Key takeaways
- The average B2B buying committee has 13 people, each with different priorities -- and each one prompts AI search engines differently about the same vendor or category.
- Fan-out analysis maps how a single core topic branches into persona-specific sub-queries, revealing content gaps that block consensus at every stakeholder layer.
- Most B2B content strategies are built around one persona (usually the economic buyer), leaving IT, legal, finance, and end-users completely unserved in AI search results.
- Each branch of a fan-out needs its own content treatment -- not a different version of the same page, but genuinely different angles, formats, and evidence types.
- AI visibility platforms that track prompt-level data can show you exactly which persona branches your competitors are winning and you're not.
The problem with targeting "the decision-maker"
For years, B2B demand gen operated on a simple assumption: find the person who signs the check, convince them, done. Campaigns were built around one ICP persona. Content was written for one job title. The funnel had one throat to choke.
That model is broken -- and has been for a while. According to research from Influ2, the average B2B buying group now has 13 people. A piece from Demand Gen Report put it plainly in early 2026: "Most B2B marketers still target one persona at a time -- even though buying committees include dozens of stakeholders."
The gap between how we market and how buying actually works has always been uncomfortable. But in 2026, it's become a concrete competitive disadvantage -- because AI search has made it visible.
When a VP of Engineering at a target account opens ChatGPT and types "does [your product] integrate with our existing data stack," they get an answer. When the CFO asks "what's the typical ROI timeline for [your category]," they get an answer. When a security analyst asks "what are the data residency options for [your product]," they get an answer.
If your content doesn't address those specific questions, your competitors' content will. And the AI will cite them, not you.
This is where fan-out analysis becomes one of the most practically useful frameworks in modern B2B content strategy.
What fan-out analysis actually means
Fan-out analysis starts with a single seed prompt -- the core question a buyer might ask about your category or product -- and maps how it branches into persona-specific sub-queries.
Think of it like a tree. The trunk is something like "best [category] software for enterprise." That's the generic query. But the moment you add persona context, the tree fans out:
- The CFO branch: "what does [category] software cost for a 500-person company," "ROI benchmarks for [category] tools," "how long does implementation take"
- The IT/Engineering branch: "does [product] have SSO support," "[product] API documentation," "[product] SOC 2 compliance status"
- The end-user branch: "how easy is [product] to learn," "[product] vs [competitor] for daily use," "does [product] have a mobile app"
- The procurement branch: "[product] contract terms," "[product] data processing agreement," "can we negotiate [product] pricing"
- The champion/internal advocate branch: "how to build a business case for [product]," "[product] case studies in [industry]," "how does [product] compare to [incumbent]"
Each of these branches represents a real person in the buying committee, asking real questions in AI search engines. And each branch needs content that speaks to that person's specific concerns -- not a watered-down version of your main product page.
Why AI search makes this more urgent than it used to be
Traditional SEO had a natural forcing function: you'd rank for a keyword or you wouldn't. The feedback loop was slow but legible.
AI search is different. When someone asks ChatGPT or Perplexity a question, the model synthesizes an answer from multiple sources and presents it as a coherent response. The sources it cites are the ones that most directly addressed the specific question asked.
This means persona-specific queries are now high-stakes. If your site has excellent content for the economic buyer but nothing for the IT evaluator, you'll get cited in CFO conversations and completely ignored in engineering conversations. The buying committee will have an uneven picture of your brand -- strong in some lanes, invisible in others.
That unevenness kills deals. As a LinkedIn analysis of B2B buying trends in 2026 noted, "every additional participant introduces new questions, different priorities, and another layer of evaluation before approval can happen." If any one of those layers hits a wall -- can't find the information they need, or finds it only from a competitor -- the deal stalls.
Mapping the fan-out: a practical framework
Here's how to actually do this for your product or category.
Step 1: Define your seed prompts
Start with 5-10 core prompts that represent how someone would discover your category. These are typically broad: "best [category] tools," "how to solve [problem]," "[category] software comparison."
These are your trunk queries. They're important, but they're also the most competitive and the least persona-specific.
Step 2: Identify your buying committee layers
For most B2B products, the committee breaks into roughly four layers:
- Economic buyers (CFO, VP, C-suite): care about cost, ROI, risk, strategic fit
- Technical evaluators (IT, Engineering, Security): care about architecture, integrations, compliance, implementation complexity
- End users (the people who'll actually use the product daily): care about usability, workflow fit, learning curve
- Gatekeepers (Legal, Procurement, Compliance): care about contracts, data handling, vendor risk
Depending on your product, you might also have a champion layer -- the internal advocate who's trying to build consensus and needs ammunition.
Step 3: Generate persona-specific branches
For each committee layer, brainstorm the specific questions that persona would ask. Don't guess -- use real data where you can. Talk to your sales team about the objections and questions that come up in each stakeholder conversation. Look at support tickets. Review the questions that come up in sales calls with different job titles.
Then map those questions to AI search prompts. The question "does this integrate with Salesforce" becomes the prompt "does [product] integrate with Salesforce" or "[product] Salesforce integration review."
Step 4: Audit your existing content against each branch
This is where most teams discover the problem. Take your content inventory and map each piece to a persona branch. You'll almost certainly find that 70-80% of your content addresses the economic buyer, with thin or nonexistent coverage of technical evaluators, end users, and gatekeepers.
The gaps aren't random. They reflect how your content team thinks about the buyer -- and it's usually shaped by whoever has the most influence in marketing meetings.
What content each branch actually needs
Understanding the branches is half the job. The other half is knowing what content format and evidence type actually works for each persona.
| Persona | Core concern | Content that works | Evidence type |
|---|---|---|---|
| CFO / Economic buyer | ROI, cost, risk | Business case templates, ROI calculators, TCO comparisons | Customer ROI data, analyst quotes, payback period examples |
| VP / Strategic buyer | Strategic fit, vendor stability | Category vision content, roadmap transparency, executive case studies | Peer company examples, analyst positioning |
| IT / Engineering | Technical fit, security, integrations | Technical documentation, integration guides, security whitepapers | Compliance certifications, architecture diagrams, API docs |
| Security / Compliance | Data handling, vendor risk | Data processing agreements, security FAQs, compliance matrices | SOC 2 reports, GDPR documentation, pen test summaries |
| End user | Usability, workflow fit | Product walkthroughs, comparison guides, "day in the life" content | User reviews, G2/Capterra ratings, tutorial videos |
| Procurement / Legal | Contract terms, pricing flexibility | Pricing transparency pages, standard contract FAQs | Reference customers, standard DPA templates |
| Internal champion | Building consensus, handling objections | Business case builders, internal pitch decks, competitive comparisons | Win/loss data, customer stories by use case |
The critical insight here: these aren't just different tones on the same content. They're genuinely different information. A CFO doesn't need to know about your API rate limits. A security analyst doesn't need your ROI calculator. Trying to write one piece that serves all of them produces content that serves none of them well -- and AI models are good at recognizing when content is actually answering a specific question versus hedging across multiple audiences.
How to prioritize which branches to build first
You can't build everything at once. Here's a prioritization framework:
First, identify which branches are actively costing you deals. Talk to your sales team: where do deals stall? Which stakeholder conversations go quiet? Which objections come up repeatedly that you don't have good answers for? Those are your highest-priority gaps.
Second, look at what your competitors have. If a competitor has a detailed security whitepaper and you don't, every IT evaluator who asks AI about your category's security posture will get cited to their content, not yours.
Third, consider the query volume for each branch. Some persona-specific queries are asked far more often than others. Technical integration questions tend to be high-volume because they're specific and searchable. Legal/compliance questions tend to be lower volume but higher stakes -- a single unanswered compliance question can kill a deal.
Platforms like Promptwatch can surface prompt-level data showing which specific questions are being asked in AI search engines, how often, and which competitors are getting cited for them. That kind of data turns the prioritization exercise from guesswork into something you can actually defend in a planning meeting.
The content gap problem in practice
Here's a concrete example of how this plays out. Say you sell a data analytics platform. Your marketing team has built solid content around "business intelligence for revenue teams" -- good ROI stories, executive case studies, comparison guides against the top two competitors.
A CFO at a target account searches for your product and gets a strong result. They're interested. They loop in their data engineering team to evaluate the technical fit.
The data engineer asks: "does [your product] support dbt integration," "[your product] Snowflake connector performance," "[your product] custom SQL support."
If you don't have content that specifically addresses those questions, the AI will either say it doesn't know or cite a competitor who does have that content. The engineer reports back that they couldn't find clear answers. The deal slows down. The CFO, who was already sold, now has to manage internal skepticism.
This is the fan-out problem in practice. The trunk query went well. The branch queries failed. And because buying committees require consensus, one failed branch can block the whole deal.
Building a fan-out content calendar
Once you've mapped your branches and identified the gaps, the execution question is how to build content systematically without overwhelming your team.
A few principles that work:
One branch per sprint. Pick one persona layer and go deep for a month. Build the 4-5 pieces that fully address that persona's core questions. Then move to the next branch. This is more effective than trying to do shallow coverage across all branches simultaneously.
Repurpose strategically, not lazily. A technical integration guide can be repurposed into a shorter FAQ for the security team, a one-pager for procurement, and a talking point sheet for the internal champion. The underlying information is the same; the framing and format change.
Write for the AI citation, not just the human reader. AI models cite content that directly answers a specific question. That means your technical FAQ needs to actually answer "does [product] support SSO" with a clear yes/no plus context -- not bury the answer in three paragraphs of marketing copy.
Use real data from your sales process. The best fan-out content comes from actual questions your prospects ask. Your sales team is sitting on a goldmine of persona-specific questions. A monthly sync between content and sales to capture the questions that came up in recent deals is one of the highest-ROI activities a B2B content team can do.
Tracking whether your branches are working
Content without measurement is just publishing. Once you've built out persona-specific branches, you need to know whether AI models are actually citing them when those persona-specific queries are asked.
This is harder than traditional SEO tracking because you're not just monitoring keyword rankings -- you're monitoring whether specific AI models cite specific pages in response to specific prompts. That requires actually running those prompts and analyzing the responses.
The practical approach: build a prompt library that mirrors your fan-out map. For each persona branch, maintain a set of 10-15 representative prompts. Run them regularly across the AI models your buyers use (ChatGPT, Perplexity, Google AI Mode, Claude). Track which pages get cited, which competitors appear, and where you're still invisible.
Over time, you'll see which branches are gaining traction and which still need work. You'll also see when competitors publish new content that starts winning citations in branches you thought you owned.
The consensus problem is a content problem
The research is consistent on this: B2B buying in 2026 is about building consensus across a committee, not convincing a single decision-maker. Every stakeholder who can't find the information they need is a potential deal-blocker.
Fan-out analysis is the framework that makes this concrete. It turns "we need to address the full buying committee" from a vague aspiration into a specific content map with clear gaps and clear priorities.
The teams that will win in AI search over the next few years aren't the ones who write the most content or optimize the most aggressively for generic keywords. They're the ones who understand that every person in a buying committee is asking different questions -- and who build content that answers each branch clearly enough that AI models cite it when those questions come up.
Start with your biggest deal-blocking gap. Map the branch. Build the content. Track the citations. Then move to the next branch.