"Do We Show Up in ChatGPT?" Here's How to Actually Answer That

A stat from a recent AgencyAnalytics survey should get every SEO’s attention: 66% of clients are now asking their agency for help with visibility in AI search tools like ChatGPT and Gemini. 64% of agencies say they’re worried about Google AI Overviews specifically. But here’s the gap — 48% of agencies admit they can’t actually track traffic from AI search or attribute conversions to it.

That’s the real problem right now. It’s not that nobody cares about AI visibility. It’s that most of the reporting stack wasn’t built to measure it, and clients are asking questions that clickthrough-rate dashboards can’t answer.

Why “are we in ChatGPT” is a harder question than it sounds

When a client asks if they show up in Google, you pull rank tracking and call it a day. When they ask if they show up in ChatGPT, there’s no equivalent single number. You’re dealing with:

• Presence — does the model mention the brand at all, for relevant queries?
• Position in the answer — is it the first source cited, or buried in a list?
• Sentiment — is the brand described accurately and favorably, or is the model working off outdated or wrong information?
• Citation pattern — which pages on the site are actually getting pulled in as sources, and why those and not others?

None of this maps cleanly onto a rank position. A brand can be “visible” in AI search and still be represented badly, or cited from a thin page instead of the one you’d want representing the company.

What to actually build into the stack

If a client is going to keep asking this question, the answer can’t be a one-off manual check in ChatGPT. It needs to be a recurring part of reporting:

Monitor AI Overviews presence and citations. Track which queries trigger an AI Overview, whether the brand is cited, and which URL gets pulled. This is the same discipline as rank tracking, just pointed at a different SERP feature.

Audit content for citability, not just rankability. Pages that get cited in AI answers tend to share a structure — direct answers near the top, clear definitions, FAQ and how-to formatting, and content that doesn’t bury the point under three paragraphs of preamble. If a page requires the reader to infer the answer, a language model will often skip it in favor of a competitor’s page that states it plainly.

Build content specifically for direct-answer capture. This means writing actual Q&A sections and definitions, not just optimizing existing paragraphs. A page that answers “what is X” or “how does Y work” in one clear sentence, followed by supporting detail, is far more likely to get lifted into an AI response than a page that treats the answer as a conclusion to build toward.

Track sentiment and context, not just mentions. Presence isn’t the goal — accurate, favorable presence is. If a model is citing outdated pricing, a discontinued feature, or an old company description, that’s worse than not being cited at all, because it’s misinformation the client didn’t sign off on and can’t easily correct.

Set up analytics for multi-session journeys. AI-driven discovery doesn’t behave like a single click-through. Someone asks ChatGPT a question, gets a brand mention, and shows up on the site two days later via a direct or branded search with no referral data connecting the dots. Standard last-click attribution will undercount this channel badly, and clients need to know that before they conclude AI search “isn’t driving anything.”

Setting expectations with clients now

The honest answer to “do we show up in ChatGPT” right now is usually some version of “partially, and here’s what we’re doing to track and improve it” — not a clean yes or no. Clients asking this question are really asking whether their agency is paying attention to where search behavior is heading. Agencies that can show a monitoring process, even an imperfect one, are in a much stronger position than ones that shrug and say it’s too early to measure.

It’s not too early. It’s early enough that most competitors haven’t built this into their reporting yet, which is exactly why it’s worth doing now. If you want a clearer picture of how your site is showing up — or not — in AI-generated answers, our GEO audit looks at citation patterns, content structure, and sentiment across the AI search surfaces your clients are already asking about.