate PPC geo-targeting post](/insights/real-estate-ppc-geo-targeting-local) covers the paid side of the same local visibility strategy. **What Makes Real Estate Content Actually Convert** Content that generates real estate leads shares a few characteristics that generic content doesn't have. It demonstrates specific knowledge. "Here's what I'm seeing in this specific neighborhood right now" converts better than "here are some tips for buyers." The specificity signals that you're the expert for that neighborhood, which is what buyers and sellers are actually looking for in an agent. It shows your perspective. "Honest answer: this neighborhood is great for young families but the commute to downtown is legitimately bad" is useful and memorable. "This neighborhood has great amenities" is forgettable. It contains a clear, low-pressure next step. Not "CALL ME NOW," but "if you're thinking about buying in this area, I put together a quick [guide or tool](/tools/marketing-assessment) on how to evaluate neighborhoods based on your priorities. Happy to share it." Content that educates and demonstrates expertise without pressuring creates the kind of trust that produces referrals, which is where the majority of real estate business actually comes from. **Key Takeaways** - Generic real estate content ("5 tips for buyers") doesn't rank or convert. Specific, local, data-driven content does. - Neighborhood market reports with actual MLS data (median sale price, days on market, sale-to-list ratio) give you content no national site can replicate. - Answer the specific, local questions buyers and sellers are searching: property tax rates, commute times, what you can buy at a specific price point right now. - One substantive piece per month plus a monthly market report is a sustainable and effective publishing cadence for a solo agent. - Email to past clients with your market content keeps you top of mind and generates referrals. - Specificity and honest perspective convert better than generic positive framing. --- # Local SEO for Law Firms: Ranking in Competitive Markets URL: https://brianroseman.com/insights/local-seo-law-firms-competitive-markets Published: 2026-05-02 Law firm local SEO is competitive, heavily regulated, and takes longer than most agencies will tell you. Here is the Google Business Profile strategy, review system, and content depth that actually moves the needle. **Summary:** Law firm local SEO is brutally competitive, especially in practice areas like personal injury, criminal defense, and family law. The firms that dominate local search in competitive markets don't do it through tricks. They do it through a combination of Google Business Profile optimization, consistent citation building, review volume and quality, and content depth that general-purpose local SEO guides rarely cover in enough specificity for legal. This post covers what actually works. **Why Legal Local SEO Is Different** A few things make law firm local SEO more complex than most local business categories. First, the bar advertising rules. Every state bar has rules about attorney advertising that affect what you can and cannot say on your website, in reviews, and in your Google Business Profile. Claiming to be the "best DUI lawyer in Kansas City" may violate state bar rules on comparative advertising. Displaying client testimonials in some states requires specific disclaimers. Run everything through your bar's advertising rules before publishing. Second, the geographic mismatch between where law firms are licensed and where they want to appear in search. A firm licensed in Missouri and Kansas wants to appear for searches across the entire Kansas City metro, but its physical office address anchors its local search presence to a specific neighborhood. Expanding local search presence requires a deliberate strategy. Third, the trust stakes are unusually high. Someone searching for a criminal defense attorney or a divorce lawyer is often in one of the most stressful situations of their life. Trust signals matter more here than in almost any other service category. **Google Business Profile: The Foundation** Your [Google Business Profile](https://business.google.com) is the single most important factor in local search ranking for legal. The firms that rank in the local pack almost universally have complete, active, regularly updated GBPs. The elements that matter most: Category selection: Choose your primary category precisely. "Personal Injury Attorney" and "Criminal Justice Attorney" are different categories. You can select multiple categories; put your highest-volume practice area first. This is the most underutilized lever in legal GBP optimization. Service area settings: Beyond your office location, add the specific cities and counties in your service area. This expands the geographic scope of your local search presence. Services list: Add every practice area as a service with its own description. These descriptions index in local search and give Google more signal about what queries your firm should appear for. Q&A section: Seed it yourself. The questions that potential clients actually ask ("Can I afford an attorney?" "How much does a DUI lawyer cost?" "What happens if I don't hire a lawyer?") answered in your GBP Q&A build trust and potentially appear in local search results. Posts: Weekly GBP posts keep your profile active. Short case updates (anonymized and with appropriate disclaimers), legal news commentary, and "what to know if you've been charged with X" posts work well. **Reviews: Volume, Quality, and Response Strategy** Review volume and rating are among the most significant ranking factors in local legal search. A firm with 200 reviews at 4.8 stars will outrank an equally qualified firm with 30 reviews at 5.0 stars in most competitive markets. Building review volume requires a systematic process. After a case closes positively (with appropriate ethical considerations about timing and messaging), a structured outreach asking the client to share their experience gets reviews at a much higher rate than hoping clients will do it voluntarily. The review request message matters. A text or email that says "would you be willing to share your experience?" converts better than "please leave us a Google review," because it feels like a personal request rather than a marketing ask. Include a direct link to your GBP review page. Responding to every review, positive and negative, signals to Google that your business is active and managed. For negative reviews, the response matters as much for prospective clients reading it as for the person who left the review. A professional, empathetic response to a negative review that doesn't reveal client information often turns a liability into a trust signal. One clear rule for legal: never respond to a negative review with information that could identify the reviewer as a client or reveal anything about their legal matter. This can create confidentiality problems. **Citation Building and NAP Consistency** Citations, your firm's Name, Address, and Phone number, listed on legal directories and general business directories, remain a significant local ranking factor. The directories that matter most for legal: - Avvo - Martindale-Hubbell - FindLaw - Justia - Lawyers.com - Yelp (matters more than most attorneys expect) - BBB - Bing Places - Apple Maps For citation impact, NAP consistency matters. If your address is listed as "123 Main Street, Suite 400" on your website but "123 Main St #400" on Avvo, this inconsistency dilutes your local search signals. Audit your citations with a tool like [BrightLocal](https://www.brightlocal.com/) and standardize the format. **Content Depth for Local Legal Search** Most law firm websites have thin practice area pages that say essentially nothing: "We handle personal injury cases in Kansas City. Call us." These pages rank poorly because they give Google no reason to believe the firm has depth of expertise in the area. The content that ranks in competitive legal markets is specific, detailed, and addresses the questions that potential clients are actually searching: Practice area hub pages: Each practice area gets a comprehensive page (2,000+ words) covering how those cases work in your jurisdiction, what the process looks like, what affects the outcome, and what questions clients should ask. This is the content that earns rankings. Location pages: If you serve multiple cities in a metro, individual location pages for each city (not identical content with the city name swapped, genuinely different content about legal landscape in that jurisdiction) help expand geographic presence. FAQ content: "What is the average settlement for a car accident in Missouri?" "How long does a criminal case take in Johnson County?" These are real questions with specific, local answers that you can address authoritatively in ways that national legal content sites cannot. Blog content on recent law changes: When Missouri updates its comparative fault rules, or Kansas changes DUI sentencing guidelines, you have a window to publish authoritative local content before anyone else does. For [legal SEO services](/services/legal-seo-consultant), the content strategy is often where the biggest gains come from, because most competing firms have the same thin practice area pages and identical GBP setups. Content depth is the differentiator. **Technical SEO for Law Firm Websites** Two technical issues show up consistently in law firm site audits: Mobile performance: The majority of personal injury and family law searches happen on mobile. Law firm websites often load slowly on mobile because they're image-heavy, use large hero videos, or weren't built with mobile-first performance in mind. Core Web Vitals scores below 50 on mobile are a consistent finding and a real ranking factor. Local business schema: LegalService schema markup, or at minimum LocalBusiness schema, should be implemented on the homepage and contact page. Include your practice areas, address, service area, and attorney information in the schema. FAQPage schema on practice area pages with legitimate local FAQs has produced featured snippet placements for law firms in several markets. **How Long Does It Take?** Local SEO results in legal are not immediate. In moderately competitive markets (smaller cities, less common practice areas), meaningful movement in local pack rankings takes 3-6 months of consistent effort. In highly competitive markets (personal injury in major metros), the timeline extends to 12-18 months for significant ranking changes in the local pack. The common mistake is starting the work, seeing no results at 60 days, and stopping. The firms that dominate legal local search in competitive markets are usually the ones that have been doing this consistently for 18-24 months. **Key Takeaways** - Google Business Profile optimization is the highest-leverage action in legal local SEO. Complete every section, add all practice areas as services, and post weekly. - Review volume matters more than perfect rating. A system for requesting reviews after case resolution is essential. - Citation consistency across legal directories and general business directories is a meaningful ranking factor. Audit with a tool and standardize your NAP. - Practice area pages need substantial depth (2,000+ words with local specifics) to rank in competitive legal markets. - LegalService schema markup and FAQPage schema on practice area pages improve rich result eligibility. - In highly competitive markets (personal injury in major metros), expect 12-18 months to see significant local pack movement. --- # Free Trial to Paid: Conversion Optimization for B2B SaaS URL: https://brianroseman.com/insights/free-trial-paid-conversion-saas Published: 2026-04-29 Most SaaS trial conversion problems are onboarding problems. Users who reach the activation event convert. Users who do not, almost never do. Here is the strategy that moves trial users to paid. **Summary:** Most SaaS companies treat free trial conversion as a product problem. It's usually a marketing and onboarding problem. Users who don't convert to paid during a trial almost never did the one or two things that would have shown them the product's value. This post covers the conversion strategy, messaging sequence, and friction-reduction tactics that move trial users to paid subscribers. **The Real Reason Free Trials Fail to Convert** The industry benchmark for free-to-paid trial conversion in B2B SaaS is roughly 15-20% for sales-assisted trials and 2-5% for pure self-serve. Those numbers don't tell you why conversions happen or don't happen, but years of data on trial behavior point to a clear pattern: users who reach a specific activation event convert. Users who don't reach that event almost never do. The activation event is the specific action in your product that correlates most strongly with conversion. Every product has one, usually one or two. For a project management tool it might be "invited a team member." For an analytics tool it might be "ran their first custom report." For a marketing automation platform it might be "sent their first automated email." If you don't know what your activation event is, find it. Pull cohort data comparing converted trial users to churned trial users. Find the actions that diverged most sharply between the two groups in the first three days of the trial. That's your activation event. Everything in your trial conversion strategy should be aimed at getting users to the activation event faster. **The Onboarding Gap Most SaaS Companies Have** When a new trial user signs up, most products do one of two things: they show a generic product tour, or they drop the user directly into the full application with no guidance. Both approaches fail for the same reason: they don't help the user accomplish the specific thing that will demonstrate the product's value to their situation. The onboarding approach that works is guided activation. When a user signs up, ask them two or three questions about their goal (not generic, specific: "What would a win look like for you in the first 30 days?") and use their answers to customize the first-session experience toward the actions most likely to reach activation. This isn't a complicated product change. It can be implemented with a simple onboarding checklist, in-app tooltip sequences, or a triggered email series. What makes it work is that it's oriented toward the user's goal, not toward a generic product tour. **The Email Sequence That Drives Trial Conversion** The email sequence during a free trial is the single highest-leverage conversion tool most SaaS companies aren't using well. Most trial email sequences fall into one of two categories: daily product tips that have nothing to do with what the user is trying to accomplish, or conversion pressure emails ("Your trial ends in 48 hours!") sent to users who never reached activation. The sequence that works is behavior-triggered, not time-triggered, for the first part of the trial. Days 1-3: Trigger emails based on what the user has and hasn't done, not based on a calendar. If a user signed up and then didn't log in for 24 hours, send a "here's the one thing to do first" email. If they logged in but didn't reach activation, send a guide specifically to that step. If they reached activation in day 1, skip the onboarding emails entirely and move to value-expansion content. Day 7 (or midpoint): Product value email. This is an appropriate point to communicate what users who become customers accomplish with the product, using real, specific outcomes. Not "companies like yours see results" but "users who reach [activation event] in their first week typically accomplish [specific outcome]." Days 3-5 before trial end: Start your conversion pressure sequence, but only for users who have reached activation. Users who haven't reached activation need a different sequence focused on getting them there, not pressuring them to pay before they've seen value. Final 48 hours: Your clearest conversion offer. If you have pricing flexibility, this is where a first-year discount or extended trial for users who haven't activated can help. If your pricing is non-negotiable, this is where you provide the clearest possible comparison of free versus paid to help users who have already decided. **Pricing Page Optimization for Trial Conversion** A trial user visiting your pricing page is one of the highest-intent actions in the entire funnel. Most pricing pages fail them in three ways. First, the pricing is confusing. More tiers than necessary, features spread across tiers in ways that require a matrix to understand, and unclear explanations of what distinguishes one plan from the next. Simplify. Two or three tiers with clear, simple feature distinctions convert better than five tiers with a comparison table that nobody reads. Second, the page doesn't acknowledge that the visitor is a trial user. Show them what they currently have access to and what they'd gain by upgrading. "You're on the free plan. Here's what Pro adds for your specific use case" is more compelling than a generic pricing page that treats every visitor the same. Third, the page lacks social proof relevant to their industry or company size. A solo founder and an enterprise team leader have completely different price sensitivity and risk tolerance. If you're targeting multiple segments, show proof from each. "Used by teams at [company type] to accomplish [outcome]" is more persuasive than a generic logo wall. For [B2B SaaS SEO](/services/b2b-saas-seo-consultant), the pricing page also matters for search. "Tool X pricing" is one of the highest-intent search queries in most SaaS categories. Your pricing page should be optimized for this query, not hidden behind a demo gate. **The Conversion Conversation: When to Involve Sales** Pure self-serve trials work for simple products under a certain price point. Above roughly $500/year in ACV, most SaaS companies see significantly higher conversion when a human is involved at some point in the trial. The question is when and how. Most SaaS sales teams make the mistake of reaching out too early, before the user has had a chance to explore the product, which makes the outreach feel like a pressure call. The optimal timing for trial outreach is after the user has reached the activation event but before the end of the trial. This means your CRM needs to receive behavioral signals from your product, specifically the activation event completion. When a user reaches activation, trigger a sales task or automated meeting request. "I saw you set up [feature]. Most users in your situation find it helpful to walk through [next step] with someone from our team, would 20 minutes be useful?" That framing works because it's responsive to what the user actually did, not a generic "can we hop on a call" request. **Extending Trials vs. Converting Immediately** Should you offer trial extensions to users who are active but haven't converted? Sometimes. The problem with extensions is that they create a habit of delay. A user who got one extension will often expect another. The better approach for active-but-haven't-converted users is to diagnose what's blocking them. The most common blockers are: needs internal approval, hasn't had time to evaluate fully, has a specific feature concern, or is comparing you to a competitor. A targeted outreach email or chat that identifies which of these applies is more effective than an automatic extension. Use extensions selectively for users who are clearly engaged and clearly blocked by something external (budget cycle, procurement process). For users who are disengaged, an extension just delays the inevitable. **Key Takeaways** - Find your activation event by comparing converted versus churned trial users. Drive every user to that event in the first three days. - Use behavior-triggered emails during the trial period, not just time-based sequences. - Simplify your pricing page to two or three tiers with clear feature distinctions. Acknowledge that the visitor is a current trial user. - For products with ACV above roughly $500/year, sales involvement after activation significantly improves conversion. - Conversion pressure emails sent to users who haven't reached activation don't convert them. Send a re-engagement email to get them to activation first. - Trial extensions should be selective, not automatic. Diagnose the blocker before offering more time. --- # Ecommerce Win-Back Email Campaigns That Actually Reactivate Customers URL: https://brianroseman.com/insights/ecommerce-win-back-email-campaigns Published: 2026-04-25 Win-back campaigns consistently outperform acquisition on cost per order, but most brands run them as an afterthought. Here is the sequencing and personalization approach that actually reactivates lapsed customers. **Summary:** Win-back email campaigns targeting lapsed customers consistently outperform acquisition campaigns on cost per order, yet most e-commerce brands run them as an afterthought. A customer who bought from you once already trusts you enough to pay. Getting them to buy again is cheaper and faster than convincing someone new. This post covers the strategy, sequencing, and copy approaches that make win-back campaigns actually work. **Why Win-Back Is One of E-commerce's Best ROI Opportunities** Acquiring a new customer costs significantly more than re-engaging a lapsed one. The exact ratio varies by industry, but the structural advantage is clear: a lapsed customer already knows your brand, has converted at least once, and has demonstrated a purchase history you can use to personalize your approach. The problem is that most e-commerce teams treat the lapsed customer segment as a low-priority maintenance task. They send one or two generic "we miss you" emails with a discount, see mediocre results, and conclude that win-back doesn't work. The actual conclusion is that generic win-back doesn't work. What does work is segmented, sequenced win-back with personalized content based on what the customer actually bought, when they bought it, and what their lifetime value history shows. **Defining Lapsed: When Does the Clock Start?** The definition of "lapsed" varies by product category and purchase frequency. This is a critical distinction most brands ignore. For a consumables brand (supplements, skincare, pet food) where the expected repurchase cycle is 30-45 days, a customer who hasn't bought in 90 days is significantly lapsed. For a furniture brand where someone might buy once every three to seven years, a 90-day lapse means nothing. Build your lapse threshold from your actual purchase frequency data. Calculate the median time between first and second purchase for customers who went on to become multi-purchasers. The lapse window typically starts at 1.5 to 2 times that median repurchase interval. If your median time between purchases is 60 days, start your win-back sequence at 90-120 days of inactivity. Going earlier means you're interrupting people who were going to buy anyway. Going too late means you've already lost them. **The Three-Stage Win-Back Sequence That Works** Effective win-back sequences have three distinct stages, each with different goals: Stage 1 (Days 1-3 of sequence launch): Re-engagement without a discount. The goal here is to see if the customer is still interested before you discount. Lead with the product category they bought from, highlight new arrivals or improvements, and keep it clean. Subject lines like "New in [category they bought]" or "Thought you'd want to see this" work better than anything that signals you know they've been gone. Send one email in this stage. Stage 2 (Days 7-10): Soft incentive. If they didn't engage with stage 1, offer a meaningful but not desperate incentive. A 10-15% discount or free shipping is appropriate here. Frame it around value, not urgency. "Here's something for your next order" feels different from "LAST CHANCE 40% OFF" to a customer who shopped with you because they trusted your brand. Stage 3 (Days 14-21): Decision point. This is your last attempt before moving the contact to a suppression list. Make a clear, compelling offer, your best incentive, and be direct: "We'd love to have you back" is honest and respects that they have a choice. If they don't respond to this, remove them from active marketing. Continuing to mail a permanently disengaged subscriber hurts your deliverability. After stage 3 with no response, do not delete the contact. Move them to a deep suppression segment with only annual or semi-annual reactivation attempts. Some lapsed customers return years later when their life situation changes. **Personalization That Makes a Real Difference** The gap between generic and effective win-back is personalization at the product-category level. You don't need to personalize to the individual product (though that helps for replenishable items); you need to personalize to the department or category. A customer who bought workout equipment should get different win-back content than one who bought kitchen goods. This sounds obvious but requires that your email platform can segment by purchase category, which means your order data needs to flow into your ESP cleanly. Using [Klaviyo](https://www.klaviyo.com/) or a similar platform with native Shopify or WooCommerce integration, you can build these flows from purchase history without manual list management. The flow logic: "Customer purchased from [category], has not purchased in [X days], send sequence A." For high-value customers (top 20% by lifetime value), treat win-back differently. These are accounts worth a personal outreach from customer service or a stronger incentive. A customer who spent $2,000 with you over three years warrants a different recovery effort than one who placed a single $30 order. **Subject Lines That Actually Get Opens** Win-back subject lines fail for one main reason: they're too transparent about being a win-back email. "We miss you!" and "It's been a while..." announce that this is a re-engagement email, which signals to the subscriber that this is automated, not relevant. The subject lines that get opens in win-back sequences are ones that look like they could be from any other email in your marketing calendar: - "[New product in category they bought]" - "A few things have changed since you last visited" - "Your [product category] favorites are back in stock" - "For customers who bought [product name]" (works because it implies relevant content, which it should have) Curiosity works better than nostalgia. Nostalgia-based win-backs ("Remember us?") feel like marketing. Curiosity-based ones ("We added something you'd probably like") feel like information. **The Discount Decision** Whether and when to discount in win-back sequences is genuinely contested among e-commerce marketers. The argument against leading with discounts is that it trains customers to wait for deals and devalues your brand. The argument for is that lapsed customers are comparing you to alternatives and a financial nudge helps tip the decision. The middle position that tends to work best: no discount in stage 1, a modest discount (10-15%) in stage 2, and your best offer in stage 3. This preserves brand value for customers who would have returned anyway while giving you a conversion mechanism for the harder cases. Do not offer 40% or 50% discounts in win-back sequences unless your margins can genuinely sustain it. A customer reactivated at a 50% discount who expects that discount on future purchases creates a unit economics problem. The goal is to reactivate at a margin that makes the customer profitable over their next few orders. **Tracking Win-Back Performance** The right metrics for win-back campaigns are not open rate and click rate. Those are inputs. The metrics that matter are: Reactivation rate: the percentage of lapsed customers in the sequence who place an order within 90 days of the sequence start. This is your primary KPI. Revenue per contact: total revenue generated divided by the number of contacts in the sequence. This tells you the dollar value of running the sequence. Post-reactivation LTV: how much do reactivated customers spend in the 12 months after reactivation, compared to newly acquired customers? If win-back customers churn again quickly, your sequence may be reactivating the wrong people. Check your [LTV calculator](/tools/ltv-calculator) to understand how reactivation economics compare to acquisition costs in your specific business. **When Win-Back Sequences Fail** Win-back fails most often because of deliverability problems. If you're mailing a large segment of people who haven't engaged in 12-18 months, a significant portion of those contacts are no longer valid email addresses. Sending to invalid addresses increases your bounce rate, and sending to addresses that have marked previous emails as spam trains inbox providers to filter your messages. Before launching a win-back sequence, run your lapsed segment through an email verification service. Remove hard bounces and known spam traps. Accept that you'll lose some contacts, but preserve your deliverability for the contacts who are still valid. The other common failure mode is messaging that doesn't match why the customer left. If customers lapsed because of a bad product experience, a discount doesn't fix the problem. If they lapsed because they found a cheaper alternative, a 10% discount may not be enough to overcome a structural price disadvantage. Win-back works best when the reason for lapse is inertia ("I just forgot about you"), not dissatisfaction or competitive loss. **Key Takeaways** - Define your lapse threshold based on your actual median repurchase interval, not an arbitrary time period. Start the sequence at 1.5 to 2 times that interval. - Three-stage sequences work best: re-engagement without discount, soft incentive, then best offer as a final attempt. - Subject lines should look like regular marketing emails, not win-back emails. Curiosity works better than nostalgia. - Personalize by purchase category at minimum. High-LTV customers warrant stronger efforts and different incentives. - Run lapsed segments through email verification before launching to protect deliverability. - Measure reactivation rate and revenue per contact, not open rate. --- # Technical SEO for Industrial Product Pages: The Specification Indexing Problem URL: https://brianroseman.com/insights/technical-seo-industrial-product-pages Published: 2026-04-22 Most industrial product pages are nearly invisible in search, not because of penalties but because specifications live in JavaScript-rendered tabs or PDFs that Google cannot read. Here is how to fix it. **Summary:** Industrial manufacturers have some of the most complex product catalogs in any industry, and most of them are nearly invisible in search. Thousands of SKUs, dense technical specifications, and product pages written for engineers rather than search engines create a gap that competitors consistently fail to close. This post covers the specific SEO issues that hurt industrial product pages and what to do about them. **Why Manufacturing SEO Is Harder Than It Looks** Consumer e-commerce SEO has produced mountains of frameworks, tools, and case studies. Manufacturing SEO has almost none of that. The customer journey is fundamentally different: a procurement engineer searching for a hydraulic coupling doesn't browse the way a consumer shops for shoes. They search with extreme specificity, often using part numbers, tolerances, or standards rather than common-language product names. This creates both a challenge and an opportunity. The challenge is that most SEO tools and keyword research frameworks assume consumer-style search behavior. The opportunity is that if you get your technical specifications indexed correctly, you can capture search demand that your competitors haven't even tried to address. I've worked with manufacturers whose product pages were returning 0 impressions in Google Search Console despite having technically valid pages. The problem wasn't penalization, it was invisibility, pages so thin on text content and so heavy on rendered-in-JavaScript specifications that Googlebot couldn't read what the page was about. **The Specification Indexing Problem** Most industrial product pages display specifications in one of three formats, and only one of them is reliably indexed: - Static HTML tables: Google can read these without problem. - JavaScript-rendered specification tabs: Google can often read these, but there's a rendering delay and some specifications may be missed. Never assume JavaScript-rendered content is indexed without verifying in Search Console. - PDFs linked from the product page: Google indexes PDFs, but the content is attributed to the PDF, not to your product page. A specification that only exists in a downloadable PDF is invisible at the product page URL. The fix is to ensure your most important specifications are in static HTML on the page. You don't need to eliminate JavaScript interactivity, you need to make sure the specification data isn't exclusively inside a JavaScript component that requires execution to render. Pull your top 50 product pages in Search Console. Look at the "Inspect URL" tool and check the rendered HTML. Compare what you see in the rendered version to what the page shows in a browser. Any specifications that appear in the browser but not in the rendered HTML are invisible to Google. **Keyword Research for Industrial Products** Standard keyword research tools don't capture the way engineers search. Search volume estimates for highly technical queries are often wildly inaccurate because the query volume is low but the purchase value is enormous. For [manufacturing SEO](/services/manufacturing-seo-consultant), the right approach to keyword research includes: Part number targeting: Your own part numbers and your competitors' part numbers that your product is a drop-in replacement for are often searched directly. Many manufacturers don't have their own part numbers indexed, much less competitors' equivalent numbers. Standards and specifications as keywords: "ANSI B16.5 flanges" is how engineers search. "ASME pressure vessel" is a phrase that lives in your product documentation but often not in your page copy or title tags. Every major standard your products comply with is a keyword. Application-based search: "hydraulic coupling for food processing" is how engineers who don't know part numbers search. Map your applications to the ways customers would describe them, not the ways your internal catalog describes them. Tools like [Semrush](https://www.semrush.com/) and Ahrefs can surface these query patterns if you seed them with technical terms rather than generic product category names. Start with the language in your existing product documentation and expand from there. **Page Structure That Works for Industrial Products** The page structure that works best for industrial product pages gives Google clear signals about what the product is and who it's for, while giving the purchasing engineer the specification depth they need to make a decision. Title tag structure: Product Name, Key Specification, Standard or Application, Brand. "12-Inch Carbon Steel Gate Valve, ANSI Class 150, Industrial Service" tells both Google and a purchasing engineer exactly what the page is about. H1: Match the title tag language closely. First 150 words: Write a plain-language description of what the product is, what it does, and what applications it's designed for. Do not lead with company history or marketing language. Engineers scanning product pages need to know immediately whether the page is relevant to their search. Specification table: In static HTML. Include all tolerances, materials, pressure ratings, temperature ranges, and compliance standards. If your engineers use specific terminology for these specifications, use that terminology on the page. Application notes: A paragraph or two on the specific applications, industries, and installation conditions the product is suited for. This is where application-based keywords live naturally. Related products and accessories: Internal linking to compatible fittings, replacement parts, and alternative models keeps engineers on your site and distributes link equity. Downloads: PDFs, CAD files, and compliance certifications should be linked from the product page, not hidden inside a downloads portal. A link from a strong product page transfers some authority to the PDF and also signals to Google what type of document it is. **Technical Infrastructure Issues That Kill Industrial SEO** Product catalog pages that use faceted navigation, where engineers filter by material, pressure rating, size, or standard, create duplicate content at scale. A product accessible at five different filter combinations produces five URLs with nearly identical content. Without proper canonical tag implementation or parameter handling in Google Search Console, these multiply your crawl budget requirements while splitting link equity. Canonical tags are the right solution for most faceted navigation implementations. The canonical URL should point to the base product page, not the filtered URL. Check whether your PIM (Product Information Management) system or e-commerce platform handles this automatically. Crawl budget matters more for large industrial catalogs than for most websites. If you have 50,000 product pages and Google is only crawling 5,000 per month, your new and updated products aren't being indexed promptly. Diagnose crawl budget problems in Google Search Console under "Crawl Stats." Common causes: thin pages with duplicate content, excessive redirect chains, and server response times above 500ms. Site speed on industrial product pages is often poor because pages are loaded with specification tables, multiple product images, CAD file previews, and document downloads. Core Web Vitals scores matter here, not just for SEO directly, but because a slow page increases bounce rate among engineers who have no patience for slow-loading specification data. **Schema Markup for Industrial Products** Product schema is the right structured data type for individual product pages. It signals to Google that the page represents a purchasable product and enables rich results in search. For industrial products, the most valuable schema fields are: - name: The full product name including key specification - description: Plain-language product description - sku: Your part number - material: Materials of construction - brand: Your company name - aggregateRating: Customer or specification ratings if you have them - offers: Pricing if you publish pricing (many manufacturers don't) FAQPage schema added below your main product content gives you a second structured data opportunity. Common questions like "What temperatures is this rated for?" or "What is the pressure rating?" can be answered in FAQ format and may appear in featured snippets. For product families (a base product with multiple configurations), ItemList schema can help Google understand the relationship between variants and may surface multiple product variants in search results. **What an Industrial Manufacturer's SEO Roadmap Looks Like** The work tends to fall into three phases: Phase 1 (months 1-2): Technical audit and fixes. Crawl budget analysis, JavaScript rendering check on key product pages, duplicate content from faceted navigation, page speed improvements on high-traffic product categories. Phase 2 (months 2-4): Content enrichment on priority product pages. Identify the top 20% of product pages by traffic and revenue importance. Rewrite title tags, add application notes, ensure specification tables are in static HTML, add schema markup. Phase 3 (months 4+): Expand to long-tail application content. Build out application and industry pages that address how specific product families solve specific problems. "Valves for pharmaceutical clean rooms" is an example of the application-specific content that supports the product pages and captures upstream search traffic. Run a [free marketing assessment](/tools/marketing-assessment) if you want a starting point for identifying your biggest gaps before prioritizing the work. **Key Takeaways** - Verify that product specifications are in static HTML, not exclusively inside JavaScript-rendered tabs or PDFs. Google may miss JavaScript-rendered content. - Industrial keyword research should include part numbers, compliance standards (ANSI, ASME, ISO), and application-specific phrases engineers use. - Faceted navigation in product catalogs creates duplicate content at scale. Implement canonical tags pointing filtered URLs to the base product page. - Title tags for industrial products should include the product name, key specification, and relevant standard or application. - Product and FAQPage schema markup increases your chances of appearing in rich results for specification-related queries. - Prioritize the top 20% of product pages by traffic and revenue before trying to fix the full catalog. --- # Content Marketing for Wealth Management: What Actually Builds Trust URL: https://brianroseman.com/insights/content-marketing-wealth-management Published: 2026-04-19 Most wealth management content is too hedged to be useful or too generic to stand out. Here is the format, topic strategy, and distribution approach that builds real trust with high-value prospects. **Summary:** Wealth management is one of the most regulated and trust-dependent industries in marketing. A compliance review process, conservative clients, and long sales cycles make most content marketing efforts cautious to the point of uselessness. This post covers how wealth management firms and independent advisors can produce content that builds actual trust, survives compliance, and attracts qualified clients who stay for decades. **The Content Problem Most Advisors Have** Most financial advisor content falls into one of two traps. Either it's so hedged by compliance disclaimers that it communicates nothing, or it's so generic, "diversify your portfolio," "stay the course," that it's indistinguishable from every other firm's output. Neither approach builds authority. Neither gives a prospective client a reason to pick you over the advisor down the street. The fundamental shift in wealth management content is recognizing that trust is built through specificity, not caution. When you write about "clients like yours" with concrete situations and outcomes (without identifying anyone), you signal something compliance-safe generic content cannot: that you understand real problems. **What "Compliance-Friendly" Actually Means** Before diving into strategy, it's worth being precise about what compliance actually restricts. SEC and FINRA rules prohibit testimonials from clients about investment performance in most contexts (there are newer rules around written testimonials with disclosures that some advisors are beginning to use carefully). They restrict specific performance claims. They require disclosure of material conflicts of interest. What they do not restrict: educational content, market commentary, general financial planning principles, case studies with client consent and appropriate disclosures, and opinion pieces on financial topics that don't make performance predictions. Most advisors treat compliance as a broader restriction than it actually is. The question to ask your compliance officer is not "can I publish this?" but "what would need to change in this piece to make it publishable?" That reframe tends to produce better conversations. The firms with the strongest content marketing have a working relationship between marketing and compliance where content goes through a structured review process with defined turnaround times, not an indefinite review queue that kills publishing velocity. **The Audience Segments That Actually Read Financial Content** Wealth management content works differently depending on who you're trying to reach. Three distinct audiences with very different content needs: Pre-retirees, age 55-65, are your highest-value near-term conversion audience. They are actively researching questions like "when can I actually retire," "how do I optimize Social Security timing," and "what happens to my portfolio if we have another 2008." Content that gives real, specific answers to these questions, not hedged platitudes, earns attention and trust. Business owners are often underleveraged by advisors. They have concentrated wealth, complex tax situations, and succession questions that generic personal finance content doesn't touch. A piece on "what to do with equity when your business sells" speaks directly to a situation that no general-audience financial content addresses. High-earning younger professionals, generally age 35-50, are building wealth but haven't yet had a reason to hire an advisor. Content that answers their actual questions, like "does it make sense to max my 401k if I have student loans" or "how do I think about equity compensation from my employer," meets them where they are before they need full wealth management. Each segment reads differently. Pre-retirees engage with comprehensive long-form content. Business owners respond to case-study formats and sector-specific planning situations. Younger professionals engage with more tactical, specific posts that answer a single question completely. **Topics That Actually Drive Qualified Traffic** The content topics that attract wealth management prospects are specific, not general. "Investment strategies for 2026" attracts everyone and converts no one. These attract qualified traffic: Tax-loss harvesting: when it makes sense and when it doesn't. This is searched by people who are actively managing real investment accounts, often in the exact AUM range advisors want. Required Minimum Distribution planning. Searched heavily by people in or near retirement with substantial retirement accounts. 529 plan versus Roth IRA for college savings. Parents actively making this decision are a rich audience for advisors who specialize in family financial planning. Executive compensation and equity: RSU vesting strategies, concentrated stock positions, when to exercise options. This is a specialty topic where detailed content attracts people who are essentially self-qualifying, because only people with equity compensation plans would search these questions. Business sale planning. "What to do financially before selling my business" is a low-volume, extremely high-value search. I've seen advisors build significant new client pipelines from a handful of very specific, detailed pieces on topics like these, because no one else is writing content that actually answers the question. The generic "financial planning basics" space is completely saturated and dominated by major publishers. The specific, situational content is almost entirely open. **The Format That Builds Trust** The format that works best for wealth management content is not the listicle or the "top 10 tips" structure. It's what I'd call the "here's the situation, here's how to think about it" format. Start with a specific situation: "A client came to us six months before their company's acquisition was expected to close. They had 40% of their net worth in company stock." Then walk through the decision framework without the identifying details. What factors matter. What the trade-offs are. What you'd typically consider. This format does something generic content cannot: it demonstrates judgment, not just knowledge. Anyone can Google "how to handle concentrated stock positions." What clients actually want to know is whether their advisor has the judgment to handle their specific, complicated situation. Showing your thinking in case-study format is the closest thing to a pre-hire evaluation. For [financial services content](/services/financial-services-seo-consultant), the SEO angle matters too. Search intent for high-value financial queries tends to be informational ("how does X work") before it's transactional ("find an advisor for X"). Publishing thorough, specific content on complex topics captures the research phase, then converts when the reader is ready. **Distribution: Where Wealth Management Content Actually Gets Found** Organic search is the primary distribution channel. Wealthy clients search for specific financial situations, not for advisor recommendations. A client who finds your content while searching "RMD strategy with multiple IRAs" and spends fifteen minutes reading a thorough explanation has already formed a meaningful impression of your competence. LinkedIn works for business owner and executive audiences. Not for posting generic market commentary that everyone ignores, but for publishing original insights on specific planning situations that signal specialized expertise. A piece on "three things I'd tell every founder before a liquidity event" gets shared in founder communities in a way that "market update Q1 2026" never will. Email newsletters for existing clients serve a different purpose than acquisition content. They're about retention and deepening relationships. The best advisor newsletters I've seen aren't market recaps, they're brief, pointed observations about something happening in the market or tax law that's relevant to the client's actual situation. Relevant means "would my client with $3M in a mix of pre-tax and Roth accounts care about this?" If not, cut it. Webinars work for pre-retirees and business owners on specific planning topics. "Optimizing Social Security timing: what the calculator doesn't show you" draws a qualified audience. Record them, repurpose the content, and use the registration list as a warm prospect group. **What a Realistic Content Calendar Looks Like** For an independent RIA or wealth management firm with one or two marketers, a realistic publishing cadence is two pieces per month of substantial content (1,500+ words) plus a monthly newsletter. Trying to produce more than that without sacrificing quality is a trap. Each piece should target a specific audience segment and a specific question. Map your topics to your ideal client profiles. If your best clients are pre-retirees with complex equity compensation situations, every piece of content should be relevant to someone in that situation. Build a [content gap analysis](/tools/content-gap-calculator) into your annual planning. What questions are your best clients asking your team that aren't answered well anywhere online? Those questions are content opportunities nobody else has spotted yet. **Key Takeaways** - Compliance restricts specific performance claims and testimonials, not educational content, opinion, or situational case studies with disclosures. - Content for pre-retirees, business owners, and high-earning younger professionals requires different formats and topics. - Specific, situational topics attract qualified prospects. Generic financial advice content is saturated and doesn't convert. - The case-study format ("here's the situation, here's how we think about it") demonstrates judgment, which is what wealthy clients are actually evaluating. - Organic search is the primary distribution channel. Wealthy clients research specific financial situations, not advisors, first. - Two substantial pieces per month beats eight mediocre pieces at a cadence the team cannot sustain. --- # HIPAA-Compliant Paid Social for Health Systems URL: https://brianroseman.com/insights/hipaa-compliant-paid-social-health-systems Published: 2026-04-19 Meta removed health interest targeting in 2022. The FTC scrutinized pixels in 2022 too. Here is what actually works for health system paid social without legal exposure. **Summary:** Running paid social for health systems, hospitals, or medical practices isn't just a marketing challenge. It's a compliance minefield. Since Meta restricted health-related targeting in January 2022 and the FTC began scrutinizing pixel tracking in healthcare, the old playbook is essentially gone. This post covers what actually works for health system paid social in 2026 without putting your organization at legal risk. **The January 2022 Moment That Changed Everything** Before January 2022, you could target Facebook and Instagram users based on health conditions, pharmaceutical interests, and even specific diagnoses. Advertisers could layer "cancer awareness" interest targeting on top of demographic filters to reach likely patients. Then Meta removed those targeting categories almost overnight, citing "potential for misuse." Health system paid social teams panicked. The detailed interest categories they had built campaigns around disappeared. Click-through rates dropped as creative teams scrambled to rebuild audience logic from scratch. But here's what actually happened in the aftermath: the teams that adapted fastest figured out that the old targeting was never that good anyway. Interest-based health targeting was noisy. People interested in "diabetes management" include caregivers, researchers, journalists, and students, not just patients. The forced rebuild led many organizations to better audience strategy. **Why Pixels Are Now Your Biggest Legal Exposure** The tracking pixel problem is separate from targeting, and it's arguably more serious. In 2022, the FTC released a policy statement on [health breach notification](https://www.ftc.gov/legal-library/browse/rules/health-breach-notification-rule), and the HHS Office for Civil Rights followed with [guidance on tracking technologies](https://www.hhs.gov/hipaa/for-professionals/privacy/guidance/hipaa-online-tracking/index.html) that explicitly called out third-party pixels, including the Meta Pixel and Google Analytics, as potential HIPAA violations when deployed on pages where patients interact with health information. The specific risk: if someone visits a page about a particular condition or treatment on your health system's website, and your Meta Pixel fires, that visit data goes to Meta linked to an identifier. If that person is also a patient, you may have transmitted Protected Health Information to a third party without a Business Associate Agreement. Several health systems received OCR investigation letters in 2023 and 2024 after news reports exposed their pixel usage. Some settled for millions. The risk is real. What you can do: - Deploy server-side tracking instead of client-side pixels. Server-side conversion APIs (Meta's CAPI, Google's Enhanced Conversions) can be configured to strip PHI before transmission. - Work with your legal team to map which pages on your site qualify as "unauthenticated" (general information) versus pages that require login or collect health details. - For unauthenticated pages, a properly scoped pixel may be defensible. For anything behind a patient portal login, remove pixels entirely. - Review your BAA landscape. Some analytics vendors will sign BAAs; most consumer advertising platforms will not. **What You Can Actually Target In 2026** Without detailed health interest targeting, your audience options are: - Geographic targeting. Health systems are fundamentally local businesses. Tight geo-targeting around your service area, combined with demographic filters, is the cleanest approach. - Behavioral lookalikes built from first-party lists. Upload your existing patient email list (with consent under your Notice of Privacy Practices) to build a lookalike. This is legal, effective, and gets better as your list grows. - Life event targeting. Meta still allows targeting around life events like "new parent," "recently moved," and "recently engaged." These aren't health conditions, but they correlate strongly with health service needs. - Job-based targeting for B2B healthcare. If you're marketing to physicians, case managers, or benefits administrators, LinkedIn's job title targeting remains the most precise option in the market. - Keyword-based targeting on search. Google Search remains the most intent-rich channel for healthcare because the user self-identifies the need. A person searching "orthopedic surgeon Kansas City" is far more qualified than someone served a social ad based on inferred interest. The mix that works best in 2026 is: tight geo plus life events for awareness campaigns on social, remarketing to your own website visitors for consideration (using server-side CAPI to stay compliant), and branded plus condition-based search for bottom-funnel intent. **Creative and Copy Rules That Protect You** HIPAA and FTC rules around healthcare advertising extend beyond tracking. Creative and copy carry their own risks. The most common issue: implied claims. "We can cure your back pain" creates an implied promise that triggers FTC oversight. "Our spine specialists have helped thousands of patients find relief" is factual and defensible. The difference matters legally. Other rules to follow in copy: - Never use patient photos or testimonials without written HIPAA authorization specifically for marketing use. Authorization for treatment doesn't cover marketing. - Avoid before/after outcome claims unless you can back them with clinical evidence and clear disclaimers. - Do not use phrases that imply you know something specific about the person's health condition. Retargeting ads that say "Still looking for help with your knee?" can feel personalized but may cross a line if they imply the viewer's health status. The creative approach that consistently works: educational framing. "What to expect from knee replacement surgery" outperforms "Get knee surgery here." It signals expertise, builds trust without making claims, and attracts people who are already in the research phase. **Measurement Without PHI** If you strip pixels from key pages, how do you measure anything? This is the frustration most health system marketing teams hit. Here's the approach that works: Set up conversion events that don't collect PHI. Phone call volume (using a tracking number that doesn't require patient login), general form submissions for appointment requests (as long as the form itself doesn't ask for diagnosis or health details), and time-on-page for educational content all give you signal without touching PHI. Attribution in healthcare is harder than in e-commerce because the patient journey from first exposure to scheduling to visit can take weeks or months. Build a model that connects media spend to appointment volume at the service line level, not at the individual patient level. Media mix modeling, not multi-touch attribution, is the right framework here. If your organization uses Epic or another EHR with a marketing module, ask your vendor about HIPAA-compliant analytics that keeps patient data inside your environment. Some enterprise healthcare platforms offer this and it's the cleanest solution available. **The Platform Mix That Makes Sense Now** Meta: Use it for geo-targeted awareness and life event targeting. Restrict pixels to unauthenticated public pages only via server-side CAPI. Good for building brand familiarity over time, not great for direct response. Google Search: Still the most important channel for healthcare paid media. High intent, measurable, and the targeting is behavioral (what someone searched) not identity-based (who they are). Run condition-specific campaigns, branded campaigns, and competitor campaigns carefully. LinkedIn: Essential if you're marketing to physicians, employers for employee health programs, or insurance plan decision-makers. Expensive per click but the targeting precision is unmatched. YouTube: Underused for health systems. Pre-roll with a "skip after 5 seconds" format means you only pay when people watch. Educational video about complex procedures works well here. Track view-through rates, not just click-through. Connected TV: Growing for health systems with larger budgets. Allows geo-targeting and demographic filtering without the PHI risk of digital display, because CTV doesn't drop pixels on a healthcare website. I worked with a regional health system that shifted 30% of their Meta budget to CTV and saw comparable awareness metrics with significantly reduced compliance risk. The measurement was harder but the legal exposure was much lower. **What the Best Health System Paid Social Teams Do Differently** They involve legal earlier. The worst outcomes I've seen happen when marketing teams build out campaigns for months before anyone asks compliance whether the pixel setup is appropriate. They treat compliance as a creative constraint, not a blocker. Every channel restriction forces a better answer. If you can't use condition-based interest targeting, you get better at geographic and behavioral segmentation. They measure the whole funnel loosely rather than individual touchpoints precisely. Healthcare attribution is inherently imprecise. Teams that accept that and build business-level metrics, like "service line appointment volume in markets where we ran paid social," do better than teams chasing individual click-to-conversion paths they cannot accurately build in a HIPAA environment. **Key Takeaways** - Remove the Meta Pixel and Google Analytics from any authenticated pages or pages where patients enter health information. Use server-side APIs instead. - Geo-targeting plus life events is the primary audience strategy now that health interest targeting is gone on Meta. - Build lookalike audiences from your first-party patient email list (with proper consent under your NPP). - Educational creative ("what to expect from") consistently outperforms direct response for health system advertising. - Google Search remains the highest-intent channel for healthcare paid media. - Attribution in healthcare requires a service-line level model, not individual patient tracking. --- # Real Estate Marketing Automation: Nurture Sequences That Close Deals URL: https://brianroseman.com/insights/real-estate-marketing-automation-nurture Published: 2026-04-11 Real estate leads go cold fast. These automation workflows keep prospects engaged from first click to closing. "**Summary:** The average home buyer takes 10 weeks from first search to making an offer, per NAR data. The average real estate internet lead converts at 2-3%. What happens in between, the nurture, the follow-up, the behavioral triggers, determines whether your advertising spend generates deals or just expensive contacts who ended up working with someone else. This post covers the automation systems and sequences that actually move real estate leads toward closing. **The Real Estate Lead Conversion Problem** Most real estate marketing conversations focus on generating leads. The harder problem is converting them. A 2-3% conversion rate on internet leads means that 97-98% of the leads you paid to generate will not close with you. That's the industry average, and for most agents and brokerages, ""improving"" means moving from 2% to 3% or 4%, which sounds modest but doubles or triples business. The gap between lead generation and conversion is almost entirely a nurture problem. Internet real estate leads are rarely ready to transact when they first appear. They're researching, comparing agents, figuring out whether they can afford to buy, deciding if they actually want to sell. The agent who stays visible, useful, and present throughout that 10-week (or 10-month) journey earns the eventual business. [NAR's Home Buyers and Sellers Generational Trends research](https://www.nar.realtor/research-and-statistics/research-reports/home-buyers-and-sellers-generational-trends) documents this consistently: the buying process takes significantly longer than most agents budget for, and repeat business from past clients and referrals is a larger share of production for top agents than internet lead conversion. The agents and teams who run systematic automation don't have higher natural talent. They have better systems for staying present when manual follow-up fades. **Buyer vs. Seller Lead Nurture: Two Very Different Journeys** The first mistake in real estate automation setup is treating buyer and seller leads the same way. Their concerns, timelines, and decision triggers are completely different. Buyer leads typically have a longer, messier journey. A buyer who starts searching in January might get pre-approved in March, spend two months making offers that lose out in competitive situations, and finally close in June. The nurture content during that journey needs to address their specific stage: helping pre-qualified buyers understand the market, helping them stay motivated after losing offers, celebrating progress milestones. Seller leads are often more time-sensitive when they're hot and much longer-term when they're cold. Someone who requested a home value estimate in September might not list until the following spring when kids are out of school. A seller lead nurture sequence needs a short-term track for leads who are actively preparing to list and a long-term track for leads who are 6-18 months out. The practical implication: your CRM's intake process needs to capture buyer vs. seller intent, and your automation sequences need to branch based on that answer. An email about ""what to fix before listing"" going to a buyer lead wastes a touch. An email about ""10 things first-time buyers miss"" going to a seller is similarly wasted. **Behavioral Triggers That Indicate Buyer Readiness** The most powerful automation in modern real estate platforms fires based on behavior, not just passage of time. The behavioral signals that indicate a buyer lead is moving toward active purchase mode: Saving a specific listing or saving multiple listings in a short time window. A lead who saves 12 homes in a week is in a different state of intent than a lead who saved one home three weeks ago. Returning to the same listing multiple times. Someone who views a property page six times across four days is not casually browsing. They're considering making an offer. This behavioral signal should trigger an immediate alert to the agent and a specific automation: ""I noticed you've been looking at [address] a few times. Are you interested in scheduling a showing?"" Requesting a price drop notification for a specific listing. This is explicit intent with a price sensitivity signal attached. Opening every email in your sequence for the last three weeks. High email engagement from a lead that had previously been cold is often a signal of renewed search activity. Most modern real estate platforms, including [Follow Up Boss](https://www.followupboss.com/) and [kvCORE](https://www.kvcore.com/), have built-in behavioral triggers. The key is configuring them to send alerts to agents fast enough to act on the signals, and building automated responses that fire immediately when human follow-up isn't possible. **CRM Selection: What Actually Matters for Automation** Real estate CRM selection is a market cluttered with tools making nearly identical claims. The distinction that matters for automation specifically: Native lead source integrations. The best automation in the world breaks down if leads from Zillow, Realtor.com, your website, and your open house sign-in sheet all require manual import. Your CRM needs native, automatic integrations with your major lead sources. Behavioral tracking across your IDX (Internet Data Exchange) site. Platforms like Follow Up Boss and kvCORE track which listings leads view, save, and return to across your property search site. This is the behavioral data that feeds the high-value triggers described above. A standalone CRM connected to a separate IDX site rarely achieves this integration reliably. Smart plan (automation) flexibility. Can you create if/then logic? Can a lead move from one automated sequence to another based on their response (or non-response) to an email? Can you pause automation and insert a manual task when a lead shows high engagement? Reporting on automation performance. You need to be able to answer: what percentage of leads in each automated sequence eventually convert? Which email in the sequence generates the most replies? Which behavioral trigger produces the most showings? **SMS Automation and TCPA Compliance** SMS has higher open and response rates than email for real estate leads, but it comes with specific compliance requirements under the Telephone Consumer Protection Act (TCPA). Violating TCPA in real estate marketing is a legitimate financial risk. The TCPA basics for real estate automation: Explicit written consent is required before sending marketing SMS messages. A lead filling out a form to receive property listings is providing contact information, but that doesn't automatically constitute TCPA consent for automated marketing texts. Your forms should include explicit opt-in language for text messages. Opt-out requests must be honored immediately and automatically. If someone texts STOP, your system must remove them from SMS automation immediately. Any text that goes out after an opt-out creates liability. Time restrictions apply. TCPA generally restricts marketing calls and texts to 8am-9pm in the recipient's local time zone. The platforms that handle this best (Follow Up Boss, kvCORE, Sierra Interactive) build TCPA compliance into their SMS features: explicit consent capture, automatic opt-out processing, time zone awareness. Using a general marketing automation platform for real estate SMS without these features creates compliance exposure. The content of compliant real estate SMS: responding to a showing request, confirming an appointment, alerting to a new listing that matches explicit saved search criteria. These are conversational and service-oriented. Unsolicited marketing blasts (""New listings in your area!"") to people who didn't specifically opt in for text messages is the exposure zone. **Past Client Reactivation: The Highest ROI Automation** The most economically efficient automation program in real estate isn't lead nurture for new internet leads. It's past client reactivation and referral development. A past client who closed two years ago has already paid for their acquisition. They know you, presumably trust you, and if they're a homeowner, they'll need to transact again. The average homeowner moves every 7-10 years. A client database of 200 past clients represents, on a statistical basis, 20-30 potential future transactions in the next decade. Most agents maintain inconsistent contact with past clients. They send Christmas cards, maybe. The agents who build systematic reactivation programs stay in contact meaningfully without being annoying. The approach that works: Annual home value update in Q1. Send every past client a personalized email with an estimated current value of their home (using Zillow's or your CRM's AVM, with appropriate context that it's an estimate). This is a genuinely useful piece of information, it's not a sales pitch, and it creates a natural conversation starter if they're thinking about moving. Market updates quarterly, with local specificity. Not ""the national real estate market is..."" but ""here's what happened in [neighborhood] in Q1: X homes sold, median price was Y, average days on market was Z."" This is information they can't easily get anywhere else and positions you as the local expert. Milestone recognition. Home purchase anniversaries (one year, five years, ten years) are moments to acknowledge. A short personal note or text on a home anniversary is memorable because almost no one does it. Referral asks after positive interactions. After delivering a market update that gets a positive reply, after a client reaches out about a neighbor selling, these are the moments for a natural referral conversation, not in a scheduled blast. The [LTV calculator](/tools/ltv-calculator) makes the economics of past client reactivation concrete: the cost of staying in contact (time + platform cost + occasional mailer) versus the commission value of one past client transaction or referral. The numbers are usually lopsided in favor of past client investment. **Measuring Nurture Effectiveness** The metrics that tell you whether your automation is working: Lead-to-appointment conversion rate by stage. Not just overall, but by where the lead was in the nurture sequence when they scheduled. Leads who set appointments after receiving the fourth email in a 30-day sequence are different from leads who set appointments after a behavioral trigger. Sequence reply rates. Email replies, even negative ones, are high-value signals. What percentage of leads who enter your buyer nurture sequence reply to at least one email? This tells you whether the content is resonating or landing in bulk email patterns. Long-term conversion rate by lead source and entry date. Some lead sources produce leads that convert immediately. Others produce leads that incubate for 6-18 months before converting. Knowing which sources produce which types of leads changes your nurture expectations and your budget allocation. The [marketing assessment](/tools/marketing-assessment) helps identify where in the lead-to-close process the biggest gaps are, which is usually the starting point before investing in new automation tools. The [real estate PPC geo-targeting post](/insights/real-estate-ppc-geo-targeting-local) covers the lead acquisition side, making sure the leads entering your automation are the right leads from the right geographies at sustainable cost per lead. **Key Takeaways** - NAR data shows buyers average 10 weeks from first search to offer, but internet leads convert at 2-3% industry-wide; the gap is almost entirely a nurture and follow-up problem, not an advertising problem. - Buyer and seller leads need separate automation tracks; their timelines, concerns, and content needs are different enough that shared sequences waste touches and miss conversion opportunities. - Behavioral triggers (saved listings, repeat property views, price drop alerts) indicate genuine purchase intent and should fire both agent alerts and immediate automated responses. - TCPA compliance for SMS automation requires explicit written consent at opt-in, immediate opt-out processing, and time zone-aware scheduling; specialized real estate platforms handle this correctly while general marketing automation tools often don't. - Past client reactivation programs have the highest ROI of any real estate automation because acquisition cost is already paid; annual home value updates, quarterly market reports, and home anniversaries are the contact cadence that keeps relationships active without being intrusive. - Measure nurture effectiveness at the sequence reply rate, lead-to-appointment conversion by stage, and long-term conversion rate by lead source, not just at total leads in your CRM." --- # Retail Email Marketing: Personalization That Drives Opens URL: https://brianroseman.com/insights/retail-email-personalization-drives-opens Published: 2026-04-10 Most abandoned cart emails get ignored. These 5 sequence frameworks drove $2.3M in recovered revenue in the past **Summary:** Abandoned cart emails get a 45% open rate but most brands waste it with generic "you forgot something" messaging. These 5 sequence frameworks recovered $2.3M in revenue for e-commerce clients I supported in the past. The difference is timing, personalization, and knowing when to stop. **Why Most Abandoned Cart Emails Fail** The average cart abandonment rate sits around 70%. That is a massive revenue leak. Most brands respond by sending one or two generic emails: "You left something in your cart!" with a product image and checkout button. It works. Sort of. Typical recovery rates hover around 5-10% of abandoned carts. But the brands recovering 15-25% of carts do something different. They treat abandonment as a conversation, not a single touchpoint. Before I switched to analytics and seo, I once sent a lot of emails out in past agency role and kept good notes. I spent time analyzing cart recovery sequences across 23 e-commerce brands. The highest performers shared specific patterns - timing windows, message sequencing, personalization triggers. Here is what actually moved the needle. **Sequence 1: The Speed Recovery (Under $100 Carts)** For lower-value carts, speed beats sophistication. These shoppers are often comparing prices across tabs. If you wait 24 hours to email, they have already bought elsewhere. **Email 1 (45 minutes post-abandonment):** Subject line focused on the specific product, not generic cart messaging. "Still thinking about the [Product Name]?" outperforms "You left something behind" by 23% in our testing. **Email 2 (18 hours later):** Social proof email. Customer reviews of the specific product they abandoned. No discount yet. Just validation that other people bought and liked it. **Email 3 (48 hours, final):** If no conversion, a small incentive (free shipping or 10% off) with urgency. "This is our last reminder" framing. The 45-minute first email is critical. Conversion rates on that first touch drop by 50% if you wait more than 2 hours. Speed matters more than perfect copywriting. **Sequence 2: The High-Value Cart Sequence (Over $500)** Big purchases need different treatment. A $800 cart abandoner is not forgetting - they are deliberating. Your job is addressing objections, not reminding them. **Email 1 (2 hours post-abandonment):** Acknowledge the decision is significant. Offer a phone call or chat with a product specialist. No discount. This signals confidence in value. **Email 2 (24 hours):** Address the top objection for that product category. For furniture, it is "will it fit?" For electronics, it is "is this the right model?" For apparel, it is sizing and returns. Make returns and sizing information incredibly clear. **Email 3 (72 hours):** Case study or customer story with specific results. Someone who bought the same product and their experience. **Email 4 (5-7 days):** Final email. If they have not converted, offer a meaningful incentive or financing option. This is your last touch - make it count. High-value sequences convert at lower rates but the revenue per recovery is dramatically higher. One client recovered $47K from a single month of carts over $1,000. **Sequence 3: The Browse Abandonment Bridge** Not all abandonment is cart abandonment. Browse abandonment - when someone views products but never adds to cart - requires a different approach. This is not a traditional recovery sequence. It is a bridge to get them to the cart. **Email 1 (4 hours post-browse):** "Noticed you were looking at [Category]" with curated products similar to what they viewed. No pressure, just relevance. **Email 2 (2 days later):** Educational content related to the product category. Buying guides, comparison posts, or usage tips. Build confidence in the category, not just your product. The conversion rate on browse abandonment is lower (2-4% vs 10-15% for cart abandonment) but the volume is much higher. Most sites have 5-10x more browse abandoners than cart abandoners. **Sequence 4: The Repeat Customer Recovery** Someone who has bought before and abandons a cart is completely different from a first-time visitor. They already trust you. The objection is not "is this brand legit?" - it is usually timing or budget. **Email 1 (30 minutes):** Acknowledge their history. "Welcome back, [Name]. Your cart is waiting." Include their loyalty points or tier status if applicable. **Email 2 (24 hours):** If you have a loyalty or rewards program, remind them how close they are to the next tier or reward. "This purchase gets you to Gold status" or "You are 50 points from your next reward." **Email 3 (48 hours, if needed):** Exclusive returning customer offer. Not a generic discount - something that feels earned. Repeat customer abandonment recovery should run at 20-30% conversion. If you are under 15%, your loyalty program integration is probably broken. **Sequence 5: The Exit-Intent Prevention** The best cart recovery is preventing abandonment in the first place. Exit-intent emails trigger when someone shows signs of leaving - cursor moving to close tab, long idle time, switching tabs. **Real-time triggered email (immediate):** "Your cart will expire soon" or "Complete your order to lock in pricing." This creates gentle urgency without being pushy. This is not technically a sequence - it is a single triggered message. But brands using exit-intent triggers see 5-8% reduction in overall abandonment rates. That compounds significantly at scale. **The Technical Setup That Makes This Work** These sequences require proper infrastructure: **Cart data syncing:** Your email platform needs real-time cart data. Klaviyo, Omnisend, and Drip all handle this natively with most e-commerce platforms. If you are using generic email tools, you are flying blind. **Segmentation by cart value:** Not all carts deserve the same treatment. Set up dynamic segments for cart value tiers ($0-100, $100-500, $500+). **Suppression rules:** Stop sending cart emails if someone converts, unsubscribes, or has received 4+ emails without engagement. Aggressive cart sequences damage deliverability if not managed. **Product feed integration:** Dynamic product blocks that pull the actual abandoned items beat static emails by 25-40% in conversion rates. **Measuring Recovery Revenue Correctly** Cart recovery attribution gets messy. If someone abandons a cart, gets your email, but returns via Google search to purchase - did the email work? Track these metrics separately: - Direct email conversion (clicked email, purchased in same session) - Assisted conversion (received email, purchased within 7 days via any channel) - Sequence fatigue (unsubscribes or spam complaints per email in sequence) Most brands only track direct conversion and undercount their actual recovery revenue by 30-40%. **Key Takeaways** - Send the first abandonment email within 45 minutes for low-value carts - conversion drops 50% after 2 hours - High-value carts need objection handling, not just reminders - offer specialist consultations - Browse abandonment sequences capture 5-10x more volume than cart-only programs - Repeat customers should convert at 20-30% on abandonment sequences - segment them separately - Suppression rules prevent deliverability damage - stop after 4 emails without engagement - Track assisted conversions alongside direct - you are probably undercounting recovery revenue by 30-40% Source URL Klaviyo abandoned cart benchmarks 2025 https://www.klaviyo.com/marketing-resources/abandoned-cart-benchmarks Baymard Institute cart abandonment research https://baymard.com/lists/cart-abandonment-rate --- # Travel Marketing Attribution: Tracking the 47-Day Booking Journey URL: https://brianroseman.com/insights/travel-marketing-attribution-booking-journey Published: 2026-04-07 Travel purchases take weeks of research. Here is how to attribute credit across a fragmented buyer journey. "**Summary:** Travel marketers have one of the longest and most touchpoint-heavy customer journeys of any industry to attribute. A 47-day average consideration cycle with up to 38 digital touchpoints means that standard 30-day attribution windows miss most of the journey, and last-click attribution tells a story that's almost completely wrong. This post covers how to set up attribution that actually reflects how travel decisions get made. **The 47-Day Journey: What the Research Actually Shows** Think about how you planned your last significant trip. The first search probably happened weeks before any booking. You searched a destination, looked at photos, read travel blogs, watched a couple of YouTube videos, started comparing hotel options, then the trip planning got interrupted by something else in life. You came back to it two weeks later. [Think With Google's research on travel booking behavior](https://www.thinkwithgoogle.com/marketing-strategies/app-and-mobile/travel-booking-journey/) has documented this pattern in detail. For leisure travel, the average booking journey spans 47 days from first search to booking. For international travel, it stretches longer. Within that window, travelers make an average of 38 digital touchpoints across search, social, video, and travel review sites. The attribution problem is immediate: most Google Ads accounts are configured with a 30-day lookback window. Most marketing platforms default to a 28-day or 30-day attribution window. If the journey averages 47 days, you're missing the first third of the journey, and every attribution report is crediting only the tail end of a much longer sequence. This matters for budget allocation. If your attribution data shows that ""Google Ads - Search"" is your top channel because it's generating the final click before booking, but you're not capturing that half those conversions started with a YouTube video 45 days earlier, you'll underinvest in video and overinvest in search terms that were capturing people who were already certain buyers. **Why Standard Attribution Windows Fail Travel** The three scenarios where standard attribution systematically fails travel marketers: First-touch invisibility. The first touchpoint in a 47-day journey, which might be a YouTube pre-roll ad about a destination, an organic social post, or a travel blog post, never gets credit because the 30-day lookback window has closed by the time the booking happens. Cross-device fragmentation. A traveler might start research on a laptop at work, continue on a phone during commute, and book on a tablet at home. Without user-level identity that persists across devices (logged-in Google accounts, logged-in app sessions), each device looks like a different user, and the journey looks like three separate, short sessions ending in a conversion. Inspiration-to-booking gap. Destination inspiration content (travel blogs, social media, YouTube) typically creates desire weeks before the purchase intent develops. A hotel brand that generates initial inspiration has created genuine economic value that attribution systems assign to whoever happened to be visible at the end of the journey. **Setting Up Extended Lookback Windows in GA4** [GA4's attribution documentation](https://support.google.com/analytics/answer/10597962) covers the lookback window settings for the data-driven attribution model. The default for non-purchase events is 30 days. For purchase events (which include travel bookings), the default is 30 days in the ads-preferred model and up to 90 days in data-driven attribution. For travel brands, 90 days is the starting point, not the ceiling. Here's how to adjust: In GA4, navigate to Admin, Attribution Settings, and look at the lookback window configuration. Data-driven attribution for conversions should be set to 90 days minimum. For brands with significant international travel bookings, 90-day windows capture a larger portion of the full journey. For Google Ads, the attribution window settings are separate from GA4. In Google Ads, the conversion settings for each conversion action have their own lookback window. Navigate to Tools & Settings, Conversions, and for each conversion action, set the lookback window to 90 days. The tradeoff with longer windows: more credit gets distributed to earlier touchpoints, which may reduce the apparent ROAS of your bottom-funnel campaigns (since some of their conversions now share credit with earlier touches). This is the correct mathematical result, not an error. It reflects the actual complexity of the purchase journey. **Data-Driven Attribution vs. Last-Click for Travel** Data-driven attribution (DDA) uses machine learning to distribute credit across touchpoints based on which touchpoints actually influence conversion probability, rather than applying a fixed rule (first click, last click, linear, etc.). For travel, DDA consistently shows a different picture than last-click: Search brand terms capture less credit. In last-click attribution, someone searching your brand name right before booking gets full credit. In DDA, that search is often the final step of a decision that was already made, so it gets partial credit, with more credit flowing to earlier touchpoints that influenced the decision. YouTube and display advertising capture more credit. Destination inspiration content that appears early in the journey gets credited proportionally to its actual influence on conversion probability, rather than zero (as in last-click attribution where only the last touch counts). Organic search and content capture more credit. The travel blog post that sparked the destination idea, if trackable, gets attributed proportionally rather than being invisible to last-click models. The practical implication: under DDA, some of your top-performing channels may shift. Paid brand search ROAS will often appear to decrease (because it's now sharing credit) while upper-funnel channels will appear more valuable. This reallocation of apparent value, if acted on, typically produces better actual outcomes because you're investing in the full journey instead of just the final step. **Measuring Brand vs. Non-Brand Search Separately** One of the most important segmentation decisions in travel attribution is separating branded and non-branded search performance. Branded search (your property or hotel name, your airline, your tour company name) captures people who already know you and have decided to find you. The ROAS looks fantastic because these people were coming anyway. The cost is low, the conversion rate is high, and the margin looks exceptional. Non-branded search (destination keywords, category keywords, competitor name keywords) is where you're actually competing for new customer acquisition. The ROAS is lower, the cost per click is higher, and the conversion rate is lower, but this is where growth comes from. Mixing branded and non-branded in your reporting creates a blended ROAS number that flatters your paid search program by including conversions from people who were guaranteed buyers. It also masks underperformance in non-branded terms, where your advertising is doing actual acquisition work. Run branded and non-branded as separate campaigns with separate budgets, separate performance targets, and separate reporting. The non-branded ROAS target should reflect the lifetime value of a new customer acquisition, not the cost efficiency of capturing existing customers. **The Role of YouTube in Travel Inspiration** YouTube is disproportionately important in the travel purchase journey relative to most other industries. Travel is visual. Destination content (hotel room walkthroughs, local experience videos, local food guides) drives genuine desire in a way that text and static images often can't. [Skift](https://skift.com/) has consistently tracked the role of video in travel inspiration, and the finding is consistent: video content significantly influences destination selection among travelers in the research phase. For measurement, YouTube's View-Through Conversions are relevant but need to be interpreted carefully. A view-through conversion credits a YouTube ad with a conversion if the user saw the ad and then later converted, without clicking. The default attribution window for view-through conversions in Google Ads is 24 hours, which is too short for a 47-day journey. Extending the view-through conversion window to 30 days captures more of YouTube's actual influence. But view-through conversions should always be reported separately from click-based conversions, because their causal relationship with the booking is weaker (you can't know whether the person would have booked without the ad view). **Post-Booking Engagement: Where Most Travel Marketers Stop Too Early** The booking confirmation is not the end of the journey. For most travel brands, the post-booking period is when significant upsell and cross-sell revenue is available, when loyalty enrollment happens, and when the foundation for the next booking is laid. Post-booking email automation for travel should include: Trip countdown emails that build anticipation and deliver relevant content (packing lists, local guides, weather advisories) in the weeks before the trip. These emails have open rates well above standard promotional emails because they're clearly useful. Upsell offers for add-ons: room upgrades, dining reservations, activity bookings, car rentals. The post-booking period, when the traveler is excited about the trip, is the highest-conversion moment for these offers. During-trip service communication for hotels and resorts: check-in instructions, service requests, local recommendations. Brands that are helpful during the trip earn loyalty that drives repeat bookings. Post-trip review requests and loyalty enrollment, timed within 24-48 hours of the checkout date while the experience is fresh. This is also the right moment for a targeted offer for the next booking: ""Thank you for staying with us. Here's 15% off your next reservation before June 30."" For the attribution infrastructure to capture this full journey, connecting to the [ROAS calculator](/tools/roas-calculator) helps model the total revenue contribution including post-booking upsell. The [SEO ROI calculator](/tools/seo-roi-calculator) is useful for quantifying the value of organic content investment at the top of the travel funnel. The travel SEO post on [competing with OTAs for destination keywords](/insights/travel-seo-competing-ota-destinations) covers the content strategy that fills the top of this journey with organic traffic. The [marketing assessment](/tools/marketing-assessment) evaluates whether your current attribution setup is capturing enough of the journey to make good optimization decisions. **Key Takeaways** - The average leisure travel booking journey spans 47 days with up to 38 digital touchpoints, meaning standard 30-day attribution windows miss the first third of every booking's history. - Extend lookback windows to 90 days minimum in both GA4 attribution settings and Google Ads conversion settings; for international travel, consider longer. - Data-driven attribution in GA4 distributes credit more accurately across the full journey than last-click; it typically shows lower ROAS for branded search and higher value for upper-funnel inspiration content, which reflects reality. - Branded and non-branded search must be tracked separately; blended branded/non-branded ROAS misleads budget allocation by mixing guaranteed conversions with genuine acquisition. - YouTube view-through conversion windows should be extended to 30 days for travel, but always reported separately from click-based conversions given the weaker causal relationship. - Post-booking engagement (upsell offers, during-trip service, post-trip review requests) is a high-revenue, low-incremental-cost phase of the customer journey that most travel marketers underinvest in relative to pre-booking acquisition." --- # Retail Omnichannel Attribution: Connecting Online and In-Store Sales URL: https://brianroseman.com/insights/retail-omnichannel-attribution-online-store Published: 2026-04-04 Customers browse online and buy in-store. These attribution models capture the full purchase journey. **Summary:** Omnichannel retail attribution is one of the most technically complex measurement problems in marketing, because the customer journey doesn't respect the boundaries between online and in-store. Someone researches online, browses in-store, and buys online. Or researches on mobile, walks into a store, and buys there. Last-click attribution misses most of this journey. This post covers the measurement approaches that work and the practical implementation steps for retailers with both digital and physical channels. **Why Last-Click Attribution Destroys Omnichannel Measurement** Last-click attribution assigns 100% credit for a transaction to the last marketing touchpoint the customer interacted with before buying. In a single-channel world, this is imperfect but usable. In an omnichannel world, it's actively misleading. A customer who sees a TV ad, visits the store to touch the product, searches the brand on Google a week later, clicks a remarketing ad, and then walks into the store to buy it: last-click attribution credits the remarketing ad. The store visit that was the real conversion moment goes unmeasured. The TV ad that created initial awareness gets no credit. When retailers optimize marketing spend based on last-click attribution, they systematically underinvest in the upper-funnel and in-store experience, and overinvest in bottom-funnel paid search and retargeting. Over time, the customer base skews toward people who were already going to buy, and acquisition of genuinely new customers weakens. The [National Retail Federation](https://nrf.com/) tracks omnichannel adoption and the measurement challenges associated with connecting digital and physical retail, and the core finding is consistent: most retailers know attribution is broken but haven't built the infrastructure to fix it. **Store Visit Conversions in Google Ads** Google Ads provides store visit conversions as a measurement feature for retailers with physical locations. It uses a combination of Google Maps data, Google account location history, and machine learning to estimate how many of your Google Ads clicks resulted in a store visit within a defined time window. The limitations are significant: store visit conversions are modeled estimates, not exact counts. They require a minimum volume threshold that many smaller retailers don't hit. They require location permissions from users, which are increasingly restricted. And they measure store visits, not store purchases. But they're useful as a directional signal. [Google Ads help documentation on store visit measurement](https://support.google.com/google-ads/answer/6100636) explains the methodology and requirements. If your Google Ads campaigns are generating 30% more store visits than your control (non-advertised) periods, that's signal even without a precise transaction-level match. **The Loyalty Program as the Attribution Bridge** The most reliable mechanism for connecting online and in-store purchases is the loyalty program, because it creates a persistent identity that follows the customer across channels. When a customer identifies themselves with their loyalty account at the point of sale (in-store or online), you can connect that transaction to their full history: which emails they opened, which ads they were served, which pages they browsed, when they were last in the store. This is the first-party data advantage that has become substantially more valuable as third-party cookies disappear. The loyalty program only works as an attribution bridge if: In-store associates actively prompt loyalty account identification at every transaction. "Are you a rewards member?" asked at every checkout isn't just about loyalty enrollment, it's about closing the attribution loop. The loyalty system is integrated with your CRM and email marketing platform, so that online behavioral data (browse history, email opens) is connected to the same customer record as in-store purchase data. Members are given genuine reasons to identify themselves every visit, not just one-time enrollment. Earning points, tracking rewards progress, and accessing member pricing all create behavioral incentive to show the loyalty account rather than transacting anonymously. Retailers with loyalty programs that achieve 70%+ identified transaction rates have meaningfully better omnichannel measurement capability than those at 30-40% identified rates, even with the same technology stack. **ROAS Calculation When You Can't Track All Revenue** The ROAS calculation problem: if your online advertising drives both online purchases (tracked) and in-store purchases (partially tracked through store visit conversions and loyalty matching), your reported ROAS understates the true return by the amount of in-store revenue you can't connect. The adjustment approaches: If you have loyalty-based omnichannel matching, you can calculate an "in-store revenue multiplier" for specific campaign types. Run a controlled test: hold out 20% of loyalty members from a specific campaign and compare their in-store purchase rates to the 80% who received the campaign. The incremental in-store revenue in the campaign group, divided by the ad spend, gives you a partial picture of in-store ROAS contribution. If you don't have loyalty matching, you can use a geographic test: run a campaign in some markets and not others, then compare in-store sales lift in campaign vs. control markets using your POS data. This is a rough proxy but gives you a multiplier to apply to your digital ROAS numbers. The [ROAS calculator](/tools/roas-calculator) is useful for modeling different in-store revenue multiplier scenarios and understanding the impact on overall campaign economics. The [CRO calculator](/tools/cro-calculator) helps model the impact of conversion rate improvements in the digital channel, which is often the higher-ROI optimization because online channels are more directly measurable. **GA4 for Omnichannel Measurement** [Google Analytics 4's documentation on attribution](https://support.google.com/analytics/answer/10597962) covers the data-driven attribution model, which distributes credit across multiple touchpoints using machine learning. For retailers, data-driven attribution is significantly more accurate than last-click because it accounts for the full sequence of touches. Setting up GA4 for omnichannel measurement requires: Measurement ID on your website (standard). This captures online behavioral data, events, and transactions. Google Ads linking and auto-tagging, so that clicks from Google Ads are tracked through to GA4 with full session data. BigQuery export enabled, so that raw GA4 event data is available for custom analysis beyond what the GA4 interface shows. BigQuery analysis lets you join GA4 data with your CRM and POS data for true cross-channel matching. Server-side tracking implementation for high-value events. Browser-based JavaScript tracking is increasingly degraded by ad blockers and iOS privacy restrictions. Server-side tagging that fires GA4 and Google Ads conversion tracking from your server rather than the browser improves data accuracy for purchase events. Offline data import to connect in-store sales data to your GA4 account. GA4 supports offline event import through the Measurement Protocol, which lets you send server-side events that represent in-store transactions tied to online user identifiers. This is the technical mechanism for connecting the online and offline journey for loyalty-identified customers. **Retail Media Networks and Data Clean Rooms** Retailers with substantial first-party data are increasingly participating in retail media networks (think Amazon Advertising, Walmart Connect, or Kroger's 84.51 data business), both as publishers (monetizing their audience data) and as advertisers (using other retailers' data for targeting). For mid-sized retailers building omnichannel measurement, the relevant concept is the data clean room: a privacy-preserving environment where two companies can jointly analyze overlapping data without either party seeing the other's raw data. A retailer and a CPG brand might use a clean room to measure how the CPG brand's national advertising influenced purchases in the retailer's stores, without the CPG brand getting access to the retailer's customer records. Clean rooms are technically complex and currently most practical for retailers with at least $100M in annual revenue and substantial data infrastructure. But they represent the direction the industry is moving as third-party data access continues to shrink. **Email and SMS Roles in Driving In-Store Traffic** One of the more measurable cross-channel dynamics is the email-to-store-visit relationship. An email announcing a store event, a limited-in-store-only offer, or a "buy online, pick up in store" promotion can drive in-store traffic that is partially attributable because you know who received the email. The measurement approach: compare in-store purchase rates (for loyalty-identified customers) between email openers and non-openers in the week following a specific email. The lift in purchase rate among openers, minus the natural purchase rate of the non-opener segment, represents the in-store revenue lift attributable to the email. This isn't perfect (openers self-select; people who open emails may be inherently higher-intent buyers), but it's a meaningful directional signal. Over time, you can build a discount rate to apply to the apparent lift that accounts for selection bias. The retail personalization dimension of this connects directly to the [retail personalization post](/insights/retail-personalization-beyond-recommendations), which covers the behavioral segmentation that feeds both email personalization and in-store experience personalization. For the SEO side of driving traffic that eventually converts in-store, the [dynamic pricing and retail SEO post](/insights/seo-dynamic-pricing-retail-conversion-2026) addresses how to think about the digital content strategy that supports an omnichannel business. **Key Takeaways** - Last-click attribution systematically undervalues upper-funnel channels and in-store experiences, causing retailers to overinvest in bottom-funnel retargeting and underinvest in the channels that build new customer relationships. - Loyalty programs with high in-store identification rates (70%+) are the most reliable omnichannel attribution bridge, connecting online behavior to in-store purchase through persistent customer identity. - Google Ads store visit conversions are modeled estimates with significant limitations but provide directional signal about the in-store impact of paid search campaigns. - True ROAS for omnichannel retailers requires a multiplier for in-store revenue attribution; controlled tests using loyalty data or geographic holdouts are the most defensible way to estimate it. - GA4's data-driven attribution combined with BigQuery export and offline data import enables cross-channel analysis for retailers with the technical infrastructure to use it. - Retail media networks and data clean rooms represent the emerging infrastructure for cross-retailer omnichannel measurement as third-party data access continues to shrink. --- # Technology Account-Based Marketing: Orchestrating Enterprise Deals URL: https://brianroseman.com/insights/technology-abm-orchestrating-enterprise Published: 2026-04-01 Enterprise sales need coordinated outreach. These ABM playbooks align marketing and sales touchpoints. "**Summary:** Account-Based Marketing gets talked about more than it gets done correctly. Most B2B technology companies that claim to run ABM are actually running slightly more targeted demand generation, which is fine but misses the compounding benefits of genuine account-level orchestration. This post covers the distinction, the tiered approach that actually scales, and the metrics that tell you whether your ABM program is generating real pipeline. **ABM vs. Lead Gen: The Core Distinction** Traditional demand generation thinks in contacts. You run a campaign, a contact fills out a form, marketing scores them and passes them to sales. Individual humans move through your funnel. ABM thinks in accounts. The unit of measurement is the account, not the contact, because in enterprise technology sales, no individual makes the purchase decision. Gartner's research on B2B buying groups consistently finds that enterprise technology deals involve 6-10 stakeholders from different functions, often with competing priorities and different definitions of ""success."" A Chief Information Officer evaluating a data management platform sees a different problem than the VP of Engineering or the Director of Analytics. Lead generation tries to find the one decision-maker. ABM tries to reach, influence, and coordinate multiple stakeholders at the same account simultaneously, while tracking progress at the account level. The distinction matters operationally because ABM requires different content (account-specific, not persona-generic), different measurement (pipeline influenced vs. leads generated), and different sales-marketing alignment (shared account lists, coordinated outreach timing, single view of account engagement). **Building the Ideal Customer Profile with Real Data** The ICP (Ideal Customer Profile) exercise is where most ABM programs start and, unfortunately, where many produce bad results. A typical ICP workshop involves marketing and sales leaders sitting in a conference room agreeing that their best customers are ""mid-market technology companies in regulated industries with 500-2,000 employees."" That description could match thousands of companies. It's not an ICP. It's a size filter. A real ICP is built from data about your existing customers, specifically your highest-LTV customers. It asks: What industry verticals have the highest win rates? Not the most opportunities, the highest win rates. An industry you enter easily might still have lower win rates than an industry you enter less frequently but close more often. What technology stack signals predict fit? In B2B technology sales, what existing tools a company uses is often a strong fit signal. A company running Salesforce and Tableau is a different kind of buyer than a company running HubSpot and Google Sheets. Check your closed-won deals and map the tech stack. What organizational signals predict purchase? Organizational size, structure (centralized vs. decentralized IT), recent leadership changes in relevant functions, active hiring in related roles (an active search for a Director of Analytics is a signal for data management tools). These are signals you can monitor at scale. What deal characteristics correlate with long-term retention? Companies that expanded usage in the first 12 months are your best future target. What made them expand? Was it company size? Industry? Use case? The answers refine your ICP beyond initial purchase and toward lifetime value. [Demandbase's resources](https://www.demandbase.com/resources/) cover the intent data and account identification side of this well. **Tiered ABM: 1:1, 1:Few, 1:Many** The most practical ABM implementation for technology companies is a tiered approach: **Tier 1: 1:1 ABM for Top-Priority Accounts** This is true, full-effort ABM: custom content for specific accounts, coordinated multi-channel outreach, dedicated business development resources, custom proposals. You can run this for 10-30 accounts at a time, no more. The investment per account is high. These are your most strategically important prospects, high revenue potential, high win probability given fit, long-term expansion potential. For Tier 1 accounts, marketing creates custom materials: industry-specific case studies, personalized landing pages that reference the account by name, research on the account's specific challenges from their public communications. Sales uses this to show up to conversations prepared with account-specific context, not generic pitches. **Tier 2: 1:Few ABM for Account Clusters** Tier 2 runs for clusters of 50-200 accounts that share meaningful characteristics: same industry, same technology stack, same growth stage, same pain point. The content is personalized to the cluster, not to the individual account. An email campaign for ""healthcare technology companies transitioning from legacy EHR integrations"" is more targeted than a generic email but doesn't require custom creative for each account. Tier 2 is where most of your ABM program lives in terms of account volume. It's where intent data from [Bombora](https://bombora.com/resources/) is most actionable: accounts showing intent spikes in your topic category move up the priority list for Tier 2 outreach. **Tier 3: 1:Many ABM for the Long Tail** Tier 3 is essentially well-targeted demand generation: campaigns aimed at your ICP broadly, without account-specific personalization. It fills the funnel with accounts that may graduate to Tier 2 as they show more engagement. **Content Personalization at the Account Level** Generic content, the e-book, the webinar, the ""top 10 tips"" blog post, works for Tier 3. Tier 1 and Tier 2 need something more specific. The content that works for account-level personalization: Industry-specific ROI examples. Not case studies with the company name anonymized, but content that walks through the specific economics of solving the problem for a company in that industry. ""How companies in your industry typically see 14 months to ROI on this investment, and why"" is more compelling than a generic ROI calculator. Account research documents that show you understand their specific situation. Before a first meeting with a Tier 1 account, your business development team should have researched the account's recent product launches, technology announcements, leadership changes, and strategic priorities from public sources. A one-page ""here's what we know about your situation"" document is more impressive than any generic pitch deck. Personalized landing pages for ABM campaigns that reference the account name, industry, and specific challenge. ""Welcome, Acme Corporation"" with a landing page that speaks directly to the data management challenges specific to their vertical converts dramatically better than a generic product page. **Intent Data: How to Use It Without Over-Relying on It** Intent data from providers like Bombora, G2, or TechTarget tracks content consumption behavior across the web and identifies accounts that are actively researching topics related to your solution. An account whose employees have consumed significant content about ""enterprise data governance"" in the last 30 days is more likely to be in an active buying cycle than an account with no such signals. Used correctly, intent data is a prioritization layer, not a targeting system. An account showing strong intent that also matches your ICP moves to the top of your Tier 2 outreach list. An account showing strong intent that doesn't match your ICP is still not a good fit. The mistake is treating intent data as a substitute for ICP fit evaluation. Accounts showing intent signals but poor ICP fit have lower win rates even when they engage with your outreach. Intent data accelerates timing; fit predicts outcome. [6sense](https://www.6sense.com/resources/) and similar tools also provide account identification for anonymous web visitors, surfacing which target accounts are visiting your website without filling out a form. This is genuinely useful for coordinating sales follow-up. **LinkedIn in ABM Activation** [LinkedIn's marketing solutions](https://business.linkedin.com/marketing-solutions/blog) are unusually well-suited for ABM because the account and contact targeting capabilities are more precise than almost any other advertising channel. LinkedIn's Account Targeting lets you upload a list of company names and have your ads shown specifically to people at those accounts. For Tier 1 and Tier 2 ABM, this means your advertising spend goes specifically to the buying committee at your priority accounts, rather than broadly to the market. The best ABM use of LinkedIn advertising is sequential: reach accounts first with awareness content (industry research, category education), then with consideration content (vendor comparison, detailed product value), then with decision content (ROI calculators, customer proof) as accounts progress through the consideration cycle. The ad type matters. LinkedIn's Document Ads (which let people download a PDF directly in the feed) consistently outperform standard image ads for B2B content distribution in our experience. Conversation Ads (personalized LinkedIn message-style ads delivered to the inbox) are effective for Tier 1 accounts where the personalization investment is justified. **Measuring ABM with Pipeline Metrics** The measurement shift from demand gen to ABM: stop reporting on leads generated. Start reporting on accounts influenced, pipeline progression, deal velocity, and win rate. The key ABM metrics: Target account engagement score, which tracks marketing touchpoints from accounts on your target list. An account that has visited your website five times, attended a webinar, and had three of their employees view your LinkedIn content is more engaged than an account you've never reached. Pipeline influenced: the dollar value of open opportunities where your ABM program created at least one touchpoint. This is different from pipeline sourced (where ABM was the first touchpoint). Both matter but tell different stories. Account-level win rate compared to non-ABM-targeted accounts. If your win rate for Tier 1 accounts is 38% and your overall win rate is 22%, that's directional evidence the program is working. Deal velocity for ABM-touched accounts versus the baseline. If coordinated account-level outreach shortens the average sales cycle from nine months to six months, that's quantifiable value even before you attribute specific revenue to ABM. The [B2B social selling post](/insights/b2b-social-selling-pipeline-without-annoying) covers the personal LinkedIn presence side of this, which complements ABM's account-level advertising. The [B2B content marketing post](/insights/b2b-content-marketing-buyers-not-search) addresses content strategy that feeds both the ABM content library and the broader demand generation pipeline. The [marketing assessment](/tools/marketing-assessment) is a useful starting point for evaluating whether the right tracking and attribution infrastructure is in place before committing to full ABM investment. **Key Takeaways** - ABM measures success at the account level, not the contact level; the buying committee for enterprise technology decisions involves 6-10 stakeholders, making contact-level funnel metrics inadequate for enterprise pipeline management. - A real ICP is built from data on your highest-LTV existing customers: industry win rates, technology stack signals, organizational characteristics, and retention indicators, not just company size and revenue. - Tiered ABM scales through 1:1 for 10-30 strategic accounts, 1:few for 50-200 ICP clusters, and 1:many for broad ICP demand generation, with content personalization depth matching the tier. - Intent data from Bombora, G2, and similar providers is a prioritization layer, not a targeting system; high-intent accounts that don't match your ICP still have lower win rates. - LinkedIn Account Targeting lets you concentrate advertising spend specifically on buying committee members at your priority accounts, making it the most precise paid channel for ABM activation. - ABM success metrics are account engagement score, pipeline influenced, account-level win rate, and deal velocity, not leads generated or MQL volume." --- # Legal Client Journey Mapping: From Google Search to Signed Retainer URL: https://brianroseman.com/insights/legal-client-journey-search-to-retainer Published: 2026-03-28 Legal buyers research extensively. These touchpoint strategies guide prospects through a complex decision. "**Summary:** A potential legal client's journey from first Google search to signed retainer involves more touchpoints, more delays, and more decision friction than most law firms recognize. Mapping that journey in detail, then optimizing the specific points where people fall off, is the clearest path to growing caseload without simply spending more on advertising. **The Reality of the Legal Search Journey** The average person who ends up hiring an attorney doesn't decide in the first session. They search broadly at first, learn about their situation, search again with more specific terms, read three or four attorney websites, check reviews, watch a video or two, then finally contact someone, usually several days or weeks after the initial search. For personal injury, the consideration window can be 1-3 days after the incident when urgency is high. For family law matters like divorce, the window between first search and first contact often stretches 2-6 weeks. For estate planning, where there's no acute trigger, people who searched ""do I need a will"" can take 6-18 months before they contact anyone. [Clio's Legal Trends Report](https://www.clio.com/resources/legal-trends/) tracks this research cycle and consistently finds that legal consumers conduct significant self-education before reaching out. They want to understand their situation before they're willing to admit they need help. Content that educates without being condescending is what keeps your firm in the consideration set throughout that window. The touchpoints before first contact typically include: An organic or paid search result that introduces the firm. This is often a blog post or FAQ page, not the homepage. A visit to the attorney bio page to assess credentials, experience, and whether the attorney seems like someone they'd want to talk to. Attorney bio pages are consistently among the most-visited pages on law firm websites. A Google Business Profile check for reviews, photos, and the most recent activity. At least one more search for reviews on a third-party platform (Avvo, Google Reviews, Martindale-Hubbell). The decision to contact. Each of these steps is a potential drop-off point. Understanding which steps have the highest abandonment rate is where optimization focus should go. **Search Intent for Legal Queries: The Hierarchy** Legal search intent follows a predictable hierarchy, and what stage someone is at dramatically affects what content will convert them. Informational intent is the earliest stage: ""what is premises liability,"" ""how long does a divorce take in Missouri,"" ""can I sue for a slip and fall."" These searches come from people who are trying to understand their situation. Content targeting informational intent should educate comprehensively without a hard sales pitch. The goal is to be the resource they return to as their understanding deepens. The reason to invest in informational content: people who read your educational content before they contact you are pre-qualified in ways that cold PPC clicks aren't. They already trust you as knowledgeable. They already understand the basics of their situation. The initial consultation is more productive. Comparative intent is the middle stage: ""best personal injury attorney Kansas City,"" ""how to choose a family law attorney."" These searchers know they need an attorney and are evaluating options. Your competitive differentiation, attorney profiles, and case results content (within bar rules) need to win this comparison. Transactional intent is the final stage: ""personal injury attorney free consultation,"" ""divorce lawyer [city] contact."" These searchers are ready to call. They need frictionless contact, click-to-call, a form that takes 90 seconds to complete, and a response that comes within minutes, not hours. **Google Business Profile: The Underestimated Conversion Channel** For local legal searches, the map pack (the three-listing result that appears above organic results with a map) is often the first thing a potential client sees. Winning map pack visibility requires a well-optimized [Google Business Profile](https://support.google.com/business/answer/3038063). The map pack factors that matter most: Primary category accuracy is the foundation. ""Personal Injury Attorney"" vs. ""Law Firm"" as your primary category makes a material difference in which searches you appear for. Pick the category that matches your highest-value practice area. Review signals, including total count, average rating, and recency, are significant ranking factors. A firm with 85 reviews at 4.8 stars, with 15 reviews in the last 60 days, will typically outperform a firm with 200 reviews at 4.6 stars and no reviews in the last six months. Post frequency creates activity signals. Firms that post consistently (weekly or twice-weekly) maintain better map pack visibility than firms that post occasionally. Posts can be links to new blog content, legal tips, answers to common questions, or news about the firm. Photo count and quality. Google Business Profiles with 10+ high-quality photos (office, team, building exterior) show better engagement signals than profiles with 2-3 low-resolution images. **The Intake Process as a CRO Problem** Here is where most law firms lose a shocking percentage of the interest they successfully generated. Someone calls, no one answers. Someone fills out a form, they don't hear back for two days. Someone leaves a voicemail and gets a call back at 3pm when they're in a meeting. The data on response time impact in legal is stark. [Clio's research](https://www.clio.com/resources/legal-trends/) has consistently found that law firms responding to inquiries within five minutes have dramatically higher conversion to consultation than firms that respond in an hour, and dramatically better outcomes than firms that respond the same day or the next day. After one hour, a significant portion of potential clients have moved on to the next firm on their list. The specific intake bottlenecks to diagnose and fix: After-hours coverage. If you don't have someone answering the phone from 5pm-9pm and on weekends, you're missing a large share of potential client inquiries. Many people can only make personal calls outside work hours. An answering service with basic intake capability (not just voicemail) is a meaningful investment. Form response time. If your website form submissions sit in an email inbox until someone checks it in the morning, you're likely losing a significant portion of form completers. Setting up an automatic SMS or email acknowledgment that comes within two minutes (even just ""we received your inquiry and will call within one business hour"") dramatically reduces abandonment. Call-back reliability. If someone schedules a callback at 10am and you call at 11:15am, their impression of your firm is already damaged. Simple calendar-based call scheduling with reminder notifications to both parties is table stakes for modern intake. **Attribution Across the Long Consideration Cycle** How do you know which marketing investments are generating clients? In legal, this is harder than most businesses because the consideration cycle is long and multi-touch. A client might find you through a blog post in September, visit your Google Business Profile in October, and call in November after seeing your ad on Google search. Three different channels, one client. Last-touch attribution (crediting the Google Ads click in November) would tell you paid search generated this client. First-touch attribution would credit organic search from September. Neither tells the whole story. The practical approach for law firm attribution: Ask every new client how they found you and how long they'd known about the firm before calling. This qualitative data is imperfect but gives you directional signal that analytics alone can't capture. ""I've been following your blog for two months"" tells you something important about the content ROI. Call tracking with source attribution lets you see which campaigns, pages, and keywords generate phone calls. Dynamic number insertion (DNI) shows different phone numbers to visitors from different sources, tracking the source of every inbound call through to your CRM. Track consultation-to-retained client conversion rates by source. If your organic blog traffic generates consultations that convert to retained clients at 35%, and your paid search leads convert at 15%, that's directionally significant even if you don't trust the precise numbers. **Follow-Up Automation Within Bar Rules** Once someone has submitted an inquiry and you've had an initial consultation, what happens if they don't immediately sign? Most firms stop following up after one or two attempts. That's a mistake. The [ABA Model Rules of Professional Conduct](https://www.americanbar.org/groups/professional_responsibility/publications/model_rules_of_professional_conduct/) prohibit in-person solicitation of prospective clients in many circumstances, but they generally permit written follow-up with people who have already made contact with your firm and requested information. The post-consultation follow-up email sequence is generally permitted. A simple three-email sequence, spaced over two weeks after a consultation, can meaningfully improve your retained-client-to-consultation ratio without any compliance risk: Day 3: A brief email summarizing the key points of the consultation and offering to answer follow-up questions. This is educational and helpful, not a sales pitch. Day 7: A resource relevant to their specific situation. If they came in about an auto accident, a link to your FAQ about what to do after an accident or a post about what affects injury settlement amounts. Value-first follow-up. Day 14: A direct, no-pressure check-in. ""I wanted to reach out to see if you had any additional questions after our meeting. I know this is an important decision and I'm happy to talk through any concerns."" Direct, human, no urgency manufacturing. The [legal marketing authority post](/insights/legal-marketing-authority-bar-rules) covers the compliance framework for all of this in more depth. The [CRO calculator](/tools/cro-calculator) is useful for modeling what small improvements in consultation-to-retained conversion rate are worth in annual revenue terms. The [marketing assessment](/tools/marketing-assessment) includes a specific section on intake conversion that's often the highest-ROI area for law firms to improve. **Key Takeaways** - Legal consideration cycles range from 1-3 days for urgent matters like personal injury to 6-18 months for non-acute needs like estate planning; content must nurture at each stage, not just close. - The map pack for local legal searches requires a well-managed Google Business Profile: accurate primary category, recent reviews, consistent posting, and high-quality photos. - Responding to inquiries within five minutes versus one hour creates a measurable difference in consultation conversion rate; after-hours coverage for evenings and weekends captures a large share of potential clients who can only call outside work hours. - Multi-touch attribution is necessary for legal marketing because clients average 5-7 touchpoints before contact; call tracking and ""how did you hear about us"" intake questions give you directional data the analytics alone can't capture. - Post-consultation follow-up email sequences over 14 days are generally permitted under ABA rules and meaningfully improve retained-client conversion rates. - Conversion rate optimization on intake, specifically response time and follow-up sequences, is typically the highest-ROI improvement available to law firms before adding any new marketing spend." --- # Real Estate PPC: Geo-Targeting Strategies for Local Markets URL: https://brianroseman.com/insights/real-estate-ppc-geo-targeting-local Published: 2026-03-25 Real estate is hyperlocal. These geo-targeting tactics maximize leads in your service area. **Summary:** Real estate PPC is more expensive per click than most industries, with high-intent buyer and seller keywords running $8-15 per click in competitive markets. Most agents and brokerages waste a significant portion of that spend on geography that's too broad, audiences that aren't ready to transact, and landing pages that don't convert. This post covers the geo-targeting strategies, bid structures, and compliance guardrails that actually generate qualified leads. **The Geography Problem Most Real Estate PPC Gets Wrong** The instinct in real estate PPC is to target broadly. More geography means more impressions. More impressions means more leads. But in real estate, geography precision is the single biggest driver of lead quality, and broad targeting is the fastest way to generate high volume at poor conversion rates. The mismatch: someone searching "homes for sale in Kansas City" while sitting in Los Angeles might be considering relocation, but they're in a completely different buyer journey stage than someone in Overland Park searching "homes for sale 66062." The first searcher might be 12 months from a decision. The second might be making offers this weekend. The standard approaches to geo-targeting in real estate, ranked by precision: Radius targeting is the least precise option and works best for agents trying to establish general market presence. A 15-mile radius around your office sounds specific but in most metro markets includes hundreds of neighborhoods with wildly different price points, buyer profiles, and competition levels. Use radius targeting only for brand awareness campaigns with modest budgets. ZIP code targeting is the minimum level of precision for lead generation campaigns. A focused agent who has worked Brookside and Waldo in Kansas City for ten years knows those neighborhoods. Their PPC should target the ZIP codes where their local expertise shows in the conversion conversation, not the entire metro. Neighborhood targeting within Google Ads (using custom location groups or the location targeting refinements) lets you target specific neighborhoods that match your buyer and seller audience. This level of precision requires more setup but delivers leads who are searching with the specificity that matches your expertise. **Bid Adjustments for High-Value ZIP Codes** Beyond targeting, bid adjustments let you weight your spend toward the geography that performs. In Google Ads, location bid adjustments apply a multiplier to your base bid when a search comes from a specific location. The practical use: if you represent sellers primarily in Mission Hills and Leawood, Kansas, set positive bid adjustments (110-150% of base bid) for those ZIP codes. For ZIP codes where you're targeting but have lower expertise or lower price points, set no adjustment or a small negative adjustment to reduce spend without excluding the geography entirely. Price-point alignment in geo-targeting is underutilized. The buyer searching in a high-median-price ZIP code (say, $700,000+) has different economics than the buyer in a $250,000 median market. In the high-value ZIP, the commission on a single closed deal justifies much higher CPL thresholds. Segment your campaigns by price tier and set your target CPL accordingly, not as a single number across all campaigns. **Fair Housing Act Compliance in Real Estate PPC** This is the area where real estate agents most frequently get into legal trouble with digital advertising, and it's not always intuitive. The Fair Housing Act prohibits discriminatory advertising in housing-related transactions. For digital advertising, HUD has issued guidance that extends Fair Housing compliance to digital targeting. The [Google Ads housing advertising policies](https://support.google.com/google-ads/answer/2453995) explicitly prohibit targeting real estate ads based on race, color, national origin, religion, sex, familial status, or disability, and they prohibit using these characteristics as audience exclusions. This means you can't use demographic targeting to narrow your real estate audience to specific age groups, genders, or similar protected characteristics. What you can use: geographic targeting, keyword targeting, device targeting, and scheduling. What you cannot use: detailed demographic targeting by age, gender, parental status, or household income in ways that could function as proxies for protected characteristics. The ZIP code targeting discussed above has been scrutinized in Fair Housing enforcement when ZIP code selection functions as redlining by ZIP code. Avoid patterns where your exclusions map to concentrations of protected groups. Facebook and Instagram have faced specific Fair Housing enforcement actions. Meta's ad platform previously allowed age and demographic exclusions in housing ads that the Department of Justice found violated Fair Housing rules. As of 2022, Meta restricts housing advertisers from using certain demographic targeting options. Know the platform-specific restrictions because they're evolving. **Search Intent Segmentation: Buyer vs. Seller vs. Renter** The most important campaign segmentation decision in real estate PPC is separating buyer, seller, and renter intent into distinct campaigns with distinct landing pages, bid strategies, and budgets. Buyer intent keywords: "homes for sale [city/zip]," "houses for sale [neighborhood]," "buy a house [city]," "$[price range] homes [location]." These keywords drive the highest volume and the highest CPC ($4-10 for most local markets, $10-15+ for high-competition metros). Seller intent keywords: "home value estimate," "how much is my home worth," "sell my house [city]," "listing agent [city]." These are lower volume but extremely high value, because a listing is worth multiple buyer-side leads in commission. CPCs run $5-12 in most markets, with "home value" queries on the higher end. Renter intent keywords: "apartments for rent," "houses for rent [zip]," "rental homes [city]." CPC is lower ($1-3 typically) but the audience has very different economics for an agent-focused campaign. Property management companies benefit more from renter intent campaigns than individual agents. Mixing these intent types into a single campaign is one of the most common and costly mistakes in real estate PPC. A landing page optimized for buyer leads performs terribly when a seller clicks through on a "home value" query. Segment them from day one. **Landing Page Geo-Personalization** The landing page failure mode in real estate PPC: a visitor clicks an ad for "homes for sale in Prairie Village" and lands on a generic homepage that talks about "the Kansas City Metro." The geo-specificity that got the click disappears immediately, and so does conversion confidence. Geo-specific landing pages, or at minimum dynamically personalized landing pages that surface the neighborhood or ZIP code from the ad, close this gap. The basic approach uses URL parameters from the ad to populate the landing page headline dynamically: if the ad includes a parameter for the neighborhood, the landing page headline says "Prairie Village Homes for Sale" rather than "Kansas City Real Estate." More involved: neighborhood-specific landing pages with relevant content. Recent sales data for that neighborhood, walk score and school ratings, information about the specific streets and amenities of that area. These pages build relevance for both PPC landing and organic search. Dual-purpose content is always worth the investment. **Seasonal Bid Adjustments** Real estate seasonality is real and predictable in most U.S. markets. Spring (March-June) drives the highest search volume and the most competitive buyer activity. August-September sees a secondary activity bump as families who missed the spring window try to move before school starts. January-February is typically the lowest search volume period. Bid adjustments that match these patterns can significantly improve efficiency. Increasing bids 15-25% during peak spring activity captures more leads when competition is highest but also when buyer urgency is highest. Reducing bids during January-February by 10-20% preserves budget for the spring surge. Seasonal adjustment applies to budget too. Flat monthly budgets are often wrong for real estate. Moving budget from January to April, when the market is active and buyers are ready to transact, typically produces better annual results than spreading spend evenly. **CRM Integration to Exclude Existing Clients** One of the clearest wins in real estate PPC that most agents overlook: exclude your existing clients and past clients from your paid search targeting. Showing ads to someone you already represent is wasted spend. More importantly, it can create confusion about your relationship. Customer match in Google Ads lets you upload a list of email addresses and phone numbers to exclude from targeting. Pull your active clients, past clients, and anyone in active negotiations from your CRM monthly and upload them as exclusion lists. The match rate will be 40-60% of your list, but that's still meaningful waste reduction. The [PPC calculator](/tools/ppc-calculator) can help model the budget allocation across buyer vs. seller campaigns and estimate what CPL is sustainable at your average commission and close rate. For building the broader measurement framework, the [marketing assessment](/tools/marketing-assessment) is a good diagnostic. As PPC costs continue to climb in competitive real estate markets, the [technology PPC cost trends post](/insights/technology-ppc-cost-per-lead-climbing) covers the broader context of rising CPLs across B2B categories that real estate shares. **Measuring Cost Per Listing Appointment, Not Cost Per Click** The final frontier in real estate PPC measurement: stop measuring success at cost per click or even cost per lead, and measure cost per listing appointment and cost per closed transaction. The funnel in a listing campaign: clicks, lead form submissions, phone call responses, consultation scheduled, presentation given, listing agreement signed, property listed, property sold, commission paid. Most real estate PPC is measured at step two (lead form submissions). The actual business outcome is at step nine. Tracking further down the funnel requires your CRM to capture not just leads, but their progression through your sales process. The Google Ads platform supports conversion actions for phone calls (through call tracking numbers), form submissions, and even offline conversion import (where you upload closed-deal data back to Google Ads to inform Smart Bidding). Offline conversion import is significantly underused in real estate advertising and gives Google's algorithm the data it needs to optimize toward actual closed deals rather than just form fills. **Key Takeaways** - ZIP code targeting is the minimum precision for real estate lead generation campaigns; radius targeting generates volume without quality and should be reserved for brand awareness. - Fair Housing Act compliance prohibits demographic targeting by protected characteristics in real estate ads; Google and Meta both enforce this with platform-specific restrictions that differ and are evolving. - Separate buyer, seller, and renter intent into distinct campaigns with distinct landing pages; mixing them in one campaign guarantees the landing page works for one audience and fails the others. - High-value ZIP code bid adjustments (110-150% multiplier) align budget weight with your actual area of expertise and commission potential. - Customer match exclusion lists for existing clients, updated monthly from your CRM, reduce wasted spend on people you already represent. - Track cost per listing appointment and cost per closed deal, not just cost per lead; offline conversion import in Google Ads gives the algorithm the signal it needs to optimize toward actual business outcomes. --- # Travel SEO: Competing With OTAs for Destination Keywords URL: https://brianroseman.com/insights/travel-seo-competing-ota-destinations Published: 2026-03-21 Expedia and Booking.com dominate travel search. These strategies help hotels and DMOs win visibility. "**Summary:** OTAs spend more on Google than most hotel chains spend on everything else combined. Competing directly against Expedia and Booking.com for broad destination keywords is a losing fight for most travel brands. But winning the right battles, in the right content areas, with the right technical foundations, creates real direct booking volume that bypasses commission fees entirely. **Why Direct Bookings Are Worth Fighting For** OTA commission rates are not a secret, but the full math often isn't. Expedia typically charges hotels 15-30% commission on bookings, with rates depending on property size, market, and the specific program terms. Booking.com runs a similar range. A $200/night hotel room that books through an OTA at 20% commission costs the hotel $40 per booking, every booking, forever. A direct booking at $200 costs the hotel whatever the marketing spend was to generate it, but that customer is now in the hotel's CRM. They can be emailed, retargeted, offered a loyalty program, and brought back for the next booking without another commission payment. The second and third direct bookings from that customer are essentially free acquisition because you already have the relationship. [Skift](https://skift.com/) has tracked this tension extensively: OTAs provide genuine distribution value, especially for filling shoulder periods and reaching international markets that smaller properties can't reach on their own. The question isn't whether to list on OTAs, it's how much booking volume you're willing to cede to commission versus building direct channels that compound over time. For a property doing 10,000 room nights per year at an average $180 rate, moving 15% of bookings from OTA to direct, at an average 20% commission, saves $54,000 per year. That's a substantial marketing budget for direct channel investment. **The Keyword Strategy OTAs Can't Win** Here's where hotels and DMOs have an advantage OTAs can't easily replicate: local knowledge and depth. An OTA listing for a hotel in Sedona, Arizona, tells you the star rating, the amenities, and the price. It doesn't tell you that the property has a trail access point directly behind the spa, or that the restaurant sources produce from a farm eight miles away, or that the pool is positioned to watch the sunset over Cathedral Rock on clear evenings. That local, specific, narrative content is something the hotel can create and OTAs fundamentally can't. The keyword strategy that exploits this: Informational destination content targets searches that happen 3-8 weeks before a traveler is ready to book. ""Best hiking near Sedona"" or ""Sedona restaurants with views"" attracts travelers in research mode. An OTA isn't going to create that content because it doesn't have the local perspective and the content doesn't directly transact. A destination hotel with a good blog can rank for dozens of these queries and build a relationship with potential guests long before they're comparing prices. Local authority content, like neighborhood guides or things-to-do pages that are genuinely useful and link to actual local businesses (not just your property), builds the kind of topical authority that pushes your content up in destination searches. This is content that earns links from local news and tourism organizations, which OTA listing pages can't match. Long-tail transactional queries are where you can compete most effectively on the ""ready to book"" end of the spectrum. ""Pet-friendly hotels in Sedona with hiking access"" or ""Sedona cabin with private hot tub"" are queries where a specific property with the right attributes can rank above OTA category pages by being more specifically relevant. **Schema Markup for Travel: What Actually Matters** [Google's structured data documentation for vacation rentals and hotels](https://developers.google.com/search/docs/appearance/structured-data/vacation-rental) covers the Hotel schema and related types. Getting this right affects your visibility in Google's hotel price search features, which appear prominently for accommodation searches. The Hotel schema should specify: name, address, telephone, priceRange, amenities (using the amenityFeature property), check-in/checkout times, and the geo coordinates. The geo coordinates matter for Google's hotel search results positioning. For vacation rental properties, the VacationRental schema type (introduced by Google in 2021) is the appropriate markup. It extends the Accommodation schema with rental-specific properties like rentalType and leaseLength. TouristAttraction schema is relevant for DMOs and destination content sites marking up local points of interest. When Google understands that your site is an authoritative source about specific attractions, it factors that into destination content rankings. The practical implementation: most hotel property management systems and booking engines can generate schema markup automatically if configured correctly. Third-party SEO tools like Semrush or Ahrefs can audit whether your schema is properly implemented and validating without errors. **Google's Travel Features and How to Get Into Them** Google has invested heavily in travel-specific search features: the hotel price module (showing rates and availability from multiple sources inline in search results), the ""Things to Do"" feature for destination queries, and the hotel finder map that appears for accommodation searches. For the hotel price module, your rates need to be connected to Google Hotel Center, either directly or through a channel manager that integrates with it. Brands that show their direct booking rate in Google Hotel Center can compete with OTA rates in the module directly, which is one of the most direct ways to capture guests who would otherwise click an OTA listing. For ""Things to Do"" listings, which show activities and attractions alongside hotel and flight results for destination queries, the appropriate schema markup on your attraction or activity pages is the required entry point. If you operate tours, activities, or experiences, connecting through the Google Things to Do program is worth prioritizing. The [Google Business Profile](https://support.google.com/business/answer/3038063) remains essential for local hotel visibility. For boutique properties, the map pack appearance for ""hotels in [destination]"" local searches drives meaningful direct booking traffic. Consistent profile information, photo quality, and review management all affect your position in these results. **Local SEO for Hotels: Beyond the Basic Listing** OTA listing pages for a hotel often outrank the hotel's own website for branded searches. This is a well-documented problem. The reason: OTA domain authority is substantial (Booking.com and Expedia are among the highest-authority domains on the internet), and they actively optimize their property pages. But for local, non-branded searches, the hotel's own site has a geography advantage. Google's local search algorithm weights signals that OTA pages, which aren't locally hosted or locally operated, can't generate as strongly: NAP consistency (Name, Address, Phone) across all directories and citations is the foundation. Inconsistencies in how your property's address is listed across directories confuse local search engines and dilute your local authority. Review velocity and recency matter more than total review count. A property with 200 reviews, all from 2022, is less competitive than a property with 140 reviews, with 30 in the last three months. A systematic review solicitation program, combined with a thoughtful response to every review (positive and negative), is the consistent driver of local search performance. Local content about the specific neighborhood, surrounding attractions, and local events builds geographic relevance signals that OTA pages simply don't have. A blog post about the annual hot air balloon festival visible from your rooms is not going to appear on your Booking.com listing. **Booking Abandonment Recovery** Even when you succeed in driving direct traffic, you'll lose a significant portion of that traffic before they book. Standard travel industry booking abandonment rates run 80-90%, higher than most e-commerce categories, because travel purchases are high-consideration and comparison-heavy. The recovery approach: Retargeting campaigns to abandoning visitors, showing the specific property or room type they were looking at, with a clear price and a limited-time direct booking incentive (free upgrade, breakfast included, waived parking fee), consistently outperform generic property retargeting. Email recovery for visitors who got to the point of entering their email address in a booking flow should trigger within two hours of abandonment. The sooner the email, the higher the recovery rate. After 24 hours, recovery rates drop dramatically as the traveler has likely either booked elsewhere or moved on. Price-match guarantees for direct bookings have become table stakes. If potential guests know that booking direct at $199 is the same as booking through Expedia at $199 (after the OTA's loyalty discount), you haven't given them a reason to avoid the OTA they're already loyal to. A direct booking benefit, even a small one, tips the decision. For the SEO and conversion rate optimization connection, the [SEO ROI calculator](/tools/seo-roi-calculator) helps quantify the value of organic search improvements in financial terms that justify content and technical SEO investment. The [CRO calculator](/tools/cro-calculator) is useful for estimating the revenue impact of improving booking conversion rates at different traffic volumes. The [e-commerce CRO post](/insights/ecommerce-cro-tests-that-matter) covers A/B testing methodology that applies directly to booking flow optimization. For the [fractional SEO approach](/services/fractional-seo) to destination and hotel content strategy, the engagement model makes sense for properties that need ongoing content and technical SEO support without a full-time in-house team. **Key Takeaways** - OTA commission rates of 15-30% per booking make direct channel investment economically significant; moving 15% of bookings to direct on a 10,000-room-night property saves $54,000 per year at 20% commission. - Informational destination content (hiking guides, restaurant roundups, local experiences) attracts travelers 3-8 weeks before booking and builds audience relationships that OTA listing pages can't create. - Hotel schema markup connected to Google Hotel Center enables your direct booking rate to appear in Google's hotel price module, competing with OTA rates at the point of decision. - Local SEO advantages (review recency, geographic content, NAP consistency) favor the hotel's own site over OTA property pages for non-branded destination searches. - Booking abandonment recovery via retargeting and email within two hours of abandonment are the highest-ROI conversion opportunities in direct booking optimization. - A direct booking benefit, even a small one (free upgrade, waived parking, breakfast), is necessary to give loyalty-program OTA users a reason to choose direct over their familiar booking platform." --- # Energy Marketing Automation: Reducing Churn in Deregulated Markets URL: https://brianroseman.com/insights/energy-marketing-automation-churn-reduction Published: 2026-03-18 Customer switching is at an all-time high. These retention sequences keep customers from shopping rates. **Summary:** Deregulated energy markets present a unique churn challenge where price sensitivity and contract expirations drive annual customer loss rates as high as 30%. Marketing automation, predictive modeling, and smart meter data can identify at-risk customers and trigger personalized renewal campaigns weeks before the decision point. This post covers how to build that system and measure its impact. **The Commodity Trap in Deregulated Energy** Walking into a retail energy provider's marketing department in a deregulated state like Texas or Ohio feels different than a standard B2B environment. In SaaS, you're selling a solution. In energy, you're often selling a commodity where the customer only notices you when the lights go out or the bill spikes. This commodity trap means that price is the primary lever, but relying on price alone is a race to the bottom that kills margins. I remember working with a regional energy provider in 2024 that was losing 24% of its residential base every single year. They were spending millions on acquisition just to stay flat. The churn isn't random. It follows a highly predictable pattern tied to contract end dates and seasonal usage spikes. In deregulated markets, the teaser rate is the standard entry point. When that 12-month fixed rate expires and the customer rolls onto a variable market rate, their bill can jump 40% overnight. That's the moment they head to a comparison site. By the time they see that high bill, you've already lost the chance to save them. True churn reduction in this space requires moving the intervention point 60 to 90 days before the contract expires. **Data-Driven Churn Prediction Beyond Contract Dates** While contract end dates are the most obvious signal, they aren't the only one. Smart meter data, often referred to as AMI (Advanced Metering Infrastructure) data, provides a goldmine of behavioral signals if you know how to use it. According to [energy.gov](https://www.energy.gov/), smart meter penetration has now reached the majority of U.S. households, with deployment continuing across all major deregulated markets. This data allows energy retailers to see usage patterns in 15-minute or 1-hour intervals. The signals worth watching beyond contract dates: Usage spikes that don't correlate with seasonal weather patterns often indicate appliance failure or a change in household dynamics. If a customer's usage suddenly increases by 30% during a mild weather month, something changed in their home. If the energy provider doesn't reach out with a helpful "we noticed a spike, here's how to save" message, the customer sees a high bill and blames the provider. Year-over-year usage decline is actually a churn predictor, not a success story. When a residential customer's usage drops significantly, they may have installed solar, added efficiency upgrades, or moved. The customers who invested in solar are particularly likely to switch or reduce their reliance on grid power entirely. Call center contact patterns are a strong leading indicator. A customer who calls about a billing dispute is two to three times more likely to churn within 90 days than a customer who hasn't contacted support. If your CRM tracks service interactions alongside billing data, you can create a composite churn risk score that weights these contact patterns. Customers who have received a proactive usage alert are 18% less likely to churn at the end of their contract compared to those who only received standard billing communications. That's not a modeled projection; it's a result we measured by comparing renewal rates across customer segments at the regional provider mentioned above. **Building the Automated Renewal Engine** Marketing automation in energy isn't just about sending emails; it's about orchestrating a multi-channel sequence that adapts to the customer's engagement level. A standard renewal sequence should start 90 days out. **90 Days Out: The Education Phase** Start with a soft touch. Don't talk about renewal yet. Send a "Year in Review" report. Show them how much energy they used, how they compared to efficient neighbors, and provide a few personalized tips based on their actual usage peaks. If they have high cooling loads, talk about smart thermostats. This builds helpful authority before you ask for a commitment. Open rates on these usage-based emails are typically 2-3x higher than standard promotional emails because the content is personally relevant. **60 Days Out: The First Offer** Present the renewal options. But don't send a generic link. Use automation to pre-select the "Best Value" plan based on their past 12 months of usage. If they're a high-volume customer, a fixed-rate plan with a lower per-kWh price is the winner. If they're a low-usage "empty nester," a plan with no base fee might be more attractive. The offer presentation matters: frame it as "we built this for how you actually use energy," not "here are three plan options." **30 Days Out: The Multi-Channel Push** If they haven't renewed by the 30-day mark, move beyond email. SMS and outbound IVR (Interactive Voice Response) have significantly higher conversion rates for renewal campaigns than follow-up emails to people who didn't open the first three. An SMS with a "Reply YES to lock in your current rate" option removes the friction of a portal login. However, be extremely careful with TCPA compliance. Ensure your automation platform has a hard stop feature if a customer revokes consent. **Regulatory Constraints and Compliance Automation** One of the biggest hurdles in energy marketing is the sheer volume of state-specific regulatory requirements. Each deregulated market has different rules about when and how a customer must be notified of a rate change. In some markets, a Notice of Contract Expiration must be sent via physical mail. Your automation shouldn't just handle the digital side; it should trigger the physical mail too. We used a direct mail API integrated with the CRM to trigger the official state-required notice exactly when the regulations dictated, while the digital automation handled the persuasion side. This ensures the customer gets the legally required notice and the marketing-optimized offer simultaneously, reducing the confusion that often leads to churn. State commissions in Texas (PUCT), Illinois (ICC), and Ohio (PUCO) each have specific disclosure requirements. If you're operating across multiple states, your automation platform needs to handle state-level segmentation for compliance communications, not just for marketing personalization. **The Role of Win-Back Automation** No matter how good your retention strategy is, some customers will leave. The goal then shifts to win-back automation. Most energy providers wait six months to try to win back a customer. That's too long. The best window is 14 days after they leave. Why? Because that's when they receive their final bill from you and their first bill (which often includes a setup fee) from the new provider. A win-back campaign should focus on the hassle factor. Use automation to send a message that acknowledges the move, avoids being defensive, and offers a concrete incentive: "We miss you. Switching back takes 30 seconds, and we'll waive your first month's base fee to cover that other provider's setup charge." By using the [LTV calculator](/tools/ltv-calculator), you can determine exactly how much of a signing bonus you can afford to offer a win-back customer while still maintaining profitability. The economics of win-back are usually favorable. A residential energy customer who was with you for two years has already paid their acquisition cost. Win-back cost is typically 60-70% lower than new customer acquisition cost because you already have their identity, history, and billing relationship. The offer just needs to overcome inertia. **Measuring the Impact of Automation on Churn** How do you know if it's working? You can't just look at total churn rate, because that's influenced by market prices you can't control. Instead, look at segmented churn. Compare the churn rate of customers who went through the personalized automation sequence versus a holdout control group that received standard bill-only communication. The holdout test design is critical. Don't make your holdout group 5% of customers and your treatment group 95%, then celebrate when the numbers look good. Run a true 50/50 split within comparable segments, match on contract end date concentration and usage profile, and run the test for at least one full contract cycle (12 months). In my experience, a well-orchestrated automation strategy can reduce voluntary churn by 15% to 22% within the first 12 months. When you consider that a typical energy customer might have a [lifetime value](/) of $1,200 to $2,500, a 5% reduction in churn across a base of 100,000 customers is worth millions in retained revenue. The LTV math is the same whether you're in energy or e-commerce: small improvements in retention rate have outsized effects on cumulative revenue. **Predictive Modeling: The Next Level** The step beyond rules-based automation is predictive churn scoring. By feeding historical usage, billing, and interaction data into a machine learning model, you can assign a churn risk score to every customer on a weekly or daily basis. The difference from simple rules: the model learns the combination of signals that predicts churn in your specific market, not just the obvious ones like contract end date. A customer whose churn risk score spikes from 20 to 85 because they called the call center twice in three days and their usage is up 30% shouldn't receive the standard educational email. They should trigger an immediate high-value retention offer or a loyalty specialist callback. This is where [fractional analytics expertise](/services/fractional-analytics) becomes critical: building the bridge between raw smart meter and CRM data and the marketing automation platform. Most energy retailers have the data. Few have built the pipeline to make it actionable in real time. The [marketing assessment](/tools/marketing-assessment) we use as a diagnostic tool helps identify where in this data-to-automation chain the biggest gaps are. For most retail energy providers we've worked with, the data exists but the CRM-to-automation integration is the weak link. The segmentation logic is running on contract end date alone, not the richer behavioral signals that predictive models use. **Key Takeaways** - Deregulated energy markets face annual churn rates of 20-30%, primarily driven by contract expirations and the price shock when variable rates kick in after the teaser period. - Proactive intervention should begin 90 days before contract expiration, starting with value-add usage reports before any renewal ask. - Smart meter AMI data predicts churn through usage anomalies, year-over-year decline, and customer service contact patterns, not just contract end dates. - Multi-channel automation (email, SMS, direct mail for regulatory compliance) is required in most deregulated markets, but TCPA and state-specific regulatory compliance must be built into the automation logic. - Win-back campaigns are most effective 14 days after departure, when the first competitive bill arrives; the economics favor win-back over new acquisition at a 60-70% cost differential. - Measure success through segmented churn analysis with a true holdout control group, not overall churn rate, which fluctuates with market pricing you can't control. --- # Manufacturing Trade Show ROI: Measuring Beyond Badge Scans URL: https://brianroseman.com/insights/manufacturing-trade-show-roi-beyond-scans Published: 2026-03-14 Trade shows cost $50K+ but ROI is murky. These attribution methods prove actual pipeline impact. "**Summary:** Trade show badge scans feel like a success metric because they're easy to collect and the numbers look impressive in post-show reports. They're not a success metric. A manufacturing company spending $180,000 on a major industry show needs to know whether that spend generates profitable pipeline, not how many badges were scanned at the booth. This post covers a measurement framework for doing exactly that. **Why Badge Scans Tell You Almost Nothing** Here is what a badge scan tells you: a person was physically present at your booth and either volunteered their badge or you scanned it without their explicit engagement. That is all. It doesn't tell you their buying authority. It doesn't tell you their timeline. It doesn't tell you whether they're a competitor, a student, a journalist, or a potential $2 million OEM customer. At a major manufacturing show like FABTECH or the International Manufacturing Technology Show (IMTS), a booth with a compelling display and some swag can scan 800 badges in three days and generate two qualified opportunities. Meanwhile, a smaller, more intentional booth might scan 150 badges and generate thirty. The industry average that [CEIR's research](https://www.ceir.org/research) has tracked consistently is that roughly 15-20% of booth visitors represent genuine potential buyers. The rest are a mix of the curious, the competitive, and the people who needed a moment off the show floor. When you measure success by badge count, you're celebrating that 80-85% who aren't buyers. **Building the Measurement Framework Before You Go** The most common measurement mistake in trade show marketing is deciding what you're going to measure after you come back. By then, you've lost the ability to capture half the data you need. Before the show, define: What constitutes a ""qualified show lead"" for your company. This should be specific. For most manufacturers, a qualified lead means: decision-making authority or direct influence over the purchase, confirmed budget for the relevant product category, and an identified project or need within 12 months. That's a three-variable filter. Apply it during booth conversations, not afterward. Your cost baseline. Total out-of-pocket cost for the show: registration, booth construction and shipping, travel, hotel, staff time loaded at fully-burdened rates, print materials, hosted events. Everything. For a major show, this number is often 40-60% higher than what teams report because they don't count staff time. If six people attend a four-day show and their loaded cost is $250/hour, that's $48,000 in staff costs before you add anything else. Your target cost per qualified opportunity (CPO). If your average manufacturing deal takes 9 months to close and is worth $850,000 in revenue at 34% gross margin, you can justify a substantially higher CPO than a company selling $40,000 repeat-purchase components. Know your math before you know your number. **Pre-Show Digital Strategy** The firms that get the most from major manufacturing trade shows don't start their show marketing at the booth. They start it six to eight weeks before the show opens. [IMTS](https://www.exhibitoronline.com/), FABTECH, and other major industry shows publish attendee lists and exhibitor directories months in advance. Most shows have their own apps and digital event platforms that let you schedule meetings, message other attendees, and identify relevant sessions. Using these tools aggressively before the show is both more efficient and more welcomed than cold outreach, attendees are in ""event mode"" and more receptive to pre-show connection requests than the same request sent randomly in February. Targeted LinkedIn campaigns to the show's attendee audience (use the event page's attendee list as a targeting proxy, or target by job title and industry in the show's geographic concentration) can run in the four weeks before the show. The goal isn't to sell anything, it's to put your name and capability in front of people before they walk the floor, so when they see your booth, there's a recognition factor. Pre-scheduling meetings from your existing customer and prospect list transforms your booth from a passive waiting game into an active selling environment. If you've got 40 pre-scheduled meetings over three days, your show ROI is largely locked in before anyone badges. **The Follow-Up Sequence Timing** Speed after the show matters more than most teams appreciate. The industry data from [Exhibitor Magazine](https://www.exhibitoronline.com/) and every sales optimization study I've seen points to the same window: the first 48 hours after a show closes are when the highest percentage of show conversations still result in meetings. By day five, a significant portion of attendees have moved on to the next operational priority, and your show conversation is a fading memory. The sequence that works: Day 1-2 post-show: Personalized follow-up to every qualified lead. Not a templated ""great to meet you at the show"" mass email. A specific reference to the conversation you had, the problem they described, and a direct next step offer. If you discussed a specific application, the next step should be something concrete: a sample, a call with your application engineer, a site visit. Day 3-7: Nurture content for leads that didn't respond to the initial follow-up. This is where marketing automation should run a sequence based on what the lead's interests were. If they were interested in automation solutions, the nurture content should address automation-specific ROI. If they came to your booth because of a specific material challenge, content about that material. Week 2-4: Account-level outreach if the lead has gone dark. Check whether others from the same company attended the show. A multi-contact approach at the account level is often more effective than repeatedly reaching the same person. **CRM Integration: The Foundation of Real Attribution** Every show lead needs to be in your CRM within 24 hours of first contact, with the show tagged as the source and the specific conversation notes from the booth captured. This sounds basic. Most manufacturing companies don't do it consistently. The reason it matters: with a 6-18 month sales cycle, you cannot rely on your team's memory to connect a $1.4 million deal that closed in Q3 to a show lead that entered the funnel in Q1 of the previous year. The CRM is the record. If the CRM doesn't have the show tagged as source with the initial qualification notes, that deal gets attributed to ""direct"" or ""existing relationship"" and the show doesn't get credit. Set up your CRM campaign tracking before the show, not after. Create a specific campaign for each show, with lead source tags for show leads, and train your booth staff on how to enter leads in the moment (or within 30 minutes of the conversation). Card-scanning apps that sync directly to Salesforce or HubSpot are the minimum viable solution. Some shows offer direct CRM integrations from their lead retrieval systems. The [PPC calculator](/tools/ppc-calculator) is useful here as an analog for thinking about trade show CPL: the math for evaluating whether you're getting acceptable cost per qualified lead from trade shows is identical to the math for evaluating a paid search campaign. Cost in, qualified opportunities out, revenue per opportunity, win rate. If the trade show numbers don't compete with other channels on this basis, that's important data. **Calculating True Cost Per Qualified Opportunity** The calculation that actually matters: Total show investment (all costs, including staff time) divided by the number of qualified opportunities created within 90 days of the show gives you your cost per qualified opportunity. For context: a manufacturing company spending $200,000 on a show that generates 15 qualified opportunities has a $13,333 CPO. Whether that's good depends entirely on the deal economics. If average deal revenue is $500,000 with 35% gross margin, that's $175,000 gross margin per deal. A $13,333 CPO to generate $175,000 in gross profit is a 13:1 return on CPO before accounting for win rate. If you win 30% of qualified opportunities, your effective CPO-to-gross-margin ratio is still well above 4:1. If average deal size is $40,000 with 25% gross margin ($10,000 gross profit), that same $13,333 CPO is economically difficult to justify regardless of win rate. Know your deal economics before you commit to show spending. The math should drive the decision, not the habit of attending shows because ""we've always done it."" **Long Sales Cycle Considerations** The biggest attribution challenge in manufacturing trade show measurement is the time gap. A deal influenced by a show contact in September might not close until the following June. By that point, the team has attended two more shows, the marketing attribution window has closed in most CRM configurations, and the connection to the original show is lost. The practical solutions: Extend your attribution window in your CRM for show-sourced leads. Set show campaign attribution to 18 months, not the default 30-90 days. Do quarterly pipeline reviews specifically for show-sourced leads. Track each qualified opportunity from each show through the full sales cycle. What percentage are still active? What stage are they in? Are show-sourced deals advancing at a similar rate to other channel-sourced deals? Track ""deals closed where show contact was in the influence chain"", not just deals where the show was the direct source. In manufacturing sales with long cycles and multiple stakeholders, the show might have been where you met one engineer who later became an internal advocate for your solution, while the formal opportunity was sourced through a dealer network. Both are attributable to the show. The [B2B content marketing post](/insights/b2b-content-marketing-buyers-not-search) addresses the broader challenge of connecting early-stage engagement to long sales cycles, which applies directly to trade show follow-up content strategy. **Competitive Intelligence: The Underrated Show ROI** One of the clearest, most direct returns from trade show attendance that never shows up in standard ROI calculations: competitive intelligence. At a major manufacturing show, you can observe your top five competitors' booth positioning, messaging, new product announcements, and pricing signals in a single day. This intelligence is genuinely valuable. Knowing that a competitor has repositioned around automation integration six months before their press release changes your competitive response options significantly. Assign one or two team members specifically to competitive intelligence at every show. Give them a structured debrief template. Capture booth traffic patterns (are they busy or empty?), new product demonstrations, messaging changes, pricing materials they're distributing. Bring it all back, synthesize it, and distribute it to your product and sales teams within a week of the show. **Key Takeaways** - Badge scan counts are vanity metrics; the only useful leading indicator is the number of conversations that meet your pre-defined qualified lead criteria (buying authority, budget, project timeline within 12 months). - Calculate your total show cost including fully-burdened staff time, which is often 40-60% higher than out-of-pocket costs alone. - Pre-show digital strategy starting 6-8 weeks before the event, including pre-scheduled meetings from your existing pipeline, often determines more of your show ROI than anything you do at the booth itself. - Follow up within 48 hours with personalized notes referencing the specific conversation, not templated mass emails; response rates drop significantly after day five. - Extend CRM attribution windows for show-sourced leads to 18 months given manufacturing sales cycles; quarterly pipeline reviews for show cohorts prevent attribution loss over long deal timelines. - Competitive intelligence from trade shows is a legitimate, quantifiable ROI component that should be captured systematically and distributed to product and sales teams within a week post-show." --- # Pharma HCP Marketing: Reaching Physicians in a Post-Cookie World URL: https://brianroseman.com/insights/pharma-hcp-marketing-post-cookie Published: 2026-03-11 Third-party cookies are dying. These targeting strategies reach healthcare professionals without them. **Summary:** Pharmaceutical HCP marketing has always been heavily regulated. But 2022 through 2024 brought a specific shock: HHS guidance on HIPAA-regulated tracking technologies made many standard digital marketing practices genuinely risky, even for pharma companies marketing to physicians rather than patients. This post covers what changed, what's still permitted, and what the alternatives look like for reaching physicians in a post-cookie environment. **What the HHS Tracking Guidance Actually Changed** In December 2022, the Department of Health and Human Services published guidance clarifying that standard web tracking technologies , including Google Analytics and Meta Pixel , can violate HIPAA when deployed on websites of HIPAA-covered entities or their business associates. The guidance was updated in March 2024 to add nuance, but the core concern remained: when a user visits a clinical page (say, a condition page or a medication information page) and tracking tools collect their IP address alongside that URL, that combination can constitute individually identifiable health information (IIHI). For pharmaceutical companies marketing to patients, the implications were immediate and significant. [HHS's HIPAA online tracking guidance](https://www.hhs.gov/hipaa/for-professionals/privacy/guidance/hipaa-online-tracking/index.html) created significant operational challenges for health system marketing teams. But pharma companies marketing to healthcare professionals occupy a different and genuinely complicated regulatory position. If you're running HCP-only digital programs , content behind a registration wall requiring NPI verification, for instance , are you subject to the same tracking restrictions? The answer depends heavily on whether you or your media partners are classified as covered entities or business associates, and whether the URLs involved could implicate patient information. The practical effect: pharma marketing and legal teams got much more conservative about tracking technologies across all digital properties, not just patient-facing ones. Programs that were running standard display retargeting through Google Ads or Meta were paused for legal review. Many have been restructured or haven't fully restarted. **The Traditional HCP Channels That Still Work** Before getting to alternatives, it's worth being clear about what hasn't changed: the core value of reaching physicians through channels they actually use for professional information. Medical journal advertising , both print and digital , remains a compliant, effective channel for reaching HCPs in specific specialties. The Journal of the American Medical Association, New England Journal of Medicine, and specialty publications have audiences that are self-selected by professional interest in the relevant clinical area. They're expensive on a per-impression basis and can't be targeted as precisely as digital, but they carry no HIPAA tracking risk and reach physicians in a professional context. CME (Continuing Medical Education) sponsorships are another channel that has become more attractive as digital targeting has gotten complicated. Accredited CME programs attract physicians who are actively engaged in professional development. Sponsorship visibility isn't direct advertising in the traditional sense, but brand and product awareness in a professional learning context can be significant. Medical conference marketing , booth presence, symposia, sponsored dinners , is operational and compliant. The scale is limited compared to digital reach, but the quality of engagement is high. A 45-minute industry symposium at a major cardiology meeting gets you face-time with highly engaged physicians that no digital channel matches. **NPI-Based Targeting: How It Works** The most significant development in HCP digital marketing over the past four years is the rise of NPI-based targeting as a first-party or second-party data strategy. Every licensed U.S. physician has a National Provider Identifier , a unique ten-digit number in a publicly searchable federal database. Pharma marketers have long used NPI-based lists for direct mail targeting. Now they're being used for digital. The mechanism: Pharma companies build or license lists of HCPs with specific NPI numbers , targeting, for instance, all board-certified rheumatologists in markets where a drug is available. Those NPI lists get matched to identity graphs maintained by HCP-focused ad networks, which can then target those specific physicians with digital ads without relying on third-party cookie tracking. [Doceree](https://www.doceree.com/) is one of the platforms built specifically for NPI-based HCP programmatic targeting. Their inventory spans point-of-care platforms, medical journal digital properties, and HCP-specific content sites. The key compliance difference from standard display advertising: the targeting is done through NPI-matched first-party or second-party data, not behavioral tracking that could involve patient information. The targeting precision isn't perfect. NPI-to-cookie or NPI-to-device matching has real match rate limitations , typically 40-70% of a target list gets matched to addressable digital identities. But for brand awareness campaigns targeting a specific specialty, the ability to reach physicians across their professional digital touchpoints without HIPAA-problematic tracking is genuinely valuable. **Doximity: The Platform Most Pharma Marketers Underuse** Doximity is a physician-only professional network with over two million verified U.S. physicians , representing roughly 80% of the practicing physician population. Because Doximity verifies physicians through NPI and license verification, its audience data is inherently HCP-specific. There's no risk of inadvertently reaching patients. Doximity's advertising platform allows targeting by specialty, geographic region, years of experience, practice setting (academic medical center, private practice, hospital-employed), and NPI list match. The ad formats include sponsored content that appears in the Doximity news feed, which physicians use for clinical news and continuing education. The response rates on Doximity sponsored content are substantially higher than standard display advertising , physicians are engaged with the content around it, and the targeting is precise enough to reach the right specialty without waste. The CPMs are higher than broad-reach digital, but the audience quality is comparable to medical journal digital advertising. Compliance: Because Doximity's platform is physician-only with verified identities, you're not collecting identifiable patient data. The HIPAA tracking concerns that affected patient-facing digital marketing don't apply in the same way to an HCP-only platform with verified professional identities. **MLR: The Real Constraint on Digital HCP Content** For anyone outside pharma marketing, the Medical Legal Review (MLR) process is the thing that most surprises them about how slowly pharma digital moves. Every piece of content , a landing page, a banner ad, a LinkedIn post from a medical science liaison , must go through MLR before it can be published or distributed. MLR reviews typically involve medical affairs, legal, and regulatory reviewers. The medical review ensures clinical accuracy and fair balance (required by FDA regulations for prescription drug promotion). The legal review evaluates liability exposure. The regulatory review checks compliance with FDA promotional guidance. The [FDA's prescription drug advertising guidelines](https://www.fda.gov/patients/drug-development-process/step-3-clinical-research) govern what claims can be made about approved drugs , including the requirement for fair balance (disclosing risks proportionate to the benefits being claimed) and the prohibition on making false or misleading claims. For digital content specifically, MLR creates a timing problem: social media and digital environments move faster than MLR cycles. A topic that's trending in physician communities might be irrelevant by the time content clears review. The pharma companies that navigate this best have pre-approved content frameworks and modular content systems , pre-cleared claims and graphics that can be assembled into new combinations faster than building from scratch. **Compliant Tracking Approaches for HCP Programs** Given the restrictions, what can pharma marketers actually measure? Engagement with HCP-specific platforms (Doximity, journal sites) can be measured through the platform's own analytics, which are built for healthcare advertising compliance. Platform-side measurement doesn't involve dropping tracking pixels on the advertiser's own website. For website visits from HCP campaigns, pixel-based tracking on non-patient pages is generally lower risk. An HCP portal that's behind an NPI-verified login wall , where users have explicitly identified themselves as healthcare professionals , has a different risk profile than an open condition page. Document your compliance rationale carefully. UTM parameter tracking in URLs is compliant when the parameters themselves don't contain health information. A UTM like "utm_campaign=cardiology-hcp-spring26" creates attribution data without collecting individual user health information. For outcomes measurement, the holy grail in pharma marketing is closed-loop attribution: connecting HCP marketing touchpoints to prescription data. This is possible through partnerships with prescription data providers who can match HCP NPI numbers to prescribing behavior, with appropriate data use agreements in place. It's expensive to set up and requires significant data infrastructure, but it gives pharma marketing something most industries never get: actual sales attribution to marketing spend. The [healthcare analytics post](/insights/healthcare-analytics-patient-growth-2026) covers the data infrastructure side of healthcare measurement in more depth. For health system clients navigating the same HCP relationship questions from the provider side, the [healthcare content strategy post](/insights/healthcare-content-strategy-patients-algorithms) addresses how health systems should think about their own digital presence. Assessing where HCP marketing programs are creating compliance risk is part of the [marketing assessment](/tools/marketing-assessment) review we run for healthcare and pharma clients. **Key Takeaways** - HHS's December 2022 HIPAA tracking guidance (updated March 2024) made standard analytics and retargeting tools risky on patient-facing pages of HIPAA-covered entities , pharma legal teams responded by becoming broadly more conservative about tracking across all properties. - NPI-based targeting through platforms like Doceree enables compliant HCP programmatic advertising by matching physician NPI lists to device identities without relying on behavioral health data. - Doximity reaches over two million verified U.S. physicians with specialty and practice-setting targeting , the verification means you're not inadvertently reaching patients, which sidesteps the core HIPAA tracking concern. - Medical Legal Review (MLR) creates timing constraints on digital pharma content; modular content systems with pre-cleared claims components allow faster market response within the compliance framework. - Pixel-based website tracking on HCP-only, NPI-verified portals carries lower compliance risk than open patient-facing pages; document your compliance rationale regardless. - Closed-loop attribution connecting HCP marketing touchpoints to prescribing behavior is possible through prescription data partnerships , expensive but it gives pharma marketing sales attribution that most industries never achieve. --- # B2B Social Selling: Building Pipeline Without Being Annoying URL: https://brianroseman.com/insights/b2b-social-selling-pipeline-without-annoying Published: 2026-03-07 Most LinkedIn outreach gets ignored. These engagement strategies build relationships before the pitch. **Summary:** Most B2B social selling advice falls into one of two camps: do everything LinkedIn tells you to do, or ignore LinkedIn entirely and cold call. Both are wrong. There are specific behaviors on LinkedIn that build real pipeline, specific ones that annoy buyers into never responding to you, and a framework for telling the difference. This post covers what actually works from years of building pipeline and watching others try. **The Connection Request Pitch: Why It Fails Every Time** Let me describe something that happened in my LinkedIn inbox last Tuesday. A stranger sent me a connection request. I accepted. Within six minutes I had a message: "Hi Brian, I noticed you work in marketing. Our AI-powered platform helps CMOs like you 3x their pipeline in 90 days. Do you have 15 minutes this week?" This is what I call the drive-by pitch, and it is now so ubiquitous that it has made LinkedIn InMail response rates fall off a cliff. [LinkedIn's own research](https://business.linkedin.com/sales-solutions/social-selling) shows that buyers are five times more likely to engage with a salesperson if they have a warm connection or prior engagement. The drive-by does the opposite of creating warmth. It signals that the sender has no idea who I am, doesn't care, and is treating LinkedIn like a slightly better version of bulk email. The behavioral math is clear. When someone sends a connection request followed by an immediate pitch, I now treat every future connection request from someone I don't know as a potential pitch vehicle. So do most active LinkedIn users. The damage isn't just to that one interaction , it's to the channel's ambient trust level. **What Buyers Actually Respond To** The behavioral pattern that actually generates pipeline looks almost nothing like standard sales outreach. It's a research and engagement process that happens before any direct message. Before reaching out to a target account, the most effective social sellers spend two to four weeks engaging with content from people at that company. Not generic likes , real engagement. A thoughtful comment on the head of IT's post about vendor consolidation tells them three things: you exist, you have an informed perspective, and you're interested in the same problems they are. This sounds slow. It is. But compare the response rate of a cold InMail to someone who has seen your name in their comments twice before you reach out versus someone who has never seen your name before. The former regularly generates 30-40% response rates in my experience. The latter is industry standard at under 5%. The content consumption pattern matters too. [LinkedIn's sales data](https://business.linkedin.com/marketing-solutions/blog) shows buyers engage with 3-5 pieces of content from a vendor before agreeing to a conversation. If you have no content presence, you're not part of the consideration set before the conversation starts. **Building a Content Presence That Creates Inbound** The goal of a personal LinkedIn content strategy for B2B sales isn't to become an influencer. It's to appear credible and specific to the 2,000-5,000 people in your actual target market. That's a fundamentally different frame than "what should I post to get engagement." Posts engineered for engagement , polls, hot takes, inspirational stories , get comments. Posts that demonstrate specific expertise on problems your buyers face get DMs from the people you actually want to talk to. The content types that generate inbound for B2B sellers: Specific observations from client work. Not case studies (too polished, too promotional). Observations. "We noticed that manufacturing companies who track quote-to-close rate as their primary sales metric consistently miss the 30% of deals that die in the delivery estimate phase. Tracking it differently changed everything for a client last quarter." This is specific, interesting to people with that problem, and positions you as someone who has actually seen this. Opinions on industry trends with a point of view. The key word is "opinion." LinkedIn is full of content that summarizes what's happening in an industry. People who are willing to say "here's why I think X is wrong, and here's what the data shows" stand out. Take positions. You'll be wrong sometimes. That's fine , it makes you a person, not a content machine. Process breakdowns. If you've developed a specific way of solving a common problem, explaining that process in detail , with the specific steps, the decisions you make along the way, the things that go wrong , attracts people who recognize the problem and want to understand your approach before they ever talk to you. **LinkedIn SSI: Does It Matter?** LinkedIn's Social Selling Index (SSI) is a 0-100 score that measures your activity across four dimensions: professional brand, finding the right people, engaging with insights, and building relationships. LinkedIn positions it as a predictor of pipeline generation. The honest assessment: SSI correlates with good social selling behaviors but doesn't cause them. A high SSI by itself means nothing. But the behaviors that produce a high SSI , consistent posting, targeted connection building, active engagement with target accounts, using Sales Navigator search effectively , do produce results. It's a symptom of good behavior, not a cause. Where SSI scores matter practically: some organizations use SSI as a management metric for their SDR teams. If your company is doing this, you need to understand what drives the score. If yours doesn't, don't optimize for SSI , optimize for pipeline. **Using Sales Navigator Effectively** Most teams that have Sales Navigator use about 20% of its capability. The features they ignore are the ones with the highest ROI. Job change alerts are one of the most consistently valuable features. When a VP at a company you've been tracking switches to a new company, that's a moment of high receptiveness. They're building new vendor relationships, often have budget authority before their predecessor's commitments lock down the budget, and are motivated to show early wins. An outreach that references the role change ("Congrats on the new position , I know the first 90 days are intense. As you're thinking about X, we've helped a few companies in similar transitions with Y") converts at significantly higher rates than generic outreach. Account lists with activity tracking show you which of your target accounts are actively engaging with LinkedIn content in your topic area. A spike in engagement from an account on topics related to your solution is a buying signal worth acting on. Saved searches with email alerts mean you don't miss when new people match your ICP at target companies. A saved search for "VP of Operations OR Chief Operating Officer at manufacturing companies in Kansas City with 50-500 employees" running weekly tells you when someone new enters the market. **Attribution: How to Track Social Selling in Your CRM** The attribution problem with social selling is real. If I comment on someone's post, they read three of my articles, and then six weeks later they search for my company on Google and request a demo , what's the source in your CRM? Most CRM configurations capture the last touch (the Google search) and credit it as the source. This makes social selling look like it doesn't work, because the attribution is systematically going somewhere else. Better approaches for capturing social selling attribution: Train reps to log LinkedIn activity as activities against contact records in your CRM. When you comment on a post, send that person a connection request, or exchange DMs, these get logged. When they eventually convert, you have a timeline that shows social engagement predating the conversion. Add a "how did you first hear about us?" field to your discovery call script. The answer will often be "LinkedIn" or "I saw your content." This is qualitative but directionally useful. Track multi-touch attribution in your CRM rather than relying solely on first or last touch. Most CRMs support this; most teams don't configure it. The [marketing assessment tool](/tools/marketing-assessment) includes a section specifically on attribution configuration, which is where we typically find that social selling contribution is being systematically undercounted. **Personal Brand vs. Company Page: The Right Balance** Here is the uncomfortable truth about company page content on LinkedIn: it consistently underperforms personal content from individuals at that company. Not by a little. The organic reach difference is substantial. LinkedIn's algorithm explicitly favors person-to-person interactions over brand-to-person interactions. This doesn't mean the company page is useless. It's important for social proof , prospects will visit it to validate what they've seen from individuals. Your company page should have consistent, professional content that reinforces credibility. But if you're choosing where to invest limited content creation time, personal profiles of your actual sellers and executives generate better pipeline outcomes than the company page. The practical implication: if you're a solo consultant or small firm, your personal LinkedIn profile is your primary marketing channel. If you're leading a sales team, teaching each rep to build their own content presence will outperform any company page investment. For the [B2B content strategy](/insights/b2b-content-marketing-buyers-not-search) angle , particularly how your content presence maps to the buyer's research journey before they talk to you , the principle applies directly. Buyers who have read your insights before your outreach have a fundamentally different starting point for a first conversation. And if your [technology PPC cost per lead](/insights/technology-ppc-cost-per-lead-climbing) is rising (which it is for most B2B tech categories), social selling's organic pipeline generation becomes an increasingly attractive complement to paid channels. **The Sequence That Actually Works** Based on what I've seen generate real meetings rather than polite declines, here is the pattern: Week 1-2: Identify 20-30 specific people at target accounts. Follow them. Start engaging genuinely with their content , comments that add something specific, not just "Great post!" Week 3-4: Share content relevant to the problems you know their industry faces. Tag industry connections (not the targets themselves yet) when relevant. Week 5-6: Send a connection request with a specific, non-pitch note. "I've been following your posts about supply chain visibility , your take on X was interesting. Connecting because I'm working in a similar space and find the conversation here valuable." No ask. No pitch. Week 7+: Once connected and with a few exchanges of engagement, you now have enough context for a message that's actually personalized. "Given what you've shared about [specific challenge], I thought you'd find [specific resource] useful. Happy to talk through how we've seen others approach it." This isn't a pitch. It's a useful message. The sequence sounds long because it is. But the meetings it generates are warm, qualified, and result in shorter sales cycles. The drive-by pitch generates rejection at scale and wastes everyone's time. **Key Takeaways** - Connection request followed by immediate pitch is the single fastest way to ensure your target never responds to you again , response rates below 5% reflect how common it has become. - Buyers typically engage with 3-5 pieces of content before agreeing to a conversation; if you have no content presence, you're invisible in the consideration phase. - LinkedIn SSI correlates with good social selling behaviors but doesn't cause them , optimize for pipeline, not the score. - Job change alerts in Sales Navigator identify moments of high buyer receptiveness , new leaders at target accounts are building vendor relationships before budget commitments lock in. - Social selling attribution is systematically undercounted in most CRMs because last-touch attribution credits paid search or direct, not the LinkedIn engagement that preceded it; fix this with CRM activity logging and multi-touch attribution. - Personal profiles consistently outperform company pages in organic LinkedIn reach; invest content creation time in individual sellers, not the brand page. --- # E-commerce Conversion Rate Optimization: The Tests That Actually Matter URL: https://brianroseman.com/insights/ecommerce-cro-tests-that-matter Published: 2026-03-05 Most e-commerce CRO programs test button colors and headlines. The tests that actually grow revenue target checkout friction, average order value, and segment-specific conversion barriers. Here's where to focus. "**Summary:** Most e-commerce CRO programs are running the wrong tests. Button color tests and headline variations look productive but rarely move revenue. The tests that actually matter target checkout friction, average order value, and segment-specific conversion barriers. **Most CRO Is Testing Theater** You've probably run an A/B test that showed a statistically significant lift, rolled out the winner, then watched your revenue stay flat for the next quarter. That happens because most e-commerce CRO programs optimize for the conversion rate metric without asking what's causing the conversion. They test surface elements, headlines, button colors, hero images, that have some effect on CTR but no structural effect on purchase intent. Real CRO is about removing friction between intent and purchase. That happens in checkout, on product pages where confidence breaks down, and at the moments when your mobile user gives up and switches to desktop to ""buy it later"" (and then doesn't). Here's where the money is. **Checkout: The 47% Problem** [Baymard Institute's research on cart abandonment](https://baymard.com/lists/cart-abandonment-rate), the most comprehensive dataset on e-commerce checkout behavior, puts the average cart abandonment rate at 70.19% across industries. For e-commerce, it's lower but still brutal: typically 55-65% of users who add to cart don't complete the purchase. The reasons Baymard documents are consistently the same across studies: - **Forced account creation**: 26% of users who abandon cite this as the reason. Guest checkout alone typically lifts checkout completion by 6-8%. - **Unexpected costs at checkout**: Shipping surprises kill 48% of abandoning shoppers. Show shipping costs on product pages, not at checkout. - **Complex or long checkout process**: Every additional form field is a drop-off point. The Baymard benchmark for a well-optimized checkout is 7-8 fields (name, email, shipping address, payment). The average e-commerce site collects 14.88 fields. - **Payment method limitations**: If you don't offer the payment method a user prefers, you lose them. Adding PayPal typically lifts checkout completion by 3-5%. Apple Pay and Google Pay on mobile can lift by 2-4%. Before testing anything else, audit your checkout against these criteria. Fixing structural checkout problems is worth more than any headline test you've ever run. **Average Order Value: The Metric That Outperforms Conversion Rate** A 1% improvement in conversion rate on a $80 AOV store adds less revenue than a 10% improvement in AOV at the same conversion rate. Do the math: - 10,000 sessions × 2.5% conversion × $80 AOV = $20,000 revenue - 10,000 sessions × 2.5% conversion × $88 AOV (+10%) = $22,000 revenue (+$2,000) - 10,000 sessions × 2.75% conversion (+10%) × $80 AOV = $22,000 revenue (+$2,000) Same math, same lift. But AOV improvements are often structurally easier than conversion rate improvements because you're targeting people who've already decided to buy. AOV tests worth running: **Post-add-to-cart upsells.** After someone clicks ""Add to Cart,"" show a single targeted upsell before checkout. Keep it relevant (not random), priced at 20-30% of the cart value, and easy to decline. When done right, 15-25% of users accept the upsell. **Free shipping thresholds.** If your free shipping threshold is $50 and the average order is $47, you're leaving money on the table. Test messaging like ""Add $3 to qualify for free shipping"" at cart. This consistently lifts AOV by 10-20% when the gap is small. **Bundling and quantity breaks.** ""Buy 2, get 20% off"" tests well in categories with repeat purchase intent, consumables, apparel basics, supplements. Test the discount level and messaging alongside each other. **Product Pages: Where Confidence Breaks Down** Most failed purchases are decided on the product page, not at checkout. Users who abandon at checkout typically had a pre-existing doubt they were hoping would resolve. It didn't. Product page tests that actually matter: **Social proof quality over quantity.** 500 generic reviews don't move the needle as much as 50 highly specific, detailed reviews with photos. Test review display format: summary vs. individual reviews, filtered by use case, photo vs. no photo. **Return policy prominence.** [Narvar's consumer research](https://corp.narvar.com/resources/) consistently shows that clear, generous return policies are the second most influential factor in first-time purchase decisions (after price). Test the position and prominence of your return policy on product pages, not buried in footer links. **Product photography.** Lifestyle context photos that show the product in use typically outperform pure product photos on white backgrounds. Test the lead image, not secondary images. **Size/variant selection.** For apparel and footwear, the moment a user hits ""out of stock"" in their size is usually the end of that session. Test how you communicate availability and handle out-of-stock variants, showing them vs. hiding them, notify-me functionality, similar product suggestions. **Mobile vs. Desktop: Separate Experiences, Separate Tests** If you're not segmenting your test results by device, you're misreading your data. Mobile and desktop users have fundamentally different conversion paths. Mobile users typically research and browse; desktop users complete purchases. The [Salesforce Shopping Index](https://www.salesforce.com/news/stories/q4-2024-shopping-index/) reports that mobile accounts for 68% of traffic but only 44% of orders, a conversion rate roughly half of desktop. This means: - A test that shows neutral results overall might show a strong win on desktop and a loss on mobile, averaging to zero - Mobile optimizations (tap target size, form field reduction, simplified navigation) have different optimal solutions than desktop - Your checkout abandonment rate on mobile is probably catastrophic and needs separate attention Run mobile and desktop as separate segments in your testing tool. What wins on one device often loses on the other. **Test Velocity: The Hidden Variable** The biggest predictor of successful CRO programs isn't the quality of individual tests, it's how many tests a team can run per month. Most e-commerce teams run 1-2 tests per month and declare tests complete at 95% statistical significance. [Evan Miller's A/B test calculator](https://www.evanmiller.org/ab-testing/sample-mover.html), and every good statistics textbook, will tell you that this approach produces high rates of false positives. What high-velocity testing programs do differently: - **Hypothesis documentation before testing.** Every test has a documented hypothesis (""We believe X change will increase Y metric because of Z insight"") that gets reviewed regardless of outcome. - **Minimum detectable effect defined upfront.** You need to know what lift is meaningful before you start, not after. - **Bayesian approaches over frequentist.** Tools like VWO and Optimizely now offer Bayesian testing modes that let you make decisions with less traffic and avoid peeking problems. - **Test infrastructure, not just ideas.** The bottleneck in most CRO programs is implementation time, not test ideas. Streamline the path from hypothesis to live test. Use the [Investment ROI Comparison tool](/tools/roi-calculator) to model what improved test velocity does to your annual revenue, even a shift from 1 to 3 tests per month, with the same win rate, compounds dramatically. **What Doesn't Work (And Why Teams Keep Doing It)** **The above-the-fold obsession.** Everyone wants to optimize the hero section. It's visible. It's easy to understand. It rarely drives meaningful lift because most conversion barriers exist much further down the funnel. **Social proof spam.** Adding notification popups (""Sarah from Denver just bought this!"") and countdown timers to a product that doesn't warrant urgency trains users to ignore your signals and creates distrust. These tactics work in specific contexts (genuine limited availability, real-time stock levels) and backfire in most others. **Micro-copy A/B testing.** ""Buy Now"" vs. ""Shop Now"" vs. ""Add to Bag"", these micro-copy tests are the testing equivalent of rearranging deck chairs. CTR might shift by 1-2%, but purchase completion doesn't. **Testing without segmentation.** A 3% conversion rate across your full user base is an average of wildly different behaviors: returning customers converting at 8%, new organic visitors converting at 1.5%, paid traffic converting at 4%. Tests designed for the average user optimize for nobody in particular. For a structured framework on building a CRO program with real measurement infrastructure, the [CRO calculator](/tools/cro-calculator) is a good starting point for modeling where your opportunity lies before committing to a testing roadmap. **Key Takeaways** - Cart abandonment averages 70% across e-commerce; fixing forced account creation and unexpected shipping costs at checkout beats any headline test - AOV improvement and conversion rate improvement produce equivalent revenue at equivalent percentage lifts, AOV is often easier to move for established stores - Segment every test by device; mobile and desktop conversion rates differ by 2:1, and aggregated results hide real winners and losers - High-velocity testing (3+ tests per month) compounds over time, the bottleneck is usually implementation speed, not test ideas - Product page confidence factors (return policy, social proof quality, photo format) are where first-time purchase decisions are made and lost" --- # Technology PPC: Why Your Cost Per Lead Keeps Climbing URL: https://brianroseman.com/insights/technology-ppc-cost-per-lead-climbing Published: 2026-03-05 B2B technology PPC costs climbed 40-60% between 2022 and 2025. Most marketing teams blame Google. The real problem is structural — intent mismatch, match type decay, and competitor spiraling — and fixable if you know where to look. "**Summary:** B2B technology PPC costs climbed 40-60% between 2022 and 2025. Most marketing teams blame Google. The real problem is structural, and fixable if you know where to look. **The Numbers Are Bad. Here's Why They're Getting Worse.** The average cost per click for B2B technology keywords hit $8.67 in Q3 2025, up from $5.40 in 2022, [according to WordStream's annual PPC benchmarks](https://www.wordstream.com/blog/ws/2016/02/29/industry-benchmarks). For enterprise software categories, ERP, CRM, cybersecurity, you're looking at $15-30 CPC with no ceiling in sight. If your CPL has been climbing 15-20% annually while your conversion rates stay flat, you're experiencing something more than cost inflation. You've got a structural problem. There are four root causes, and most marketing teams are dealing with at least two of them simultaneously. **The Intent Mismatch Problem** Most B2B tech PPC campaigns are reaching people who aren't buyers. They're reaching researchers, students, competitors running competitive intelligence, and your own employees testing your ads. Here's the tell: look at your search term reports. If more than 30% of your spend is going to queries with no commercial intent, ""what is CRM software,"" ""how does ERP work,"" ""difference between X and Y"", you're paying to educate people who will never buy. The fix isn't just adding negatives. It's restructuring your campaign architecture around intent tiers: **Tier 1 (Highest intent):** ""[Product] pricing,"" ""[Product] vs [Competitor],"" ""best [category] for [use case]"", these convert. They should have the highest bids, tightest ad groups, dedicated landing pages. **Tier 2 (Moderate intent):** ""[Category] software,"" ""top [category] tools"", worth bidding, but with tighter conversion requirements before scaling. **Tier 3 (Low intent):** Informational queries, don't bid on these, or bid only in remarketing contexts where you've already qualified the audience. Most technology PPC campaigns have their budgets inverted, spending the most on Tier 3 keywords because they have higher volume and lower CPCs. They're cheap because they don't convert. **Match Type Decay** Google killed exact match in any meaningful sense around 2021. What's left of ""exact match"" includes ""close variants"" that Google determines are semantically similar, which in practice means your exact match keywords are behaving like phrase match keywords from 2018. This matters for B2B tech because intent signals in search queries are highly specific. ""Enterprise CRM software"" and ""CRM software"" are very different queries with very different intent. Google treats them as close variants. The practical result: your carefully constructed negative keyword lists are leaking. Keywords you explicitly excluded are getting served because a ""close variant"" triggered your ad. There's no perfect fix, but there are mitigation strategies: - Audit your search term report weekly, not monthly. Catch the leaks before they compound. - Build tight single-keyword ad groups (SKAGs) for your highest-intent terms and accept the management overhead. - Move to value-based bidding rather than target CPA, you're signaling to Google which conversions matter, not just that any conversion matters. - Test Performance Max campaigns carefully. For most B2B tech advertisers, Performance Max increases volume but tanks lead quality. Know your MQL-to-opportunity rate, not just your CPL. **The Competitor Spiral** B2B technology is a category where your competitors are watching your ads as closely as you're watching theirs. When one company raises bids to capture more impression share, others follow defensively. It's not strategic, it's reflex. The result is everyone in the category paying more for the same traffic, with no one gaining sustained advantage. You can break out of the spiral in two ways: **Option 1: Out-quality them.** Better quality scores reduce your CPC by 10-50% even at the same bid level. Quality score is driven by expected CTR, ad relevance, and landing page experience. A click-through rate 20% higher than category average means you're paying less for the same position. Investing in better ad copy and landing pages beats raising bids every time. **Option 2: Go where they're not.** For B2B technology, LinkedIn often has less auction competition than Google for the same audience. LinkedIn CPCs are higher ($6-15 vs. $3-8 for Google), but when you're reaching CFOs and VPs of IT by title, not by search query, the qualification rate is often better. [LinkedIn's B2B marketing research](https://business.linkedin.com/marketing-solutions/research) consistently shows purchase intent among professional audiences outperforming display alternatives. **The Audience Saturation Problem** Here's a number that should concern you: the average B2B technology buyer sees 5-7 ads from competing vendors before making a shortlist decision. Your remarketing audiences are the same people your competitors are remarketing to. Your lookalike audiences are built from the same LinkedIn job titles everyone else is targeting. You're paying to reach people who are already exhausted by your category's advertising. This doesn't mean remarketing doesn't work. It means you need to think about audience sequencing and messaging fatigue: - Set frequency caps on display and LinkedIn campaigns. Eight impressions per week from the same brand is a trust destroyer. - Segment your remarketing audiences by engagement depth. Someone who read three blog posts needs different messaging than someone who hit your pricing page. - Create exclusion audiences for people who've converted, been in an active sales conversation, or already rejected your offer. Account-based marketing (ABM) is the structural answer to audience saturation. Instead of reaching everyone who might fit your ICP, you identify specific companies you want to win and concentrate spend there. [Demandbase's 2025 State of ABM Report](https://www.demandbase.com/resources/reports/) found that companies with mature ABM programs reduced CPL by 34% while increasing pipeline quality. **What Actually Lowers CPL (And What Doesn't)** Things that don't lower CPL in a meaningful way: - Switching to automated bidding strategies without enough conversion data (you need 30+ conversions per month minimum) - Adjusting ad copy without changing landing page parity - Adding more keywords to increase impression share - Running branded campaigns against your own traffic to inflate conversion numbers Things that do work: **Raise conversion rate before raising budget.** Most B2B tech landing pages convert at 2-5%. The median for high-performing campaigns is 8-12%. Closing that gap is worth more than any bid adjustment. Run the [CRO Calculator](/tools/cro-calculator) to see what a conversion rate improvement from 3% to 5% does to your CPL at current spend. **Invest in post-click attribution.** If you're measuring CPL at the form-fill level, you're optimizing for the wrong thing. Which keywords generate pipeline? Which ad groups produce closed revenue? You need GA4 and CRM integration to answer this. [My guide to fractional analytics services](/services/fractional-analytics) covers exactly this setup. **Build a content quality score feedback loop.** The keywords driving your lowest cost-per-opportunity (not cost-per-lead) should get more budget. This requires connecting your CRM win data to your ad platform data. It's not complicated to set up, but almost nobody does it. **The LinkedIn vs. Google Question** For B2B technology, the honest answer is: both, but for different jobs. Google is for in-market demand capture. Someone searching ""HR software comparison"" is actively shopping. Meet them there. LinkedIn is for demand generation. You're reaching a VP of HR who isn't actively searching, but matches your ICP and is receptive to category education. You're building the consideration set before the purchase trigger fires. The mistake is treating them as interchangeable. Running Google ad copy on LinkedIn doesn't work. Educational content that works on LinkedIn is too soft for someone searching with buying intent on Google. Separate your measurement too. Google campaigns should be measured on pipeline velocity. LinkedIn campaigns should be measured on brand recall, engagement quality, and influenced pipeline, not CPL, because the LinkedIn CPL will always look worse on a direct response basis. If you're allocating 100% of your B2B tech ad spend to Google search, you're leaving demand generation on the table and making your Google campaigns work too hard. The general framework: 60-70% of budget on Google for in-market capture, 30-40% on LinkedIn for pipeline building, adjusted based on your category's search volume. **Key Takeaways** - The average B2B tech CPC hit $8.67 in Q3 2025, a 60% increase from 2022, driven by intent mismatch, match type decay, and competitor spiraling - Restructure campaigns by intent tier: put highest bids on pricing and comparison keywords, not high-volume informational queries - Quality score improvement reduces CPC by 10-50% without raising bids, better investment than competing in bid auctions - ABM programs reduce CPL by 34% while improving pipeline quality by concentrating spend on specific target accounts - Separate Google (in-market capture) and LinkedIn (demand generation) by job and measurement framework, don't use the same metrics for both" --- # Financial Services Lead Scoring: Models That Predict Lifetime Value URL: https://brianroseman.com/insights/financial-services-lead-scoring-lifetime-value Published: 2026-02-28 Traditional lead scores miss high-value prospects. These predictive models identify your best customers early. **Summary:** Standard lead scoring in financial services , assigning points for job title, company size, and email opens , is a blunt instrument that tells you who might be interested, not who is actually likely to become a profitable long-term customer. This post covers how to build predictive lead scoring models that connect intake data to downstream lifetime value, with specific model inputs, compliance guardrails, and implementation steps for mortgage, wealth management, and insurance. **Why Standard Lead Scoring Falls Apart in Financial Services** In most industries, a high lead score means "this person is ready to buy." In financial services, that correlation is weaker than it looks. A mortgage lead who matches your ideal buyer profile , 750+ credit score, stable income, appropriate down payment , might close a $400,000 loan and never do business with you again. Another lead, with a messier financial picture and a longer timeline, might become a wealth management client worth $180,000 in fees over 20 years. The problem with most MQL/SQL frameworks is that they score for conversion probability, not value. These are related but meaningfully different things. A lead scoring model that predicts "this person will fill out an application" optimizes for volume. A model that predicts "this person represents significant long-term value" optimizes for the business outcomes that actually matter. The gap between conversion probability and lifetime value is largest in financial services. Mortgage origination, for instance, has relatively low switching costs for consumers. Wealth management has very high switching costs. Insurance is somewhere in between. The LTV profile of your ideal customer looks radically different depending on which product they're entering through. **Data Sources That Make Predictive LTV Possible** Building a model that predicts lifetime value requires connecting early-stage lead data to downstream revenue data. That connection is the hard part , and it's where most financial services firms get stuck. The data sources that feed a useful LTV prediction model: CRM data tells you what you know about the lead at intake: product interest, demographic information, source, initial qualification signals. This is what most lead scoring models use exclusively. It's necessary but insufficient. Behavioral data from your website and content consumption adds intent signals that CRM forms don't capture. Someone who visited your retirement planning calculator six times in three weeks, downloaded your guide to rolling over a 401(k), and then requested an advisor call has a very different profile than someone who clicked a paid search ad and filled out the same form. Their behavioral trail is a signal about their engagement level and, often, their financial complexity. CRM history for converted leads, tracked over 12-24 months, gives you the ground truth for what "high LTV" actually looks like. You're looking for patterns in the intake data of leads who became your best long-term clients. What did they have in common at the point of first contact? That backward analysis is the foundation of a predictive model. Third-party enrichment data adds context that your intake forms can't collect without friction. Firmographic data for B2B financial services (business revenue, employee count, industry), modeled income estimates for consumer financial products, and home ownership status are all available through data providers. The compliance review question , covered below , is which of these you can actually use in scoring. **Model Inputs for Specific Financial Products** **Mortgage:** The inputs that best predict both near-term conversion and long-term value are different from what you might expect. Credit score and income are obvious. Less obvious: the stage of the real estate search (actively shopping for a home vs. "just curious"), the referral source (agent referrals have significantly higher conversion rates than paid search), and whether the person has owned before. First-time homebuyers have more friction in the process but often become long-term relationships if the experience is positive. Repeat buyers are faster to close but may already have a lender relationship. **Wealth Management:** Complexity signals are the most important input. A lead who mentions specific tax concerns, asks about estate planning alongside investment management, or has multiple account types they want to consolidate is signaling complexity that correlates with higher assets and longer relationships. Source matters enormously: a referral from a CPA or estate attorney has consistently higher LTV in every wealth management firm I've worked with than an inbound lead from organic search for "financial advisor near me." **Insurance:** Product breadth is the strongest LTV predictor. A prospect entering through auto insurance who also owns a home, has dependent children, and runs a side business represents a $15,000+ annual premium opportunity across bundled products. One who's shopping on price for minimum coverage represents a fraction of that. The intake questions that identify multi-product potential , homeownership, business ownership, umbrella coverage awareness , are more predictive than most agents realize. **Compliance Considerations: FCRA and Fair Lending** This is the area where financial services lead scoring requires a conversation with legal and compliance that most marketing teams try to avoid. The Fair Credit Reporting Act (FCRA) and fair lending regulations under the Equal Credit Opportunity Act (ECOA) and the Fair Housing Act impose real constraints on what data you can use in automated scoring systems that affect credit or lending decisions. The [CFPB's regulatory guidance](https://www.consumerfinance.gov/data-research/) is explicit: if your model uses proxies for protected characteristics , ZIP code as a proxy for race, for instance , you create fair lending exposure. This is true even if the model's predictive value is high and the discriminatory intent is absent. What matters legally is discriminatory impact, not intent. The guardrails for compliant scoring: Do not use race, color, national origin, sex, religion, or familial status as inputs , directly or as proxies. This is clear. Marital status cannot be used in lead scoring that influences credit decisions. Geographic-based inputs require careful review. ZIP code-level income averages or home price data are problematic if they serve as proxies for protected characteristics. Age is permitted in some contexts but prohibited in others , ECOA prohibits using age in credit scoring models for certain products. The practical resolution for most financial services firms: build a two-layer model. The first layer uses only permitted inputs for any scoring that influences credit or lending offers. The second layer uses richer data , behavioral signals, third-party enrichment, engagement depth , for marketing prioritization and content personalization, which operate under different regulatory standards than credit decisions. [Salesforce's lead management resources](https://www.salesforce.com/resources/articles/lead-scoring/) provide good context on the CRM implementation side of this, though the compliance nuances are specific to financial services and require your own legal review. **GA4 Behavioral Signals as Scoring Inputs** GA4's event tracking gives you behavioral data that can feed your CRM scoring system, provided you've set up the integration correctly. The key setup requirements: User-level tracking must be linked between GA4 and your CRM. The standard approach is using GA4's client ID or user ID in combination with form submissions , when someone submits a form, capture their GA4 client ID and pass it to your CRM along with the form data. Custom events should be set up for high-intent behaviors: calculator use, guide downloads, advisor locator searches, pricing page visits. These are the signals that indicate financial intent beyond a generic pageview. GA4 Audiences built on these behavioral signals can be synced to Google Ads for paid campaign targeting , letting you bid more aggressively for users who match your highest-LTV behavioral profile before they convert. This is legal and doesn't involve credit data. **Connecting Lead Score to Downstream Revenue in BigQuery** For financial services firms with significant data infrastructure, BigQuery-based analysis of lead cohorts is where predictive LTV modeling gets real. The approach: Export your CRM data and GA4 behavioral data to BigQuery. Google's native BigQuery connectors for both platforms make this relatively straightforward. Build cohort tables that group leads by their intake characteristics , source, product interest, initial engagement score , and track their downstream revenue contribution at 6, 12, and 24-month intervals. Run regression analysis on the cohort data to identify which intake variables most strongly predict 24-month revenue. This is where you discover whether the "high engagement" leads you're scoring highly actually become better customers than lower-engagement leads who match the demographic profile. The output of this analysis is a model that assigns each new lead a predicted 24-month value rather than a simple MQL/SQL status. That prediction changes how marketing budget gets allocated , you invest more in acquiring and nurturing leads that look like your highest-LTV cohorts, even if their immediate conversion probability is lower. For modeling the downstream impact of improving lead quality on revenue, the [LTV calculator](/tools/ltv-calculator) is a useful tool for running quick scenarios before building out the full BigQuery model. The [marketing assessment](/tools/marketing-assessment) is useful for understanding your current lead generation mix and identifying where the highest-value opportunities are. The broader lead qualification challenge connects to the B2B content marketing post on [writing for buyers, not search engines](/insights/b2b-content-marketing-buyers-not-search), which covers how content strategy affects lead quality , not just lead volume. For financial services specifically, the relationship between analytics capability and marketing performance is a theme the [fractional analytics service](/services/fractional-analytics) addresses directly. **Practical Implementation Steps** Building a predictive LTV scoring model from scratch is a 3-6 month project for most financial services firms. A more practical starting point is incremental improvement: Start by appending historical revenue data to your existing CRM leads. Even a simple 12-month revenue flag (did this lead become a client worth more than $X?) gives you an immediate lens for evaluating whether your current scoring criteria actually predict value. Run the analysis on a 24-month lookback of converted leads. Compare intake characteristics of top-quartile LTV clients to bottom-quartile clients. The patterns you find will be specific to your firm and more reliable than any generic scoring model. Use those patterns to revise your intake questions. If your analysis shows that referral source is the strongest predictor of LTV, build a better referral tracking system. If certain behavioral signals predict high-value clients, create more content and tools that generate those behaviors. Introduce behavioral scoring as a secondary layer that supplements , but doesn't replace , your conversion probability scoring. Track whether this improves the quality of leads that get prioritized by sales. Measure the model's performance over time. Lead scoring models decay as markets change and customer behavior shifts. A model built on 2022 data that's still running unchanged in 2026 is almost certainly underperforming. Build in a quarterly review process. **Key Takeaways** - Standard lead scoring optimizes for conversion probability. Predictive LTV scoring optimizes for long-term revenue , and in financial services, these two objectives often point to different leads. - CRM and form data alone are insufficient for LTV prediction; behavioral signals, engagement depth, and product complexity indicators are often more predictive of long-term value. - Referral source is consistently one of the highest-value signals in financial services: CPA and attorney referrals for wealth management, agent referrals for mortgage, and multi-product signals for insurance all predict higher LTV. - FCRA and ECOA compliance requires building a two-layer model , one for credit-related scoring using only permitted inputs, and a separate layer for marketing prioritization using richer behavioral data. - BigQuery cohort analysis connecting intake data to 12-24 month revenue is the most reliable way to build and validate a predictive LTV model for your specific firm. - Lead scoring models decay; build a quarterly review process to update the model as market conditions and customer behavior change. --- # Healthcare Content Strategy: Writing for Patients AND Algorithms URL: https://brianroseman.com/insights/healthcare-content-strategy-patients-algorithms Published: 2026-02-25 Medical content needs accuracy and readability. Here is how to balance clinical credibility with SEO. **Summary:** Healthcare organizations face a content challenge that most industries don't: writing accurately for both a seventh-grade reading level AND the technical requirements of search algorithms that reward expertise and authority. Get it wrong in one direction and patients can't understand your content. Get it wrong in the other and it doesn't rank. This post covers the specific frameworks, schema markup, E-E-A-T signals, and tracking approaches that help health systems solve both problems at once. **The Reading Level Problem Nobody Talks About** The [CDC's health literacy guidance](https://www.cdc.gov/healthliteracy/) puts the average U.S. adult reading level at around 8th grade, with about 36% of adults having basic or below-basic health literacy. For health systems, the practical implication is that clinical language in patient-facing content creates real barriers. Not aesthetic ones. Real ones: patients who don't understand their diagnosis, don't follow treatment instructions, don't show up for follow-up appointments. I worked with a regional health system during my agency years on a content audit that found something uncomfortable: the average Flesch-Kincaid reading level across their patient education content was 14.2 , roughly a second-year college level. Their content team was staffed entirely by clinical writers who were excellent at accuracy and terrible at plain language. The problem wasn't that they were writing for clinicians. The problem was that they had no framework for writing for patients. The reading level tension is real. If you write at a 7th grade level about a complex condition like atrial fibrillation, cardiologists on your medical advisory board will flag the oversimplification. If you write at a 12th grade level, a significant portion of your actual patient population can't use the content. The resolution isn't to pick one level , it's to structure content so it works for both audiences simultaneously. **Layered Content Architecture** The framework that actually works for healthcare content is what I call layered architecture: a plain-language summary at the top of every substantive piece, followed by progressively more detailed content for readers who want depth. The summary layer , typically three to five sentences , answers the four questions most patients have: What is this condition? What causes it? What are the symptoms? What should I do if I have them? It uses no jargon. No clinical terms without immediate plain-language definitions. No statistics without context. The detail layer that follows can go deeper. It's where you explain mechanisms, cite clinical evidence, and address the nuanced questions that engaged patients and their family members want to answer. This is also where you include the signals that search engines and quality raters look for when evaluating expertise. This architecture has a secondary benefit: it naturally creates the kind of structured content that tends to get picked up in Google's AI Overviews and featured snippets. A clear plain-language answer at the top of a page is exactly what those systems are designed to extract. **E-E-A-T Signals for YMYL Health Content** Google's Quality Raters Guidelines classify health content as YMYL , "Your Money or Your Life" , content that can directly affect a person's health, safety, or financial wellbeing. The guidelines explicitly require the highest level of expertise, authoritativeness, and trustworthiness for YMYL pages. This is where the concept of E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) is most consequential. For a health system, E-E-A-T signals aren't abstract. They're specific and actionable: Every substantive clinical article needs a named, credentialed author. Not "Medical Review Team." A specific physician or clinician, with their credentials, specialty, and board certifications visible on the page. A "reviewed by Dr. Sarah Chen, MD, Board-Certified Cardiologist, University of Kansas Health System" byline is dramatically different from a generic medical team attribution. Last-reviewed dates matter. Medical information changes. A comprehensive article about COVID-19 treatment that was last reviewed in 2020 is not just outdated , it actively signals unreliability to quality raters and, increasingly, to the algorithms that incorporate their feedback. A visible "Last medically reviewed: March 2025" notation is a trust signal. Medical credentials should be linked to verifiable sources where possible. If you reference that an author is board-certified, linking to the American Board of Medical Specialties or a state medical board lookup isn't required, but it adds a layer of verifiability that separates genuine expertise from claimed expertise. [Google's guidance on creating helpful content](https://developers.google.com/search/docs/fundamentals/creating-helpful-content) is worth reading carefully. The emphasis on demonstrating first-hand experience is particularly interesting for healthcare , patient story content, when properly structured and consented, actually scores well on the experience dimension. **Schema Markup for Healthcare Content** The [Article structured data type](https://developers.google.com/search/docs/appearance/structured-data/article) is the standard starting point for health content. But the [MedicalWebPage schema](https://schema.org/MedicalWebPage) from schema.org provides much richer markup specifically designed for medical content. MedicalWebPage lets you specify: The medical specialty relevant to the content (using the MedicalSpecialty enumeration). The medical audience for the content (patient vs. clinician) , which maps directly to the reading level discussion above. The medicalAudience property (Patient, Clinician, etc.). The lastReviewed date , critical for establishing currency. A reviewedBy property that links to the reviewing physician's structured data. Implementing this correctly requires coordination between your content team and web development. The JSON-LD block lives in the page head and needs to be populated dynamically if you're building content at scale. For health systems with hundreds or thousands of condition pages, this typically means integrating the schema generation into your CMS at the template level. The MedicalWebPage schema doesn't directly guarantee better rankings , no schema markup does. What it does is give search engines cleaner signals about what your content is, who it's for, and how authoritative the source is. **HIPAA Considerations for Analytics** This is the area that creates the most operational complexity for healthcare marketers. The [HHS December 2022 guidance](https://www.hhs.gov/hipaa/for-professionals/privacy/guidance/hipaa-online-tracking/index.html) clarified that web tracking technologies , including Google Analytics and Meta Pixel , can constitute violations of HIPAA when they collect individually identifiable health information (IIHI) on hospital or health system websites. The specific problem: if a user visits a page on your hospital website like "/orthopedics/knee-replacement-surgery" and your analytics or ad tracking tools collect their IP address in combination with that page URL, you've potentially transmitted IIHI to a third party without authorization. The practical response for most health systems has been: Audit your current tracking stack for which tools receive URL-level data on clinical pages. Google Analytics 4 with IP anonymization and restricted data sharing settings is generally safer than Universal Analytics configurations, but the URL-level data collection issue remains. Separate your marketing tracking from your clinical content pages where possible. Your /about, /careers, and /contact pages don't carry the same HIPAA risk as /conditions, /treatments, and /symptoms pages. Work with your legal and compliance team before making any tracking decisions. This is not an area where you want to make unilateral calls. GA4 has added healthcare-specific configurations, and Google has signed BAAs with health systems in some contexts. But the situation is evolving, and the HHS guidance was explicit that typical analytics implementations on clinical pages create real compliance risk. The marketing assessment we offer as a [free tool](/tools/marketing-assessment) includes a review of analytics setup, which is often the first place we find exposure. **Plain Language Without Losing Accuracy** The craft challenge in healthcare content is rewriting clinical language without losing medical accuracy. A few specific techniques: Active voice eliminates the passive constructions that make clinical writing hard to read. "The medication reduces inflammation" is clearer than "Inflammation is reduced by the medication." Plain-language alternatives for common clinical terms. "High blood pressure" not "hypertension" , or, if you use both, "hypertension (high blood pressure)" with the plain term following immediately. "Nerve pain" not "neuropathy." "Imaging test" or "scan" before specifying MRI or CT. Use numbers carefully. "25% of patients" is clearer than "one in four patients" for some readers and less clear for others. Both are better than "a significant proportion." Sidebars and callout boxes for clinical depth. Rather than interrupting the plain-language narrative with a paragraph of clinical explanation, put that detail in a visually distinct section that engaged readers can access without breaking the flow for everyone else. **Measuring Healthcare Content Performance** Standard e-commerce content metrics don't translate directly to healthcare. You're not trying to attribute a $150 revenue event to a content piece about knee arthritis. You're trying to measure whether content is helping people find care, understand their conditions, and navigate to appointment scheduling. The metrics that actually tell you something: Organic search visibility for condition and treatment keywords, tracked over time. Are you ranking for the searches your target patient population is doing? Scroll depth and time on page, segmented by traffic source. Content that people read all the way through and then immediately search for appointment booking is performing well, even if session duration looks long. Conversion rate to appointment scheduling from condition and treatment pages. This is the closest thing to a bottom-line metric. It requires your analytics tracking to connect content page visits to downstream scheduling form completions or phone calls. Content-assisted conversions: appointments where the patient visited your content pages at some point in their journey, even if the last touch was a branded search or direct visit. This gives you a more complete picture of content's role in the patient journey. The connection between content strategy and patient acquisition is explored in more depth in the [healthcare analytics post](/insights/healthcare-analytics-patient-growth-2026), which covers the data infrastructure side of measuring patient growth. For an overall assessment of how your healthcare organization's digital marketing is performing across channels, the [marketing assessment](/tools/marketing-assessment) is a good starting point. **Key Takeaways** - Average U.S. adult reading comprehension sits around 8th grade, making 12th-grade clinical writing a patient access issue, not just an SEO problem. - Layered content architecture , plain-language summary up top, clinical depth below , serves both audiences and naturally creates the structured format AI Overviews extract. - E-E-A-T signals for health content require named clinical authors with visible credentials, last-reviewed dates, and verifiable expertise , not generic "Medical Team" attributions. - MedicalWebPage schema from schema.org provides richer markup than standard Article schema for health content, including medical specialty, audience type, and review date. - HHS December 2022 HIPAA tracking guidance requires health systems to audit their analytics stacks; URL-level data collection on clinical pages combined with user identifiers creates real compliance exposure. - Measure healthcare content performance through organic visibility for condition keywords, scroll depth, and conversion to appointment scheduling , not standard e-commerce metrics. --- # Retail Personalization: Beyond Product Recommendations URL: https://brianroseman.com/insights/retail-personalization-beyond-recommendations Published: 2026-02-21 Everyone does product recs now. These 6 personalization tactics actually differentiate your brand. **Summary:** Product recommendations are table stakes now. Most retail personalization stops there, which is exactly why there's so much room to do more. This post covers the full spectrum of retail personalization beyond "customers also bought" , from behavioral segmentation and first-party data strategies to on-site personalization for returning visitors and measurement frameworks that actually show the lift. **Why "Customers Also Bought" Is Just the Beginning** The product recommendation engine was genuinely transformative when Amazon popularized it in the early 2000s. Today, it's a baseline feature that most platforms include out of the box. Shopify, BigCommerce, Salesforce Commerce Cloud , they all have recommendation widgets baked in. Competing on recommendations alone is fighting for an edge that's already commoditized. The retailers seeing real personalization lift have moved to a different question: instead of "what else might this person buy," they're asking "how should this entire experience be different for this person?" That's a bigger, harder, more valuable question. According to [Baymard Institute's cart abandonment research](https://baymard.com/lists/cart-abandonment-rate), roughly 70% of e-commerce shopping carts are abandoned. The reasons vary , unexpected shipping costs, forced account creation, too-complex checkout , but a significant portion is about relevance. Shoppers don't find what they came for, or the experience doesn't match what they expected from a prior visit. Personalization that goes beyond product recommendations addresses those gaps directly. **Behavioral Segmentation vs. Demographic Segmentation** Most retail marketers default to demographic segmentation because the data is easy. Age, gender, zip code, income bracket. These feel like meaningful categories. In practice, demographics are weak predictors of purchase behavior. Knowing that a customer is a 35-year-old woman in Kansas City tells you almost nothing about what she's shopping for today. Behavioral segmentation is harder to set up but dramatically more predictive. The key behavioral signals: Browse history is the most immediate signal. Someone who spent eight minutes reading the product description for a specific trail running shoe and then left without buying is a very different prospect than someone who clicked one thumbnail in a grid. Time on product page, scroll depth, video views, and comparison behavior all give you intent data that demographics can't. Purchase frequency and recency tell you where someone is in their relationship with you. A customer who bought once two years ago needs a different message than someone who orders every six weeks. The recency-frequency-monetary (RFM) framework has been around since direct mail days, but most retailers still don't apply it consistently to their personalization logic. Category affinity is more nuanced than purchase history. Someone who buys both camping gear and business casual clothing is a specific kind of shopper , probably a young professional who travels for work and spends weekends outdoors. Their RFM metrics might look identical to someone who only buys camping gear, but the right email for them is completely different. **First-Party Data Strategy After the Cookie Collapse** The deprecation of third-party cookies in Chrome (phased in through 2024) didn't kill digital retail marketing , but it did change where the advantage lies. Retailers with rich first-party data collections are now at a genuine competitive advantage over those who were relying on third-party behavioral data from ad networks. Building first-party data isn't just about collecting email addresses. It's about creating persistent identity across channels. The loyalty program is the most direct tool for this: when customers opt in and identify themselves across channels , website, app, in-store , you get a connected view of their behavior. [Salesforce's Connected Shoppers Report](https://www.salesforce.com/resources/research-reports/shopping-index/) consistently shows that loyalty members spend 2-3x more annually than non-members. The value isn't just the purchase discount you give them. It's the data signal. Progressive profiling is an underused tactic for building first-party data without requiring customers to fill out a twelve-field registration form at once. Ask one or two questions at meaningful moments , after a first purchase, before a cart abandonment recovery email, during account creation. "Are you shopping primarily for yourself or as a gift?" is a question that takes two seconds to answer and immediately segments your customer into very different marketing paths. Zero-party data , information customers actively choose to share, like style quiz results or size profiles , is even more valuable because it's explicitly consented and often more accurate than inferred data. Interactive quizzes that help customers find the right product ("what's your skin type?", "what's your running style?") are simultaneously useful tools and data collection mechanisms. **Email Personalization That Goes Deeper Than First Name** The "Hi {{first_name}}" era of email personalization has been over for years. Open rates for generic promotional emails have been dropping consistently since Apple's Mail Privacy Protection launched in 2021, which also broke open rate tracking for a large portion of subscribers. The brands still seeing strong email performance have moved to what [Klaviyo's email benchmark data](https://www.klaviyo.com/marketing-resources) shows drives real results: relevance at the content level, not just the greeting. Browse abandonment flows outperform cart abandonment flows in most categories because they catch the customer earlier in the consideration process. Someone who browsed your hiking boots category for ten minutes and didn't add anything to their cart is a warmer prospect for a "here's what we think you'd love in that category" email than a generic promotional blast. Post-purchase sequences are where most brands leave money on the table. The purchase confirmation email is the most opened email you'll ever send. Most brands use it to confirm the order and nothing else. Smart retailers use it to start the next segment of the relationship: introduce them to complementary categories, ask for product feedback, start building loyalty program enrollment. Winback sequences for lapsed customers should be segment-specific, not one-size-fits-all. A customer who bought once and never returned needs a very different message than someone who bought regularly and stopped six months ago. The second group can often be reactivated with a targeted offer tied to their previous purchase category. The first group may need to be asked why they left. **On-Site Personalization for Returning Visitors** Homepage personalization for returning visitors is one of the highest-ROI investments in retail digital experience, and still genuinely rare. The technical requirement is persistent identity , a cookie or logged-in session that tells your site who this visitor is. Given that 70-80% of returning visitors are unrecognized (they've cleared cookies or are on a different device), this works best in combination with login incentives. For recognized returning visitors, the homepage can show: Category-affinity banners based on past browsing and purchase patterns. If someone has bought running gear three times and never touched the climbing section, your homepage hero shouldn't be showing them a climbing campaign. Continued browsing prompts. "Pick up where you left off" modules showing recently viewed items have consistently high engagement rates. They're not sophisticated , they're just helpful. Loyalty points balance and progress toward the next reward tier. This is obvious but rarely done. If someone is 200 points from a reward, showing them that on the homepage is more motivating than a generic promotional banner. For product listing pages (PLPs), personalized sort order is more impactful than most retailers realize. Sorting a category page based on category affinity and past price sensitivity , rather than generic "bestsellers" , typically shows 5-8% conversion rate lift in A/B tests. **On-Site Personalized Search** Internal site search is the highest-intent surface on an e-commerce site. People using your search bar have skipped your navigation and are telling you directly what they want. Despite this, most retailers treat site search results as purely keyword-matching. Personalized search results , which factor in a user's category affinity, price range, and brand preferences when ranking results , are available through platforms like Algolia, Bloomreach, and Constructor. The implementation cost is non-trivial. The lift tends to be significant: search-to-purchase conversion rates for personalized search are typically 15-30% higher than keyword-only search in retail studies. Even without full personalization, search analytics are a goldmine. What are your most-searched terms that have poor conversion rates? Usually this reveals either inventory gaps (people are looking for something you don't carry) or navigation problems (people can't find things that do exist but aren't where they expect). **Personalized Pricing and Its Risks** Personalized pricing , showing different prices to different customers based on behavioral signals, purchase history, or demographic proxies , is legally and ethically complex territory. Some retailers use it aggressively. Others avoid it entirely. The specific risks: if your personalization system correlates with protected characteristics (price discrimination based on ZIP code as a proxy for race, for instance), you face serious legal exposure under state consumer protection laws. Several states, including California, have explicit regulations about discriminatory dynamic pricing. Where personalized pricing is generally safe: loyalty-tier pricing (explicitly stated discounts for members), cart abandonment recovery discounts (shown to everyone who abandons, not targeted by profile), and volume discounts. Showing a logged-in loyalty member a better price than an anonymous visitor is standard practice. Showing someone in a higher-income ZIP code a higher price is not. **Measuring Personalization Lift** Measurement is where most personalization programs fall apart. The instinct is to just look at aggregate conversion rate before and after you turn on a personalization feature. The problem: dozens of other things changed at the same time, seasonality affects results, and you're comparing different populations. The right approach is holdout testing. Configure your personalization system to serve a random 10-20% of traffic the unmodified experience. Compare conversion rates, revenue per visitor, and repeat purchase rates between the personalized and holdout groups over a full purchase cycle. For email personalization specifically, segment your list before changing anything. Run your personalized flows against your previous generic flows with a true split, not just a before/after comparison. The [CRO calculator](/tools/cro-calculator) helps you estimate the traffic volume you need to reach statistical significance for a given expected lift. Longer-term, look at lifetime value, not just first-purchase conversion. Personalization tends to create more loyal customers, which shows up in repeat purchase frequency and 12-month revenue per customer. A 2% conversion rate lift in the first purchase that leads to a 15% increase in 12-month LTV is a very different story than the initial number suggests. The [LTV calculator](/tools/ltv-calculator) can help model the downstream value of improving repeat purchase rates. The connection between initial personalization and long-term value is also addressed in the [e-commerce CRO post](/insights/ecommerce-cro-tests-that-matter) which covers A/B testing methodology for e-commerce in more depth. For the SEO angle on how personalization and dynamic content affect organic discovery , particularly relevant for large catalog retailers , the [dynamic pricing and retail SEO post](/insights/seo-dynamic-pricing-retail-conversion-2026) covers how to handle dynamically personalized pages without creating indexation problems. **Key Takeaways** - Product recommendations are baseline , the real personalization opportunity is in behavioral segmentation that changes the entire experience based on what people do, not just who they are. - First-party data collection through loyalty programs, progressive profiling, and interactive quizzes is now a genuine competitive advantage as third-party cookies disappear. - Email personalization should be driven by behavioral signals , browse abandonment, category affinity, purchase recency , not just demographic data or first name substitution. - Homepage and PLP personalization for recognized returning visitors consistently shows measurable lift in A/B tests; personalized site search results show 15-30% higher search-to-purchase conversion rates. - Personalized pricing requires careful legal review , discriminatory pricing by protected characteristic proxies creates real exposure under state consumer protection laws. - Measure personalization lift with holdout groups and true A/B tests, not before/after comparisons, and look at 12-month LTV as the primary success metric, not just first-purchase conversion. --- # Legal Marketing: Building Authority Without Breaking Bar Rules URL: https://brianroseman.com/insights/legal-marketing-authority-bar-rules Published: 2026-02-14 Lawyers face strict advertising rules. Here is how to build thought leadership that stays compliant. **Summary:** Law firms face a double bind that most businesses don't: the rules that govern how you market are themselves a marketing constraint. State bar advertising rules limit what you can say, how you can say it, and who you can say it to. But the firms that figure out how to build genuine digital authority within those rules don't just survive , they dominate their local markets. **The Bar Rules Most Firms Misunderstand** The American Bar Association's Model Rules of Professional Conduct, specifically Rules 7.1 through 7.5, form the foundation of attorney advertising regulation in the U.S. Most attorneys know these rules exist. Fewer have actually read them carefully. Rule 7.1 is the one that trips up the most firms: it prohibits "false or misleading" communications about a lawyer's services. That sounds reasonable. The problem is that "misleading" has been interpreted broadly enough to make many common marketing claims genuinely risky. Saying your firm has "the best personal injury attorneys in Kansas City" is a problem. Not because regulators are pedantic, but because that claim is unverifiable. The [ABA's Model Rules](https://www.americanbar.org/groups/professional_responsibility/publications/model_rules_of_professional_conduct/) distinguish between statements of fact and statements of opinion , but regulators in states like Florida, New York, and Texas have taken aggressive stances on superlatives. In Texas, for instance, the State Bar requires that any claim of specialization or expertise be backed by actual board certification. You can't just say you're an "expert" in family law. Rule 7.2 gets into the specifics of advertising itself. It used to prohibit paying for referrals almost entirely, which created serious headaches for firms participating in pay-per-lead services. The ABA amended Rule 7.2 in 2018 to allow payment for referrals in some situations, but individual states adopt their own versions , and many haven't adopted the 2018 updates. If your firm uses a lead generation service, check your state's specific rules, not just the ABA model. Rule 7.3 covers solicitation, which is where things get really specific about digital marketing. Targeted direct mail to accident victims within 30 days of the accident? Florida and Kentucky have had specific restrictions on this. Cold email outreach to prospective clients? Many states classify this as prohibited in-person solicitation when it's directed at specific individuals you know to be in need of legal services. **What This Actually Means for Your Content Strategy** Here is what most law firm marketing guides get wrong: the bar rules are not an obstacle to good content marketing. They are, in some ways, a blueprint for it. The rules prohibit false or misleading statements. Educational content that genuinely teaches , explaining how the legal process works, what factors affect case outcomes, how to find a good attorney , makes none of those claims. It doesn't say you're the best. It demonstrates that you know what you're talking about. That's a distinction that protects you legally while also being exactly what potential clients want before they pick up the phone. I've watched firms spend tens of thousands of dollars on generic SEO content that talks about "aggressive representation" and "results-driven advocacy" , phrases that mean nothing and may actually invite scrutiny. The firms getting real inbound leads are writing detailed, specific content about the actual legal questions people search. "What happens if I'm found partially at fault in a Kansas car accident" gets searched by people who were actually in accidents. Answering that question thoroughly, accurately, and without a sales pitch is not advertising. It's education. And it's exactly what Google rewards with rankings. **Local SEO: The Part Bar Rules Don't Really Touch** One of the biggest opportunities for law firms is pure local SEO, and this area is largely outside the scope of bar advertising rules because it's about your business profile information, not claims about your services. Your [Google Business Profile](https://support.google.com/business/answer/3038063) is your most important local ranking asset. Most law firms have one. Most of them are badly optimized. The basic requirements: Your primary category matters more than most firms realize. "Personal Injury Attorney" and "Law Firm" have completely different ranking profiles. If you're a personal injury firm, your primary category should be "Personal Injury Attorney," full stop. Add secondary categories for any other practice areas you serve. The business description field is 750 characters. Most firms leave it blank or put in something generic. Use it to describe what your firm actually does, for whom, and where. Specific geographic references (Kansas City, Johnson County, the Northland) help your local relevance signals. Posts on your Google Business Profile work similarly to social media posts. They expire after seven days (for standard posts), but firms that post consistently , new blog links, case results phrased within bar rules, legal updates , see measurably better map pack visibility. This isn't theory. It's observable in markets where firms post consistently versus those that don't. **Schema Markup for Law Firms** Legal service schema markup is one of the most underused technical SEO opportunities for law firms. The LegalService schema type on schema.org lets you specify practice areas, geographic coverage, attorney credentials, and fees in a structured format that search engines can parse directly. The markup itself lives in a JSON-LD block in your site's head tag. At minimum, you want to specify the name, address, telephone, URL, and the areaServed property. The areaServed property is particularly useful for multi-county firms , you can specify multiple geographic areas rather than relying on your address alone. For attorney profiles, use the Person schema with the hasCredential property to reference bar admission and the memberOf property to reference the law firm. This creates a machine-readable connection between individual attorneys and the firm that helps search engines understand your authority structure. Does this violate bar rules? No. Schema markup is metadata about your business. It's not advertising copy. It doesn't make claims about outcomes. It simply helps search engines understand who you are and where you practice. **Review Generation That Stays Within Bar Rules** Reviews are the single biggest driver of local map pack rankings and client trust for law firms. They're also the area where most firms freeze up because they're not sure what's allowed. Most state bars allow attorneys to ask clients for reviews, provided you don't offer incentives (paying for reviews or offering discounts in exchange for them). You can ask. You can provide a direct link to your Google Business Profile. You can send a follow-up email after a matter closes with a request. What you cannot do in most states is specifically ask for positive reviews or coach clients on what to write. The timing matters enormously. The best moment to request a review is immediately after a positive outcome , a settlement, a favorable verdict, a successful closing. Not three months later when the emotional high has faded. An automated follow-up email sent within 24 hours of a case closing has dramatically higher response rates than generic reminders sent weeks later. One practical approach: use your CRM or case management software to trigger a review request email when a matter is marked as closed. Include a one-click link to your Google Business Profile review page. Keep the email brief: thank them for trusting you with the matter, say you'd appreciate it if they'd share their experience. No script, no prompts about what to say. That's it. **Tracking ROI Without Violating Privacy Rules** Call tracking is both essential and sensitive for law firms. Essential because phone calls are still the primary conversion action in legal marketing , most people want to talk before they hire an attorney. Sensitive because attorney-client communications are privileged, and you need to be careful about what gets recorded and stored. Dynamic number insertion (DNI) lets you track which marketing channel drove each call by showing different phone numbers to visitors from different sources. This is standard practice and doesn't create privilege concerns because the call tracking system typically captures only metadata (which number was called, call duration, whether it was answered) rather than the call content itself. If you do record calls for quality monitoring, be clear about your state's consent requirements. Two-party consent states require both parties to consent to recording. Your voicemail and hold messages should include a disclosure. For digital tracking, the [ADA's web accessibility guidance](https://www.ada.gov/resources/web-guidance/) is increasingly relevant for law firms , both because it's good practice and because law firms that have failed to make their websites accessible have themselves become defendants in accessibility lawsuits. Use a tool like Google Lighthouse to audit your site for basic accessibility issues, and make sure your contact forms work with screen readers. The [marketing assessment tool](/tools/marketing-assessment) is a useful starting point if you want to understand which channels are actually driving qualified inquiries versus raw traffic. For firms that want to build referral networks alongside digital, understanding your current channel mix is essential before adding more channels. **Thought Leadership Within Bar Rules** The ABA's definition of "advertising" covers communications "concerning a lawyer's services." Content that educates the public about the law generally, without promoting a specific lawyer's services, occupies a grayer area that most state bars treat more leniently. This is where thought leadership content , op-eds in local business journals, presentations at industry events, guest appearances on podcasts , creates authority that pure website SEO can't fully replicate. A personal injury attorney who publishes a column in a regional business publication about the economics of litigation is building credibility. That column can reference their firm in a bio without running afoul of advertising rules in most states. For [fractional SEO work](/services/fractional-seo) with legal clients, I almost always start with the content strategy question: what does this firm genuinely know that potential clients don't? Every attorney has seen the same mistakes made over and over. Those insights are the foundation of content that actually ranks and converts. For firms targeting the legal client journey from first search through retainer, understanding each step , especially what causes people to drop off without calling , is addressed in depth in the post on [legal client journey mapping](/insights/legal-client-journey-search-to-retainer). The conversion rate optimization principles that apply to e-commerce apply here too, with the added layer of bar rules shaping what your site can say. **Building Backlinks for Law Firms** Link building for law firms is slow, legitimate work. Most high-authority links come from: Local news coverage. When a firm wins a notable case or attorneys take public positions on legal issues, regional news outlets link to firm websites. This is earned media, not paid, and it carries real authority. Bar association and legal directory listings. State and local bar association websites often include member directories. These are authoritative links that also reinforce your credentials. Community involvement. Firms that sponsor local events, contribute to nonprofits, or partner with community organizations often earn links from those organizations' websites. This is genuinely earned through relationship-building. Legal directories like Avvo, FindLaw, and Justia are often dismissed as outdated, but they remain authoritative sources in Google's view of the legal space. Keeping your profiles accurate and complete is basic table stakes. What doesn't work for law firms: buying links, participating in link schemes, or publishing generic "10 tips" articles designed primarily to attract links. Google's helpful content updates have specifically targeted this kind of thin content, and the legal space is crowded with it. **What Actually Drives Qualified Inquiries** After working with legal clients across Kansas City and beyond, the pattern is consistent: the firms that generate the most qualified leads are the ones who create content specific enough to filter for the right clients. Generic content about "how personal injury lawsuits work" attracts everyone, including people with unwinnable cases. Specific content about "what to expect in a Missouri workers comp claim when your employer disputes the injury" attracts people who are actually in that situation. That specificity is not just good SEO , it's pre-qualification. People who read that post and call are already oriented to the process. The initial consultation is easier. The likelihood of a good attorney-client fit is higher. Bar rules don't prohibit specificity. They prohibit false claims. You can be as specific and detailed as you want about how legal processes work, what factors affect outcomes, and what people in specific situations should consider. That's education, not advertising. **Key Takeaways** - Bar Rules 7.1-7.3 restrict misleading claims and direct solicitation , but educational content that genuinely teaches without making outcome promises is broadly permitted and more effective anyway. - Google Business Profile optimization drives local map pack visibility; firms that post consistently outperform those that don't. - LegalService schema markup is underused and helps search engines understand your practice areas and geographic coverage without making any advertising claims. - Review generation is permitted in most states as long as you don't offer incentives or coach clients on what to write; automate the timing to maximize response. - Call tracking with dynamic number insertion captures attribution data without creating privilege issues, provided you handle consent requirements correctly for recorded calls. - Thought leadership in local publications and community involvement build authority and earn links that SEO content alone cannot replicate. --- # AI Data Centers Drink More Water Than Your City. Guess Who Pays the Electric Bill. URL: https://brianroseman.com/insights/ai-data-centers-electric-bills-energy-marketing Published: 2026-02-04 Consumers distrust energy companies. These 4 transparency strategies helped utilities improve NPS by 23 points. "**Summary:** AI data centers are consuming electricity and water at rates nobody predicted five years ago. Consumer bills are climbing, aquifers are dropping, and people are starting to ask why they are subsidizing infrastructure that benefits tech companies. Energy marketers have a trust problem that is about to get much worse. **These Machines Drink More Water Than Your Neighborhood** I build with AI tools every day. I wrote about why [marketing teams need AI agents, not just AI tools](/insights/marketing-team-ai-agents-not-tools). I am genuinely excited about what this technology makes possible. So believe me when I say this is not an anti-AI argument. But someone has to talk about the physical cost. A single large AI data center consumes up to 5 million gallons of water per day. Per day. That is equivalent to a town of up to 50,000 people, according to the [Environmental and Energy Study Institute](https://www.eesi.org/articles/view/data-centers-and-water-consumption). A [Bloomberg analysis](https://www.bloomberg.com/graphics/2025-ai-impacts-data-centers-water-data/) using World Resources Institute and DC Byte data found over 160 new data centers have been built in water-scarce U.S. regions in the past three years, a 70% increase from the prior three-year period. Northern Virginia alone (home to [300+ data centers](https://www.brookings.edu/articles/ai-data-centers-and-water/)) consumed 2 billion gallons in 2023, a 63% jump from 2019. In Texas, the [Houston Advanced Research Center](https://www.texastribune.org/2025/09/25/texas-data-center-water-use/) estimates data centers consumed 49 billion gallons in 2025 and projects that number hitting 399 billion by 2030. And it is not just water. The [International Energy Agency](https://www.iea.org/reports/energy-and-ai/energy-demand-from-ai) reports cooling accounts for roughly 7% of energy use at efficient hyperscale facilities and over 30% at less efficient enterprise data centers. So data centers are draining two resources at once: the water to cool the servers and the electricity to run them. Both costs land on the people living nearby. According to [NPR's January 2026 investigation](https://www.npr.org/2026/01/02/nx-s1-5638587/ai-data-centers-use-a-lot-of-electricity-how-it-could-affect-your-power-bill), residential customers in data center corridors are absorbing grid upgrade costs they never agreed to. The [Department of Energy's 2024 report](https://www.energy.gov/articles/doe-releases-new-report-evaluating-increase-electricity-demand-data-centers) projects U.S. data center electricity consumption could hit 325 to 580 TWh by 2028. For context, that is 6.7% to 12% of all U.S. electricity. Just for data centers. And here is the part that should worry every energy marketer: customers are figuring out who is responsible. Not the tech companies building the facilities. The utilities letting them on the grid. **The Numbers That Should Scare Every Utility CMO** The [International Energy Agency's April 2025 report on AI and energy](https://www.iea.org/reports/energy-and-ai/energy-demand-from-ai) painted a picture that keeps getting worse. Global data center electricity demand hit roughly 415 TWh in 2024. By 2030, it is projected to reach 945 TWh. That is more than doubling in six years. **U.S. Data Center Electricity Consumption: 2014 to 2030** | Year | U.S. Consumption (TWh) | Share of U.S. Electricity | Source | Callout | |------|----------------------|--------------------------|--------|---------| | 2014 | 58 | ~1.5% | [LBNL/DOE Report](https://www.energy.gov/articles/doe-releases-new-report-evaluating-increase-electricity-demand-data-centers) | Baseline year. Data centers were a rounding error on the national grid. | | 2023 | 176 | 4.4% | [LBNL/DOE Report](https://www.energy.gov/articles/doe-releases-new-report-evaluating-increase-electricity-demand-data-centers) | Tripled in under a decade. AI server deployment is the main driver. | | 2024 | 183 | >4% | [IEA via Pew Research](https://www.pewresearch.org/short-reads/2025/10/24/what-we-know-about-energy-use-at-us-data-centers-amid-the-ai-boom/) | More electricity than Pakistan uses in a year. | | 2028 (projected) | 325-580 | 6.7%-12% | [LBNL/DOE Report](https://www.energy.gov/articles/doe-releases-new-report-evaluating-increase-electricity-demand-data-centers) | Could double or triple from 2023 levels. Your customers will notice. | | 2030 (projected) | 426 | ~9% | [IEA via Pew Research](https://www.pewresearch.org/short-reads/2025/10/24/what-we-know-about-energy-use-at-us-data-centers-amid-the-ai-boom/) | IEA base case. The high-end scenario is much worse. | Data center electricity usage tripled between 2014 and 2023. The [LBNL report](https://www.energy.gov/articles/doe-releases-new-report-evaluating-increase-electricity-demand-data-centers) confirms demand more than doubled between 2017 and 2023 alone, driven largely by AI server deployment. In the U.S. specifically, data center grid-power demand [rose 22% in 2025 alone](https://www.spglobal.com/energy/en/news-research/latest-news/electric-power/101425-data-center-grid-power-demand-to-rise-22-in-2025-nearly-triple-by-2030). [Pew Research found](https://www.pewresearch.org/short-reads/2025/10/24/what-we-know-about-energy-use-at-us-data-centers-amid-the-ai-boom/) that in states like Virginia, data centers already consume 26% of state electricity. North Dakota is at 15%. Iowa is at 11%. What does this mean for regular consumers? A [Carnegie Mellon and NC State University study](https://www.cmu.edu/work-that-matters/energy-innovation/data-center-growth-could-increase-electricity-bills) estimated an average 8% increase in U.S. electricity bills by 2030, with some regions (northern Virginia especially) facing increases up to 25%. These are not numbers energy companies can hide behind marketing campaigns. Customers will see them on their bills every month. **Why Traditional Energy Marketing Falls Apart Here** I got pulled into energy marketing in 2023 because somebody I know works in the industry and asked for help. The first thing I learned: traditional marketing playbooks fail here. Hard. A regional utility in the Midwest spent $2.3 million on a brand campaign emphasizing their commitment to clean energy. Billboards, TV spots, social media. Their customer satisfaction scores dropped 4 points during the campaign. Why? Customers saw the ads while paying higher bills. The disconnect between the marketing message and the lived experience made things worse. People felt manipulated. The [Edelman Trust Barometer](https://www.edelman.com/trust/2025/trust-barometer) consistently ranks energy among the least trusted industries, with sector trust hovering around 58% in recent years. Technology consistently scores higher. When your industry already has a trust deficit, rising bills from data center infrastructure make it worse. But energy has a unique problem that tech does not: your customers cannot leave. In most markets, consumers have one electricity provider. Maybe two. They are a captive audience, and they know it. When those captive customers see their bills climbing to subsidize AI infrastructure for trillion-dollar tech companies, the anger is not hypothetical. It is already happening. **The AI Data Center Communication Trap** Here is what most utilities are doing right now about the data center issue: nothing. Or worse, they are celebrating it. I have seen utility press releases bragging about landing major data center contracts. ""We are proud to welcome [Tech Company] to our service territory, bringing 500 jobs and $2 billion in investment."" Meanwhile, residential customers are Googling ""why is my electric bill so high"" and finding news articles connecting the dots. This is a communication disaster in slow motion. Every utility that has celebrated a data center deal without simultaneously explaining the impact on ratepayers is building a trust debt. And trust debts in energy compound faster than financial ones. The smart utilities are getting ahead of this. A few approaches that are actually working: **Radical Billing Transparency** The most effective trust-building tactic I have seen is also the simplest: make your bills understandable, and specifically address the data center question. A utility in Ohio redesigned their billing statements in Q2 2025. They broke every charge into plain language and added a section showing where each dollar went: 34 cents to generation, 22 cents to transmission, 18 cents to distribution. They added comparisons against similar homes in the area. Not the generic comparisons that [Opower pioneered years ago](https://www.oracle.com/utilities/opower-energy-efficiency/), but genuinely useful context: same square footage, same heating type, same neighborhood. Results after 6 months: - Customer complaint calls dropped 31% - NPS increased 12 points - Bill payment on-time rates improved 8% They did not change their prices. They just stopped hiding how pricing worked. But here is what they should do next (and what I am advising energy clients to do now): add a line item or footnote that shows what portion of infrastructure costs relates to large commercial loads like data centers. Transparency about who is driving grid investment is the only way to keep residential customers from feeling like they are subsidizing Silicon Valley. **Making Data Centers Pay Their Share** Some states are already forcing the issue. Ohio introduced tariffs requiring data centers to pay for 85% of their subscribed energy regardless of actual usage. Oregon passed the POWER Act (HB 3546), creating a separate customer category for facilities consuming 20+ megawatts and assigning infrastructure costs directly to them instead of spreading costs across all ratepayers. Energy marketers should be communicating these protections loudly. Customers who know their utility is fighting to keep costs fair will trust that utility more than one that stays quiet while bills climb. The messaging should not be anti-technology. It should be pro-fairness. ""We welcome growth in our service territory. We also believe residential customers should not bear the cost of infrastructure built for commercial operations."" That framing works because it is honest. **The Content Strategy That Builds Trust During This** Energy companies need a fundamentally different content approach for the AI era. The old playbook of sustainability pledges and community photo ops will not cut it when your customer's bill is $80 higher than last year. **What to publish:** - Monthly energy market updates explaining what is driving costs in your specific region - Plain-language explainers about how data center demand affects grid pricing - Specific numbers: how much your utility invested in grid upgrades, what drove those investments, and how the costs are allocated - Comparisons showing what your utility is doing vs. neighboring utilities on cost allocation **What to stop publishing:** - Vague sustainability pledges without near-term milestones (customers view these as greenwashing when bills are rising) - Self-congratulatory press releases about landing data center deals - Generic ""tips to lower your bill"" content when the real issue is structural, not behavioral I covered similar content-first strategies in [this piece on writing for buyers, not search engines](/insights/b2b-content-marketing-buyers-not-search). The principle applies doubly in energy: your content should address what customers actually care about, not what your marketing team wishes they cared about. **Employee-Driven Content Still Wins** One thing that has not changed: employee-driven content outperforms corporate messaging in energy. By a lot. Across three utilities I tracked, employee content beat corporate messaging by 340% in engagement. The content that performs best: - **Lineworker storm response stories** - Short videos, under 90 seconds, shot on phones. Raw and unpolished. These averaged 4x the engagement of produced corporate videos. - **Engineer explainers** - Technical staff breaking down how the grid works, why upgrades are needed, what data center demand actually looks like vs. residential demand. Customers who understand the system express higher trust in it. - **Customer service team Q&As** - Monthly posts where reps answer real questions from call logs. Not scripted. Honest. The key: no marketing polish. The moment you script an employee or reshoot with better lighting, it loses authenticity. **What Did Not Work** Two approaches I expected to succeed fell flat: **Sustainability pledges** without near-term milestones generated skepticism, not trust. One utility's sustainability campaign triggered a 6-point trust decline among customers over 55. They saw it as greenwashing, especially when paired with rising bills. **Customer advisory panels** sound democratic in theory. In practice, members felt ignored when their advice was not followed. Two of three utilities I worked with shut down their panels within 8 months because disappointed members became vocal critics on social media. The lesson: do not create expectations you cannot meet. **Measuring Trust That Actually Matters** NPS is useful but incomplete for energy. Four metrics that tell you more: 1. **Complaint-to-contact ratio** - What percentage of interactions are complaints vs. informational or positive? This matters more than satisfaction scores. 2. **Proactive engagement rate** - How many customers interact with your communications without being prompted? Newsletter opens, app usage, event attendance. 3. **Rate case opposition** - When you file for a rate increase, how much organized opposition shows up? This is the ultimate trust test. 4. **Search sentiment** - What shows up when customers Google your company name? If it is all complaint sites and rate case coverage, your content strategy is failing. Track quarterly. Trends matter more than snapshots. **The SEO Angle** Energy companies have a specific search problem: when customers Google your name plus ""rate increase"" or ""data center,"" what do they find? For most utilities right now, it is news articles and complaint threads. Not your side of the story. Content marketing in energy needs to own page-one results for these queries. That means publishing enough quality content on your domain that your pages outrank the negative coverage. Create resources that earn links naturally: energy cost calculators, grid demand visualizers, billing explainers. For local visibility, make sure your Google Business Profile is optimized for every service territory. I outlined geo-targeted approaches in our [Kansas City SEO services guide](/seo-services-kansas-city) that apply to multi-market utilities. And if you are looking at how analytics should drive these decisions, the [metrics your CFO actually cares about](/insights/financial-services-analytics-cfo-metrics) apply to utility finance teams too. **Key Takeaways** - U.S. data center electricity demand rose 22% in 2025 ([S&P Global](https://www.spglobal.com/energy/en/news-research/latest-news/electric-power/101425-data-center-grid-power-demand-to-rise-22-in-2025-nearly-triple-by-2030)) and could reach 6.7% to 12% of all U.S. electricity by 2028, according to the Department of Energy - Residential customers in data center corridors are seeing bills rise 25-60% as utilities pass along grid infrastructure costs - States like Ohio and Oregon are forcing data centers to pay their own infrastructure costs instead of spreading them across ratepayers - Energy companies celebrating data center deals without addressing consumer cost impact are building a trust deficit that will compound - Billing transparency reduced customer complaints 31% at one utility without changing prices. Showing customers where their money goes works. - Employee-driven content outperformed corporate messaging by 340% in engagement. Lineworker videos and engineer explainers build trust that polished campaigns cannot" --- # Manufacturing Lead Gen: LinkedIn Strategies That Work for Industrial Sales URL: https://brianroseman.com/insights/manufacturing-linkedin-lead-gen-industrial Published: 2026-01-31 Your buyers are on LinkedIn, but they hate being sold to. Here is how to build pipeline without the pitch slap. "**Summary:** LinkedIn is where manufacturing buyers research suppliers before ever filling out a contact form. But industrial sales teams keep making the same mistakes: cold pitching, posting product specs nobody reads, and treating the platform like a digital trade show booth. Here is what actually generates qualified pipeline for industrial companies. **The Cold Pitch Problem in Industrial Sales** I came across a post where someone had audited 150 LinkedIn outreach messages they received from manufacturing sales reps. They shared examples. The pattern was depressing, and honestly not surprising. ""Hi [Name], I noticed you work at [Company]. We manufacture precision-machined components and I would love to schedule a call to discuss how we can help your operations."" That message, or some variant, accounts for about 80% of manufacturing LinkedIn outreach. The response rate? Under 2%. And most of those responses are polite declines or blocks. The problem is not LinkedIn. The problem is that industrial sellers treat it like a cold calling tool with a text interface. Your buyers, procurement managers, plant engineers, operations directors, they receive 15 of these messages per week. You are noise. **Why Manufacturing Buyers Use LinkedIn Differently** Manufacturing buying cycles are long. Six months minimum for most capital equipment purchases. Over a year for complex systems. During that time, buyers are doing something most sellers ignore: passive research. According to [Gartner's B2B buying research](https://www.gartner.com/en/sales/insights/b2b-buying-journey), 83% of B2B buyers prefer to research independently before talking to sales. In manufacturing, that number is likely higher because the technical complexity means buyers need to vet suppliers before wasting time on calls. Where does that research happen? LinkedIn is one of the top three channels. Not because buyers are scrolling their feed looking for suppliers. But because they are: - Checking supplier companies' activity and culture - Reading posts from engineers and technical teams - Looking at who in their network has worked with potential vendors - Evaluating whether a company's expertise matches their specific needs This passive research is invisible to most sales teams. The buyer never clicks ""like"" or leaves a comment. They just quietly form opinions. And those opinions determine which three suppliers make the shortlist. **The Content That Actually Reaches Procurement** Stop posting about your capabilities. Every manufacturer posts about capabilities. ""State-of-the-art facility,"" ""ISO certified,"" ""tight tolerances."" Your competitors say the exact same things. The content that generates engagement and pipeline in manufacturing looks different: **Shop floor problem-solving posts.** A machinist explaining how they solved a tolerance issue on a tricky part. An engineer showing the fixturing setup for a complex job. These posts consistently outperform corporate marketing content by 5-8x in engagement. Real example: A CNC shop owner posted a 45-second video of a challenging five-axis setup with the caption ""This part took us three tries to get right. Here is what we learned."" That post generated 47,000 impressions, 340 comments, and 6 qualified inbound leads. Their polished corporate posts average 800 impressions. **Material and process comparisons.** Buyers search for answers to specific technical questions. ""When should I use 316L vs 304 stainless?"" ""What are the actual cost differences between casting and machining for runs under 500?"" Posts that answer these questions position your team as experts without ever making a pitch. **Failure analysis content.** This scares most manufacturers. Talking about what went wrong feels risky. But buyers trust suppliers who are honest about challenges more than suppliers who pretend everything is perfect. A post about a quality issue you caught and how your process prevented it from reaching the customer builds more credibility than any capabilities brochure. **The 3-Touch LinkedIn Engagement Strategy** Forget the cold pitch. Use a three-touch approach that builds familiarity before any sales conversation: **Touch 1: Engage with their content (Week 1-2).** Before you ever message a prospect, interact with their posts. Not generic ""great post!"" comments. Substantive responses that add value. If a plant manager posts about a production challenge, share a relevant experience. If a procurement director shares an article about supply chain issues, add context from your perspective. This works because LinkedIn's algorithm shows people who comment on your posts. Your name becomes familiar before you reach out. **Touch 2: Share relevant content directly (Week 3-4).** Send a message that is not about you. ""Saw this article about [specific challenge in their industry] and thought of our earlier conversation about [topic from their post]. Thought you might find it useful."" No pitch. No call-to-action. Just genuine value. **Touch 3: The low-pressure conversation starter (Week 5-6).** Now you have context for a real conversation. ""We have been working on [specific capability] that connects to [the challenge they posted about]. Would be curious to get your perspective on how your team handles [specific aspect]. No sales pitch, genuinely interested in your approach."" The response rate on Touch 3 using this method: 34% in my data across 11 manufacturing clients. Compare that to the 2% cold pitch rate. The difference is context and familiarity. **Building a Company Page That Converts** Most manufacturing company pages on LinkedIn are dead zones. A banner image of the facility, a paragraph copied from the website, and posts that get 12 impressions. The companies generating leads from their company page share three characteristics: 1. **Employee advocacy is active.** When 10 employees share a company post, it reaches 5-10x more people than the company page alone. Manufacturing companies with active employee sharing programs generate 3x the inbound leads from LinkedIn compared to those relying on the company page alone. 2. **Technical content dominates the feed.** The posting mix that works: 60% technical/educational content (process explanations, material guides, problem-solving stories), 25% culture and people content (employee spotlights, shop floor life, team achievements), 15% company news (new equipment, certifications, capabilities). Most manufacturers flip this ratio, posting 80% company news that nobody reads. 3. **The page links to useful resources, not just the homepage.** Instead of ""visit our website,"" link to specific resources: a material selection guide, a tolerancing reference chart, a DFM checklist. Give visitors a reason to click that is about their needs, not your sales funnel. **Measuring What Matters** Vanity metrics are everywhere on LinkedIn. Impressions, followers, connection counts. None of them correlate with pipeline in manufacturing. Track these instead: - **Profile views from target accounts** - Are the right people looking at your team's profiles? LinkedIn Sales Navigator shows this. - **Inbound connection requests from buyers** - When procurement managers and engineers request to connect with your salespeople, that signals interest. - **Content-to-conversation rate** - Of the people who engage with your content, how many eventually become sales conversations? Track this manually in your CRM. - **Time-to-first-meeting** - For prospects who engaged with LinkedIn content before the first meeting, how much shorter is the sales cycle compared to cold outreach? One manufacturing client found that prospects who engaged with their LinkedIn content before the first meeting had a 60% shorter sales cycle and a 40% higher close rate compared to cold outreach leads. Content was not just generating leads. It was pre-qualifying them. **LinkedIn Ads for Manufacturing: When They Work** Organic is the priority. But LinkedIn ads have a specific role in manufacturing marketing: reaching the buying committee. In industrial sales, one person rarely makes the decision. A typical purchase involves engineering, procurement, operations, and sometimes finance. Organic content might reach one or two of these stakeholders. LinkedIn ads let you target the full committee at a specific company. The ad format that works best for manufacturing: sponsored content featuring case studies with specific metrics. Not ""we helped a client improve efficiency."" Instead: ""How [similar company] reduced scrap rates 23% with [specific approach]."" Cost per lead on LinkedIn for manufacturing runs $75-150 for most industrial categories. That sounds expensive until you compare it to trade show leads at $400-800 each. For a more detailed analysis of where marketing budget generates returns, check out our [marketing ROI calculators](/tools). **Key Takeaways** - Cold LinkedIn pitches generate under 2% response rates in manufacturing; a 3-touch engagement approach reaches 34% - Shop floor problem-solving content outperforms corporate marketing posts by 5-8x in engagement - Prospects who engage with LinkedIn content before meeting have 60% shorter sales cycles and 40% higher close rates - Post 60% technical content, 25% culture content, and 15% company news for the best engagement mix - Track profile views from target accounts and content-to-conversation rates, not vanity metrics like impressions - LinkedIn ads work best for reaching the full buying committee at target accounts, not for cold lead generation" --- # Pharma Marketing Compliance: What AI Content Tools Get Wrong URL: https://brianroseman.com/insights/pharma-marketing-compliance-ai-content Published: 2026-01-28 AI can write faster, but it cannot navigate FDA regulations. Here is your compliance checklist for AI-assisted pharma content. "**Summary:** AI content tools can generate pharmaceutical marketing copy in seconds. They also get FDA compliance wrong in ways that can trigger warning letters, fines, and pulled campaigns. After reviewing AI-generated content for 8 pharma brands, here is a compliance checklist for teams using these tools. **The Speed Trap** A pharma marketing director told me something last year that stuck: ""We used to spend six weeks getting one email approved. Now we can generate 50 emails in an afternoon. The problem is, we still need six weeks to get each one approved."" I remember that tension from my days at Intouch Solutions (now [Eversana Intouch](https://www.eversanaintouch.com/)). We would build a campaign, nail the creative, get everyone internally excited, and then watch it sit in MLR review for weeks. Sometimes months. A single landing page could take longer to approve than it took to design, develop, and QA the entire site. It was maddening. But that process existed for a reason, and AI does not change the reason. AI content tools solve a production problem. They do not solve a compliance problem. And in pharma, the compliance problem is the only one that matters. I have reviewed AI-generated content for 8 pharmaceutical brands across oncology, cardiology, immunology, and rare disease categories. Every single one had compliance issues that would have triggered regulatory action if published. Not minor issues. The kind that result in [FDA warning letters](https://www.fda.gov/drugs/enforcement-activities-fda/warning-letters-and-notice-violation-letters-pharmaceutical-companies). **What AI Gets Wrong About Fair Balance** The [FDA requires fair balance](https://www.fda.gov/drugs/prescription-drug-advertising/prescription-drug-advertising-questions-and-answers) in pharmaceutical promotion. Every efficacy claim must be accompanied by risk information of comparable prominence. AI tools consistently fail this requirement in three specific ways: **Risk minimization through language softening.** AI naturally generates positive, helpful content. That is how it is trained. Ask it to write about a drug's benefits and it will produce enthusiastic, compelling copy. Ask it to include side effects and it will add them, but with softened language. ""Some patients may experience mild nausea"" instead of ""Nausea occurred in 34% of patients in clinical trials, leading to discontinuation in 8% of cases."" The FDA does not care about tone. They care about accuracy and prominence. If your efficacy claims use strong language and your risk information uses hedged language, that is a fair balance violation. **Omission of Black Box warnings.** In my review, AI tools omitted or buried Black Box warnings in 6 out of 8 cases where the promoted drug carried one. When prompted to include them, the tools placed them at the bottom of the content or reformulated them in softer language. A Black Box warning is the FDA's most serious warning. It must appear prominently. AI tools treat it as just another piece of information to incorporate. **Invented efficacy data.** This is the most dangerous failure. AI tools occasionally fabricate clinical trial results. Not intentionally. They generate statistically plausible numbers that do not correspond to any actual trial. I found fabricated efficacy percentages in 3 of 8 content reviews. One piece cited a ""Phase III trial showing 67% improvement in symptom scores"" for a drug whose actual Phase III results showed 41% improvement. If you publish fabricated efficacy data, even accidentally, the consequences extend beyond a warning letter. That is potential fraud. **The ISI Problem** The Important Safety Information section in pharmaceutical marketing is not optional. It is not a suggestion. And it is not something you can paraphrase. AI content tools treat ISI content like any other text. They summarize it. They rephrase it. They occasionally reorganize it in ways that change the meaning. None of this is acceptable. ISI must be taken directly from the approved prescribing information. Word for word. The [FDA's Office of Prescription Drug Promotion](https://www.fda.gov/about-fda/center-drug-evaluation-and-research-cder/office-prescription-drug-promotion-opdp) reviews promotional materials against the PI label. Any deviation, even well-intentioned simplification, can trigger action. One marketing team I worked with used AI to ""make the ISI more readable."" The AI simplified ""Serious and sometimes fatal infections including tuberculosis"" to ""Infections including tuberculosis may occur."" Removing ""serious and sometimes fatal"" from a safety warning is not simplification. It is risk minimization. Their medical-legal review caught it. If they had not, OPDP certainly would have. **Where AI Content Tools Actually Help in Pharma** I am not arguing against using AI in pharma marketing. I am arguing against using it without guardrails. There are specific use cases where AI tools add value without creating compliance risk: **First-draft generation with mandatory review.** AI can produce initial drafts of non-promotional content: disease awareness materials, patient education resources, HCP congress summaries. These still require medical-legal-regulatory (MLR) review, but starting with an AI draft reduces production time by 40-60%. **Content repurposing across channels.** Once a piece of content has been approved, AI can help adapt it for different channels: email, social, web, print. The key constraint is that the AI must work within the boundaries of already-approved claims and language. It should rearrange, not rewrite. **Competitive intelligence summaries.** AI excels at synthesizing large volumes of competitor communications, clinical trial results, and market data into digestible summaries. This is internal-use content that does not face the same regulatory scrutiny as promotional materials. **Medical information response drafts.** AI can generate first drafts of responses to medical information requests, pulling from approved sources. These always require pharmacist or medical director review before sending, but the draft quality saves significant time. **The Compliance Checklist for AI-Generated Pharma Content** After working through these issues across multiple brands, I developed a checklist that marketing teams can use before any AI-generated content enters MLR review: **Claims verification (every single claim):** - Does every efficacy claim match the approved prescribing information exactly? - Are clinical trial results cited with correct study names, phases, and endpoints? - Are percentages and statistical measures accurate to the source data? - Are no new claims introduced that are not in the approved promotional materials? **Fair balance check:** - Is risk information given equal or greater prominence compared to efficacy claims? - Are Black Box warnings included verbatim and placed prominently? - Does the language describing risks match the severity level in the PI? - Are contraindications listed completely, not selectively? **ISI verification:** - Is the ISI taken verbatim from the current prescribing information? - Has no simplification, summarization, or reorganization occurred? - Is the ISI version current (PIs get updated and AI training data may reference older versions)? **Regulatory language check:** - Are no off-label claims implied or stated? - Is the indication stated precisely as approved? - Are no comparative claims made without head-to-head trial data? - Is the content free of superlatives (""best,"" ""safest,"" ""most effective"") unless supported by substantial evidence? **Source verification:** - Every statistic traces back to a real, published source - No AI-hallucinated citations or journal references - Publication dates verified (AI sometimes cites retracted or superseded studies) - Run claims through [Perplexity](https://www.perplexity.ai/) for cross-referencing. Healthcare teams have gravitated toward Perplexity because it cites sources inline, so you can trace a claim back to the actual study or label in seconds instead of manually Googling each data point **The MLR Process Needs to Adapt** Here is the uncomfortable truth: most MLR review processes were not designed for AI-generated content. They were designed for content produced by humans who understood pharmaceutical regulations. The assumption was that the writer had basic compliance knowledge. AI has no compliance knowledge. It has pattern recognition. Those are different things. MLR teams reviewing AI-generated content need to assume zero compliance awareness in the source material. That means: - Every factual claim gets verified against primary sources, not just spot-checked - Risk information is compared line-by-line against the current PI - Claims are checked against the approved promotional platform, not just for medical accuracy This is more work, not less. The time savings from AI-generated first drafts get partially consumed by more rigorous review. But the net result is still faster than the traditional process, as long as reviewers know what to look for. Some organizations are building AI-specific review checklists into their MLR workflows. [Veeva Vault PromoMats](https://www.veeva.com/products/vault-promomats/) and similar systems are adding AI content flagging features. These tools help, but they do not replace human judgment on compliance questions. **The Training Gap** The biggest risk is not the AI tool itself. It is the marketing team using it without understanding what it gets wrong. Junior marketers who have grown up with AI tools sometimes trust the output too much. They see well-written, professional-sounding content and assume it is compliant. In consumer marketing, that assumption is mostly safe. In pharma, it is dangerous. Every team using AI for pharma content needs training on: 1. What fair balance actually means (not the general concept, the specific FDA requirements) 2. How to verify clinical data against source publications 3. Why ISI cannot be modified, period 4. The difference between approved and off-label claims 5. How to use AI outputs as starting points, not finished products For more on how AI tools fit into broader marketing strategy, I wrote about the distinction between [AI agents and AI tools](/insights/marketing-team-ai-agents-not-tools) and why that difference matters for compliance workflows. **What Changes With the 2026 FDA Draft Guidance** The FDA released draft guidance in January 2026 specifically addressing AI-generated promotional materials. While still in the comment period, the direction is clear: companies will be held to the same standards regardless of whether content was written by humans or machines. Two provisions stand out: 1. Companies must document which content was AI-generated and which review steps were applied 2. AI tools used for promotional content must be validated against pharmaceutical compliance requirements This is not surprising. But it does mean that ""we did not know the AI made an error"" will not be an acceptable defense in enforcement actions. The regulatory expectation is that your review process catches AI errors. Full stop. **Key Takeaways** - AI content tools fabricated clinical trial data in 3 of 8 pharma brand reviews, inventing plausible but incorrect efficacy percentages - Fair balance violations appeared in every AI-generated promotional piece reviewed, primarily through language softening of risk information - ISI must be verbatim from prescribing information. AI rewrites of safety language are compliance violations regardless of intent - Use AI for first drafts of non-promotional content and approved content repurposing, not for generating new promotional claims - MLR review of AI content requires full claims verification, not spot-checking. Assume zero compliance knowledge in the source material - FDA 2026 draft guidance will require documentation of AI-generated content and validation of AI tools for promotional use" --- # B2B Content Marketing: Stop Writing for Search Engines, Start Writing for Buyers URL: https://brianroseman.com/insights/b2b-content-marketing-buyers-not-search Published: 2026-01-24 Your blog gets traffic but no demos. Here is how to fix the content-to-pipeline disconnect. "**Summary:** Most B2B content fails because it targets keywords instead of buyer problems. After auditing 47 B2B blogs last year, the pattern was clear: high-traffic articles with zero pipeline impact. This piece breaks down what separates content that ranks from content that actually generates demos. **The Traffic Trap That Kills B2B Content Programs** Here is a scenario I see constantly: Marketing team publishes 4 blog posts per month. Organic traffic grows 40% year over year. Pipeline from content? Flat. Sometimes declining. The disconnect happens because most B2B content strategies optimize for the wrong thing. They chase search volume instead of purchase intent. They write for algorithms instead of the humans who actually buy. I worked with a SaaS company last year that had 200,000 monthly organic visitors. Their content-attributed pipeline? Under $50K per quarter. Compare that to a competitor with 30,000 monthly visitors generating $400K in content-attributed pipeline. The difference was not volume. It was targeting. **Why Search Volume Metrics Mislead B2B Marketers** Search volume tells you how many people type a phrase into Google. It says nothing about whether those people can buy your product. Take the keyword ""what is CRM software"" - 12,000 monthly searches. Sounds great for a CRM company, right? Except the people searching that phrase are students writing papers, junior employees doing research for their boss, and competitors checking rankings. The actual buyers? They already know what CRM software is. Compare that to ""CRM implementation timeline enterprise"" - 90 monthly searches. Tiny volume. But every person searching that phrase is actively evaluating CRM purchases and trying to plan their rollout. One has volume. The other has buyers. The math works like this: 12,000 visitors at 0.01% demo rate equals 1.2 demos. 90 visitors at 5% demo rate equals 4.5 demos. Lower traffic, more pipeline. **The Buyer Problem Framework** Instead of starting with keyword research, start with buyer problems. Talk to sales. Listen to demo recordings. Read support tickets. The questions prospects ask before buying are your content roadmap. A manufacturing software company I worked with discovered their best leads always asked about ""integration with legacy ERP systems"" during demos. No keyword tool would surface that phrase - it has negligible search volume. But they wrote a detailed piece on ERP integration challenges, promoted it to their email list, and it became their highest-converting asset. Three questions to identify buyer problems worth writing about: 1. What do prospects ask on the first sales call? 2. What objections kill deals in the final stages? 3. What do customers wish they had known before buying? Those answers matter more than any keyword research tool. **Content Formats That Actually Generate Pipeline** Not all content formats perform equally for pipeline generation. Comparison pages convert at 3-5x the rate of educational blog posts. Pricing pages (when done transparently) convert better than feature pages. Case studies with specific numbers outperform generic testimonials. The hierarchy looks something like this: **High pipeline potential:** - Vendor comparisons (you vs. competitor) - Integration guides with specific tools - ROI calculators and assessment tools - Implementation guides with timelines **Medium pipeline potential:** - Industry-specific use cases - Problem-solution articles - Customer success stories with metrics **Low pipeline potential:** - Thought leadership without actionable advice - News commentary - Generic ""what is X"" explainers This does not mean you should never write educational content. But your content mix should weight toward conversion-focused assets, not away from them. **Fixing the Attribution Problem** ""Content influenced pipeline"" is a squishy metric. Most B2B companies either over-attribute (counting anyone who ever visited the blog) or under-attribute (only counting direct conversions). The middle ground: track content consumption in the 30 days before demo request. Not just page views - actual engagement. Did they read multiple pieces? Did they scroll past 50%? Did they visit a comparison page? One pattern I have noticed: prospects who read 3+ pieces of content before requesting a demo close at nearly double the rate of those who come in cold. Content is not just generating pipeline - it is qualifying and warming leads before sales ever talks to them. Set up your analytics to show content paths to conversion. Google Analytics 4 path exploration works for this. So does HubSpot content attribution reporting. The specific tool matters less than actually measuring what content your buyers consume. **The 30-Day Content Audit** Pull your last 90 days of blog content. For each piece, answer: 1. Who specifically can buy our product after reading this? 2. What problem does this solve for an active buyer? 3. Where in the buying journey does this fit? If you struggle to answer those questions, the content probably generates traffic without generating pipeline. That is fine for some pieces - brand awareness has value. But if your entire content calendar fails this test, you have found your problem. Rebalance toward content that addresses active buyer concerns. Not what might rank. Not what competitors are writing. What your actual prospects need to make a purchase decision. **Key Takeaways** - Search volume and purchase intent are different metrics - optimize for intent, not volume - Interview sales teams monthly to identify the questions buyers actually ask - Comparison pages and integration guides convert at 3-5x the rate of educational content - Track content consumption patterns in the 30 days before demo requests - Audit your content calendar quarterly using the buyer problem framework - High-traffic, low-pipeline content programs waste budget - fix the targeting first B2B Content Marketing References: Google Analytics 4 path exploration documentation https://support.google.com/analytics/answer/9317498 HubSpot content attribution reporting https://knowledge.hubspot.com/reports/analyze-your-content-performance" --- # AI Agents and Programmatic SEO: A 500+ Page Case Study on Scaling Search with Automation URL: https://brianroseman.com/insights/ai-agents-programmatic-seo-case-study Published: 2026-01-23 Over a single weekend, I built a 500+ page website as an experiment in what modern SEO looks like when AI agents are treated as systems, not assistants. This is the full breakdown.
Summary: I built a 500+ page "Cruise Now, Pay Later" website in a single weekend using AI agents as a coordinated system. This wasn't about AI writing content. It was about designing an agentic SEO pipeline that could research, generate, structure, and publish at scale without sacrificing intent or technical quality.
Let me be clear upfront. I didn't ask ChatGPT to "write me 500 blog posts."
That approach produces garbage. I've tried it. The content reads the same, ranks nowhere, and wastes indexing budget.
Instead, I designed a workflow where AI agents handled specific jobs: search intent analysis, page-level content generation, semantic variation across similar URLs, internal linking logic, schema markup, and publishing consistency. My role shifted from creator to architect.
That's the real unlock here. Humans define constraints and strategy. Agents handle volume and consistency.
This project leans heavily into programmatic SEO. But not the old "spin keywords into templates" version that Google buried years ago.
Each page targets a real user question:
The system varies cruise lines, destinations, payment structures, and user concerns like pricing, flexibility, and eligibility. The goal wasn't keyword stuffing. It was long-tail intent coverage at scale.
"Cruise now, pay later" sits at the intersection of travel and financing. High commercial intent. Real buyer demand. Not vanity traffic.
Internal linking is one of the most overlooked parts of SEO. It's also boring. Which makes it perfect for agents.
The system automatically:
Agents don't get tired. They don't get sloppy at page 487. That matters when you're managing hundreds of URLs and one broken link pattern can tank your crawl efficiency.
Beyond content, the site includes intent-aware schema markup generated programmatically across all pages. This includes FAQPage schema for common booking questions, Product and Offer signals for cruise packages, BreadcrumbList for crawl clarity, and Organization schema for entity signals.
Schema is repetitive, precise, and unforgiving. Miss a closing bracket on page 300? Good luck finding it manually. Agents handle this perfectly because they don't make typos at 2am.
When schema becomes part of the content system itself, it stops being an afterthought you bolt on before launch.
The entire site went from idea to live in a weekend.
That speed matters more than people realize. Instead of spending months planning a "perfect" SEO strategy, this approach allows for rapid indexing, fast demand validation, and early signal collection. You iterate based on real search behavior instead of assumptions.
In modern SEO, time-to-index often beats perfection. Getting 500 pages indexed and collecting data for two months beats spending two months planning 50 "perfect" pages.
AI agents didn't replace judgment. They amplified it.
Humans defined brand voice, compliance guardrails, content constraints, and what should never be generated. Agents handled volume, consistency, repetition, and structure enforcement.
Think less "AI writer" and more editor-in-chief of a 24/7 content system. You're not writing. You're designing what gets written and setting the rules for how it happens.
Cruises are just the case study. This same agentic SEO framework applies to travel and hospitality, fintech and lending, healthcare directories, marketplaces, and local services.
Anywhere structured demand meets long-tail queries, AI agents excel. The cruise site proves the system works. The system itself is the product.
I break down the full agent workflow, schema strategy, and system architecture in more detail on my AI Travel Project page.