How to Pitch GEO to Your Boss or CMO (With a 1-Page Brief)

#01Why do GEO pitches get turned down?

Most pitches fail for predictable reasons. They start with the technology rather than the business. They quote big market claims the leader has already seen debunked. They ask for a new budget line, a new hire and a new tool all at once. Or they promise results nobody can measure yet.

Senior leaders hear about new channels constantly. What earns their attention is a specific risk or opportunity in their own business, shown with evidence they can check, and a request small enough to say yes to without a long debate. Our view: your goal in the first pitch is not to win a programme. It's to win permission to learn, with a date when you'll report back.

#02What business problem should you lead with?

Pick the one your leader already worries about. If pipeline from organic search is softening, frame AI answers as a change in how buyers research. If competitors keep appearing in sales conversations, frame it as competitors being recommended where you aren't. If brand accuracy is a concern, show an AI answer that describes your product wrongly.

Use your company's own language. If the CMO talks about "share of consideration," talk about being in the shortlist an assistant gives. If they talk about pipeline, connect AI answers to the questions buyers ask before they book a demo. Don't introduce new jargon unless you must. "Generative engine optimisation" can come later; "making sure AI assistants recommend us accurately" works on day one.

#03What evidence should you bring?

Bring evidence you gathered yourself, because it's specific and hard to dismiss. Write ten to fifteen questions real buyers ask, using sales call notes, support tickets or reviews. Run each question three times in the two or three assistants your buyers most likely use. Record whether you're named, how you're described, which competitors appear and which sources are cited.

Summarise it in one simple table and two screenshots. One screenshot should show a competitor being recommended for a question you should win. The other, if you have it, should show an inaccurate description of your product. Those two images often do more than any slide.

Add one or two signals from your own analytics. Many analytics tools can show visits referred from AI assistants if you set up a channel or segment for them. Even small numbers are useful as a baseline. Don't extrapolate them into big forecasts.

#04What should the pilot look like?

Make it small, specific and time-boxed. A good pilot focuses on one product line or topic, a fixed set of prompts and a short list of changes, run over about ninety days. For example: improve five key pages so they answer buyer questions directly, fix inaccurate facts in major listings, publish one comparison page, and re-test the same prompts monthly.

Define success before you start. Pick two or three measures, such as how often you're named across the prompt set, how often descriptions are accurate and how many visits arrive from AI assistants. Agree what result would justify continuing, and what would mean stopping. Leaders like pilots that can end.

#05How do you ask for resources?

Ask for the smallest set that makes the pilot real. That's usually a share of your own time, a few hours a week of help from a writer or web developer, and possibly a modest tool budget for tracking. If you can run the pilot with a spreadsheet and manual prompt runs, say so; it lowers the bar to yes.

Be clear about what you'll stop doing to make room. Leaders worry about distraction more than cost. "I'll pause the monthly newsletter refresh for one quarter" is a trade they can weigh.

#06How should the one-page brief be structured?

Use six short sections:

1. The problem: one or two sentences in business terms.
2. What we found: the summary table and two screenshots.
3. Why it matters: the link to pipeline, brand accuracy or competitive position.
4. The pilot: scope, changes and timeline.
5. What it needs: time, people and any tool cost.
6. How we'll decide: measures, review date and the decision you'll bring back.

Keep it to one page. Put the full prompt list and raw results in an appendix for anyone who wants detail.

#07How do you deliver the pitch?

Send the brief a day ahead, then use the meeting for questions. Open with the screenshot of a competitor being recommended. Explain the pilot in two minutes. Then stop talking and ask what concerns they have.

Expect three questions. "Isn't this just SEO?" Answer that it builds on SEO but adds new checks, like how assistants describe you and which third-party sources they rely on. "How do we know it works?" Answer that the pilot is designed to find out, with agreed measures and an end date. "What does it cost?" Give the honest number, including your time.

#08What should you do after a yes or a no?

If you get a yes, confirm the scope, measures and review date in writing the same day. Send short monthly updates even if nothing dramatic has changed. Trust builds through consistency.

If you get a no, ask what would change their mind, and when you could revisit it. Keep running a small monthly test on your own time if you can. A second pitch with three months of data is much stronger than the first.

#09From MentionX data: where brands go missingOriginal data

Method. Read-only aggregate queries over MentionX scans from 22 August to 29 October 2026 (UTC): 2,217 engine runs (2,154 completed) across 18 anonymised projects and six engines.

MentionX's anonymised aggregates show that brand mentions fall sharply as questions move away from the brand name:

Our view: this is the slide most CMOs need. Being named on your own brand questions is expected; the opportunity is in recommendation and problem questions, where brands are named far less often.

Limits. A convenience sample of mostly B2B software projects, run through engine APIs, which can behave differently from the consumer apps. Perplexity, Copilot and DeepSeek figures come from a single day (26 October 2026), so treat them as a snapshot, not a trend.

#10Worked example: Ledgerline's one-page pitch

Ledgerline is a fictional accounts-payable tool. Figures are illustrative.

A content lead at Ledgerline, a fictional accounts-payable tool, notices sales reps hearing "the AI recommended Competitor A." She runs 12 buyer questions three times in three assistants. Ledgerline is named on brand questions but rarely on "how do I automate invoice approvals?" One answer lists an integration Ledgerline dropped two years ago.

Her one-page brief asks for 90 days, four hours a week of a writer's time and no new tools. She'll fix the outdated integration listing, rewrite four guides and publish one comparison page, then re-test monthly. The CMO says yes, on condition that she reports at day 45. She confirms by email that afternoon.

#11What can you do this week?

Collect ten buyer questions from sales or support, run them in two assistants and take screenshots. Draft the six-section brief. Show it to one trusted colleague before your leader, and cut anything they don't understand. For the cost side, see building a GEO business case.

#12How do you tailor the pitch to different leaders?

A CMO focused on brand will respond to accuracy and consistency: an assistant describing your product wrongly is a brand problem, not a technical one. A demand-generation leader will respond to the buyer journey: questions asked before a demo request, and whether you're in the shortlist. A CFO-minded leader will want the cost, the time commitment and the point at which you'll stop if results don't come. A CEO may care most about competitors, so lead with the screenshot where a rival is recommended.

You can use the same one-page brief for all of them, but change the first two lines. Put the concern that matters most to that person at the very top. If you're pitching a group, ask your sponsor beforehand which concern will decide the meeting.

#13Which claims should you leave out?

Leave out market-size claims you can't source, predictions about the end of search, and any figure from a vendor's marketing that you haven't checked. Leaders who have seen bold forecasts fail will discount everything else in your brief if one number looks inflated. If you want context, cite primary sources, such as Google's own guidance on AI features in Search, and say clearly what they do and don't tell you.

Also leave out promises about rankings or guaranteed mentions. Nobody controls how assistants answer. What you can promise is a disciplined test, clear reporting and a recommendation at the end, backed by evidence from your own market.

#14How do you build allies before the meeting?

Talk to sales and support first. They hear what buyers say about AI tools, and a quote from a sales rep often carries more weight with a CMO than any data you bring. Ask the web or SEO lead whether they'd support the pilot, and what they'd need. If legal or brand teams usually review content, give them a heads-up that you may propose page changes. Walking into the meeting with two colleagues already supportive changes the conversation from "should we?" to "how do we start?"

Finally, ask your sponsor how decisions are usually made. Some leaders decide in the meeting; others want a written follow-up. Match their habit and your pitch will feel easy to approve.

#15Common questions

How long should a GEO pitch be? One page, plus an appendix with prompts and raw results.

Should I ask for a new hire? Usually not at first. Ask for a small pilot that proves the need.

What if my boss says it's just SEO? Explain that it builds on SEO but adds checks on how assistants describe you and which sources they use.

How do I measure the pilot? Pick two or three measures before you start, such as mention rate, accuracy and AI referral visits.