Account-based GEO: how to reach every role on the buying committee in AI answers

Forrester's State of Business Buying 2026 reports that, on average, 13 internal stakeholders and nine external participants influence buying decisions. Each of them can ask an AI assistant a different question at a different time. An ABM program that coordinates ads, email and sales plays by role but ignores what those people read in AI answers leaves a large part of the account experience to chance.

Most GEO work starts from category prompts such as "best X software". That is useful and it is not enough for ABM. A procurement lead at a target manufacturer asks about contract terms and data residency; an engineer asks about integration with a specific system; a CFO asks about payback. Our article on how B2B buyers research vendors in ChatGPT summarizes the survey evidence for this behaviour.

Start from your ABM tiers and buying group map. For each target segment, list the roles you already plan plays for, then write prompts for each buying stage in the words those people use on sales calls. Semrush found that 61% of AI users describe their specific use case or problem and 45% include constraints such as budget, features or compatibility, so include constraints in the prompts.

Keep the set small enough to run every month: around five prompts per role for each priority segment is usually enough to see patterns. Record the exact wording so results are comparable over time.

Answers can only include facts that exist somewhere crawlable. When we audit enterprise sites, the facts the economic buyer and the security reviewer need are often missing or locked in PDFs behind forms. Google states that there are no additional requirements to appear in AI Overviews or AI Mode, but a page must be indexed and eligible to show with a snippet to be a supporting link. Plain HTML pages with direct answers remain the foundation.

Write each page for one role and one question. Lead with the answer, give the specific fact (the integration method, the certification, the pricing model), then the proof. Our buying committee content by role article lists the assets each seat needs; account-based GEO makes sure those assets are also findable and quotable by AI tools.

• One industry use-case page per tier 2 cluster, naming the constraints that cluster shares.
• Comparison pages for the competitors that appear in answers for your target segments.
• A trust page with current certifications, data residency and subprocessors.
• A pricing model page that explains cost drivers even if exact prices are scoped.
• Integration pages for the systems your target accounts run.

Take a tier 2 cluster of mid-size manufacturers that share a compliance requirement. The ABM plan already has an industry message, a LinkedIn campaign and a sales play. Account-based GEO adds four things: around 25 prompts across the five roles that matter for this cluster, a monthly run and score, a use-case page that answers the compliance question directly with the specific standard named, and a check of the two or three industry sources that the answers cite.

After the first run, the team usually finds one of three patterns. You are absent because the facts are missing; you are present but described with outdated facts; or a competitor is named because a third-party comparison favours it. Each pattern has a different fix: publish, correct or build source presence.

AI answers lean heavily on third-party sources. G2 reports that citations from software review sites are the top trust signal that increases buyers' confidence in an AI answer. For an ABM team, that makes review profiles, partner directories, analyst pages and industry publications part of the account experience, in the same way an event sponsorship or a LinkedIn campaign is.

In practice, record which sources are cited for each segment's prompts, then correct outdated facts, add missing profiles and prioritize the publications your target industries read. Our AI shortlist authority work covers this source alignment in detail.

1. Add account-based GEO to the ABM plan with the same tiers, roles and owners as the rest of the program.
2. Run the prompt matrix monthly and score presence, accuracy, position and cited sources for each segment and role.
3. Feed gaps into the content and source backlog, ordered by tier.
4. Brief sales on what AI tools say about you and competitors for their accounts, so they can address it early.
5. Report coverage by tier and role alongside engagement and pipeline, and review it in the monthly ABM meeting.

Measurement is the part teams skip. Our article on measuring AI visibility for target accounts shows how we combine prompt testing, AI referrals by account and self-reported attribution in one report.

Account-based GEO is not a new channel to staff separately. It is ABM thinking applied to the place buyers now do much of their research. Teams that already map buying groups have most of what they need: the roles, the questions and the accounts.

Our generative engine optimization service builds and runs the prompt matrix, the pages and the source work with your ABM team.

If you want to see what your tier 1 accounts are likely to read today, talk to our team.