AI Visibility Report: How to Show Clients Where They Appear in AI Answers

Every client has now heard that people "search in AI", and sooner or later each one asks the same thing: "Do we show up when someone asks ChatGPT about what we sell?" An AI visibility report is how an agency answers that with evidence instead of a screenshot of one lucky answer. Done well, it shows how often the brand is named, how often the site is cited as a source, and whether those mentions turn into visits.

Done badly, it becomes a vanity chart that falls apart the first time the client types the same question and gets a different answer. This guide covers what the section should measure, how to pick the prompts behind it, how to explain the difference between mentions and clicks, and the mistakes that make AI numbers look unreliable in front of a client.

What does an AI visibility report measure?

It measures how often AI answer engines name or cite a client when people ask questions in the client's category. The core unit is a tracked prompt: a question you run through each engine on a schedule, recording whether the brand appears in the answer and whether the client's pages are listed as sources. Everything else in the section is a summary of those results.

Mentions and citations are worth separating. A mention means the engine recommends or names the brand. A citation means it used the client's page as evidence, which is the part that can send a click. A brand can be mentioned without being cited (the engine learned about it elsewhere), and a page can be cited for a fact without the brand being recommended. Both matter, and the gap between them is often the most useful finding in the whole client SEO report.

Competitor context helps too. If you record which other brands appear for the same prompts, the client can see who the engines recommend instead of them, which turns an abstract score into a concrete gap to close.

How do you choose the prompts to track for a client?

Choose prompts the way a real buyer would phrase them, grouped by the services the client actually wants to win. A prompt list built from the agency's favourite keywords will flatter the client; one built from customer questions will tell the truth. Work through it in order:

1. List the money services. Start with the three to five products or services that drive revenue, not every page on the site.
2. Write buyer questions for each. Use the client's sales calls, support tickets and Search Console queries. Phrase them as full questions, such as "which accountant in Leeds is best for contractors".
3. Add comparison and "best" prompts. These are where engines name brands most often, and where competitors show up.
4. Include a few brand prompts. Questions that name the client check whether the engines describe it accurately.
5. Add location where it matters. For local businesses, put the city or region in the prompt, the same way you would handle local keyword ranking checks.
6. Freeze the core set. Keep most prompts the same every period, and add new ones in a clearly labeled extra group.

Share the list with the client before the first report. It takes ten minutes, it surfaces prompts the agency missed, and it removes the most common objection later: "but nobody asks that".

How should you read and present the results?

Present mentions and clicks as two halves of one story. Mention data shows whether the brand is in the answer; analytics data shows whether anyone clicked through. Neither is complete on its own, and clients draw the wrong conclusion when they see only one.

The click side lives in GA4. Since Google added an AI Assistant channel to the GA4 default channel group, visits from assistants such as ChatGPT, Gemini, DeepSeek, Copilot and Grok have their own row. The same definition excludes Google AI Overviews and AI Mode, so those clicks still count as organic search.

Google's guide to AI features and your website confirms that appearances in AI Overviews and AI Mode are included in the Performance report in Search Console, within the Web search type, as part of overall search traffic. So for Google's AI answers, the visibility data is your only direct signal of presence, and the clicks are mixed into ordinary search traffic.

In practice, structure the section in three layers. First, a headline: share of tracked prompts where the brand was mentioned, and how that moved. Second, the engine split and the prompts where the brand is missing, which become next quarter's content work. Third, AI assistant sessions, engagement and conversions from GA4, so the client sees that mentions lead somewhere.

Keep paid results in their own section, too. In its September 2026 expansion announcement, OpenAI states that ads in ChatGPT are labeled, separate from the answers and do not influence them, so a client who also runs ChatGPT ads should never see that spend credited to organic visibility. If both appear in the same report, label them clearly as paid placement and earned mentions.

Common AI visibility report mistakes

Tracking a tiny prompt set. Five prompts means each answer moves the score by 20 points. One changed answer looks like a collapse or a breakthrough, and the client loses trust in the number. A broader set smooths the noise and makes trends meaningful.

Treating one answer as a ranking. AI answers are generated fresh each time. Google itself says AI Overviews and AI Mode use different models and techniques, so the responses and links they show will vary, and AI Overviews often do not appear at all. Talk about how often the brand appears, not where it "ranks".

Ignoring variability between runs. If the client checks a prompt on their phone and gets a different answer, that is normal, not a reporting error. Say so in the first report, and show results over time rather than from a single check.

Changing the prompts every month. A new prompt set resets the baseline. Freeze the core list and report additions separately.

Claiming engines you do not track. If the data covers five engines, say which five. Promising coverage of every AI tool invites the question you cannot answer.

How does AgencyReportr show AI visibility?

AgencyReportr, which is marketing reporting software for agencies, adds AI visibility to the SEO report: brand mentions and citations in ChatGPT, Gemini, Perplexity, Claude and Google AI Overviews. It does not cover Google AI Mode or Grok, and the report does not pretend otherwise. The same SEO report carries an AI assistant traffic section from GA4 with AI sessions, engagement rate, AI conversions, conversion rate and a source and medium table, so mentions and clicks sit on the same page. You can see the module on the AI mention tracking page.

Everything appears under the agency's logo, colors and custom domain as a fully branded client dashboard, and each report opens with an AI-written summary of the period. If clients prefer a document, weekly or monthly PDF emails deliver the same AI visibility report on a fixed schedule, next to Search Console, rankings and backlinks, so AI search becomes a standing section rather than a one-off slide.

AI Visibility Report FAQ

What should an AI visibility report include?

At minimum: the list of prompts tracked, how often the brand is mentioned in the answers, how often the client's site is cited as a source, the split by AI engine, and the trend against the previous period. Pair it with AI assistant traffic from GA4 so the client sees clicks as well as mentions.

How many prompts should I track for a client?

Enough to cover the client's main services, buyer questions and comparisons, usually a few dozen rather than a handful. A set of five prompts swings wildly from one run to the next, while a broader set shows a trend you can defend in a report. Keep the core set fixed so periods stay comparable.

Is being mentioned in ChatGPT the same as ranking first?

No. AI answers are generated each time and can change between runs, users and days. Google says AI Overviews and AI Mode use different models and techniques, so the responses and links they show vary. Report mention rates across many prompts and runs, not a single position.

Where do clicks from AI answers show up in Google Analytics?

Visits from assistants such as ChatGPT, Gemini, DeepSeek, Copilot and Grok fall into GA4's AI Assistant channel. Clicks from Google AI Overviews and AI Mode are excluded from that channel and count as organic search, and Search Console includes them in the Web search type of the Performance report.

Can buying ChatGPT ads improve a brand's AI visibility?

No. OpenAI states that ads in ChatGPT are labeled, kept separate from the answers, and do not influence what ChatGPT says. Organic mentions and citations come from the content and reputation the AI systems can find, so they need to be reported separately from any paid results.