AI Search Case Study 2026: How an Executive Community Got Recommended

Here is the work in one view. The numbers come from Pomo's audits of the same questions across the same four AI engines.

Pomo did the checking and the pointing. The team did the deciding and the writing. That split is why the work moved quickly without the team handing over its voice or its judgment.

• Pomo asked the AI engines the questions executives ask and recorded whether Open Future Forum was mentioned or cited
• Pomo reported which competitors and outside sources appeared in answers instead
• Pomo sent a daily email naming the pages that were cited or missed and the specific updates to make next
• The team chose which gaps mattered for the audiences it serves
• The team wrote and published the FAQ changes and the five articles
• The team kept the claims to what it could back with its own experience and research

Being known and being recommended are different problems. An accurate description helps someone who already knows your name. Discovery happens earlier, when a CEO asks for candid peer conversations, or a director asks for AI governance guidance, without naming anyone. You have to show up for the need, not just for your own name.

Pomo's September 17 report showed where Open Future Forum was and wasn't showing up for those needs. The samples were small, so the team used them to decide what to fix next rather than as a verdict on any single page.

Not every gap is worth closing. A gap matters when it overlaps with an audience you can actually serve. So the team ran each proposed article through three questions before writing it, which kept them from chasing every search opportunity.

Take private equity operating partners. An article for them has to cover what makes peer exchange worth their time: revenue growth, margins, working capital, management capability, technology and exit readiness. It also has to show when an operating partner needs other operating partners in the room, and when a CFO, CISO or CEO would be more useful.

• Who would ask this question?
• What decision are they trying to make?
• What evidence or experience can we contribute?

Much of the useful work was practical. Who qualifies? How do you apply? What does it cost, and how much time does it take? How are guests chosen? How is a private gathering different from an open event? The Open Future Forum FAQ now answers these, because an executive can't decide whether to take the next step without them.

The team also published a five-article series dated September 18. Three pieces were for specific audiences: CISO peer groups and roundtables, how general counsel choose a peer group, and a peer agenda for private equity operating partners. The other two were for anyone choosing a community: ten criteria for comparing executive peer groups, and what makes a peer group useful for decision quality.

The goal was to make the offering easy to evaluate. For a selective community, content that helps a reader decide they aren't a fit is still doing its job, so quality is judged by how relevant the people it attracts are, not by how many.

An audit tells you where you stand once. AI answers keep changing, so the harder part is keeping up. Each morning, Pomo's email covered three things: which tracked questions had changed, which competitors or sources had appeared in answers, and which Open Future Forum pages were cited or missed. It ended with a short list of specific updates, such as adding participation costs and selection criteria to the FAQ, or linking the CFO community pages to the forum's AI investment research.

The team stopped planning content quarterly. They read the email in a short daily check-in and assigned one or two updates. Both the FAQ changes and the five-article series came out of those recommendations.

The report found six outside domains cited in the sampled answers. The team didn't turn that list into a placement schedule. A site that AI engines cite often still has to reach the right readers and suit the story.

Instead, the two kinds of content got different jobs. Open Future Forum's own site states the facts: who the community serves, how participation works, which events are available and where the research lives. Outside publications put those facts in context, through stories about how executives, founders and private equity leaders approach AI decisions. Murray's view is that a participant, partner or writer describing their own experience is more convincing than the same company description repeated across several sites.

AI-search analysis can point to a missing explanation, but you still have to supply the evidence. Open Future Forum's research program studies how executives adopt, fund, buy and evaluate AI, so a CFO piece can cite findings on AI budgets and returns, and a security piece can cite research on agent access and governance.

Murray measures the work at three levels and keeps them separate. The move from 18% to 23% is progress at the discovery level. The community-growth level is still being measured.

• Content: are the relevant questions answered clearly and backed with evidence?
• Discovery: in a comparable follow-up audit, do the same questions produce more consistent mentions and citations? Note any change in prompts or engines so the comparison holds.
• Community growth: are those routes bringing relevant applications, suitable event participants and relationships that last past a first visit?

Nothing here is specific to executive communities. The same steps work for any brand whose buyers ask an AI assistant for a shortlist.

• Write down the questions your buyers ask before they know your name, and test them across the major AI engines
• Note where you are missing, and which competitors and sources show up instead
• Close only the gaps that match an audience you can serve, using the three questions above
• Answer the practical questions buyers need to evaluate you: who it's for, what it costs, how to start
• Review changes daily and make one or two updates at a time, rather than planning quarterly
• Re-run the same questions on the same engines, so the before-and-after comparison holds