AI visibility tracking is the habit of asking AI engines the questions your buyers ask, on a fixed schedule, and recording whether your brand is named, whether your pages are linked, and how you are described. It turns "does ChatGPT know us?" into a number you can watch move month by month.
📺 Watch: ChatGPT and Google AI Run Hidden Searches. Here’s How to Rank for All of Them
🔥 Get the Agent OS as a free bonus: AI Profit Boardroom members get the full Agent OS zip, prompt libraries, daily tutorials and weekly live coaching calls. → Get inside
This guide explains what it measures, how to set it up in an afternoon, why the answers change from one run to the next, and what Google Search Console and Bing Webmaster Tools can and can't show you. If you'd rather have it done for you, book a free SEO strategy session and Goldie Agency will walk through your current AI visibility with you.
Prefer to build the skill in-house? Inside the AI Profit Boardroom you can learn AI SEO yourself, with step-by-step trainings and four live coaching calls a week.
What AI visibility tracking actually is
Classic rank tracking checks where a URL sits for a keyword. AI answers have no fixed order, so that model breaks.
Instead, you track presence. You pick a set of prompts, run them through each engine, and log what comes back. Over time the log tells you which questions you win, which ones a competitor owns, and which engines ignore you.
Two kinds of data feed it: prompts you or a tool run yourselves, and first-party reports the engines publish about your site. Use both.
The six things AI visibility tracking measures
- Mentions. Your brand name appears in the answer text, linked or not.
- Citations. A link to one of your pages is shown as a source. Peec AI's site makes a useful point here: your content can be cited even when your brand isn't named.
- Share of voice. Your mentions divided by all brand mentions across the same prompt set. Ten prompts, thirty brand mentions, six of them yours, gives you 20%.
- Sentiment. Whether the answer frames you as the safe pick, a budget option, or with a warning attached.
- Branded versus unbranded prompts. "Is Acme Plumbing any good?" tests reputation. "Best emergency plumber in Leeds" tests whether you get found at all. Track them in separate columns.
- Per-engine results. ChatGPT, Gemini, Perplexity, Claude, Copilot and Google's AI Overviews and AI Mode each build answers their own way, so one combined score hides the story.
How to set up AI visibility tracking in five steps
- Write 20 to 30 prompts. Pull them from sales calls, support emails and Search Console queries. Make about two thirds unbranded.
- Pick three to five competitors. Use the names your prospects actually compare you with, not the biggest brand in the industry.
- Choose the engines. Start with the ones your buyers use. For most small businesses that means ChatGPT, Google AI Overviews and AI Mode, Gemini and Perplexity.
- Set a cadence. Monthly is enough for a manual sheet. Tools can run daily or weekly; Peec AI, for example, offers daily, weekly or monthly prompt tracking.
- Fix the log format. One row per prompt per engine per run, with columns for date, mentioned, cited URL, competitors named and a one-line sentiment note.
Save the first sweep as your baseline and judge every change against it.
Want a second pair of eyes on your prompt list? Grab a free strategy session and we'll check whether it mirrors how your buyers really search.
Why the same prompt gives different answers
Run one question five times and you can get five slightly different answers. That's how these systems work.
Google says AI Overviews and AI Mode may use a "query fan-out" technique, firing off several related searches across subtopics before writing a response. It also says the two features may use different models and techniques, so their answers and links will vary. In my video on this page I break down those hidden searches and how to show up for them.
Location, chat history and fresh content shift results too. Bing says citation activity can move with evolving AI models, competing content, freshness and changes in demand.
The fix is to read trends, not single runs. Run each prompt more than once, average the results, and only act when a pattern holds for two or three cycles.
What Search Console and Bing can and can't show
The free first-party reports are the only data that comes straight from the engines about your site.
| Source | What it shows | What it can't show |
|---|---|---|
| Search Console Generative AI performance report | Impressions when links to your site appear in AI Overviews and AI Mode, by page, country, date and device | No query or click columns are listed; nothing about ChatGPT, Perplexity or Claude; Search Labs experiments excluded |
| Bing AI Performance | Citations across Copilot, AI summaries in Bing and select partners; cited pages; a sample of grounding queries; Citation Share | Bing says the counts don't indicate ranking or placement, and Citation Share doesn't reveal competitor domains |
| Your own prompt log or a tracking tool | Mentions, citations, sentiment and competitors for the exact prompts you choose, in every engine | It's a sample, and no third-party tool sees Google's internal data |
Google rolled the Generative AI report out to every website worldwide as of 31 August 2026. If you can't see it, Google says your site likely hasn't had enough AI impressions yet, or it has been excluded from Search generative AI features.
Bing's June 2026 update added Intents, Topics, Citation Share and Compare. Citation Share is your share of all citations shown for the same grounding query, which is the closest thing to an official share-of-voice number any engine publishes.
Tools that automate the tracking
A spreadsheet works for 20 prompts. Past that, a tool saves hours. Five options, described by their own sites:
- Ahrefs Brand Radar runs prompts through AI Overviews, AI Mode, ChatGPT, Perplexity, Gemini and Copilot, with share of voice and sentiment.
- Semrush AI Visibility Toolkit offers prompt tracking on Google AI Mode and ChatGPT plus brand performance reports.
- Otterly.AI tracks mentions and citations across seven engines and tags prompts as branded or non-branded.
- Peec AI lets you filter visibility by model, country and prompt tag.
- SE Ranking's AI Visibility Tracker covers AI Overviews, AI Mode, ChatGPT and Gemini, with competitor comparison.
This page explains the method. For a ranked comparison, read our list of the best AI visibility tools.
Whatever you pick, keep Google's warning in mind. Its guide to generative AI features says no third-party tool has access to its internal ranking or AI systems, so treat any tool that claims otherwise with suspicion.
Turning the numbers into fixes
Read each monthly sweep with three questions.
- Which unbranded prompts name a competitor but not you? Look at the pages the engine cites for them. That is your content gap.
- Which engines never cite you at all? Check robots.txt first. OpenAI says sites that block OAI-SearchBot won't be shown in ChatGPT search answers, and Perplexity says PerplexityBot is what surfaces and links sites in its results.
- Where is the sentiment off? Wrong facts usually trace back to an old directory listing or review page. Fix the source and the answers follow.
If tracking shows a deeper problem, run a full AI visibility audit to check crawl access, entity facts and citations end to end.
Want us to read your first month of data with you? Book a free strategy session with Goldie Agency and we'll point to the prompts worth fighting for.
You can also hand the weekly sweep to an agent. The Agent OS can run the same prompt list, fill the log and flag changes, and Goldie Bench shows which AI models handle that kind of structured work well.
What we track on our own sites
We test on our own properties first. aiprofitboardroom.com went from 328 Google clicks a month in October 2025 to 22,443 in October 2026, per its own Search Console. That is Julian's own published number, and it's the reason we watch the Generative AI report alongside every prompt sweep.
Want the prompts we start every sweep from? Grab the free 200+ AI SEO prompts, and the free link building book for the mentions side of the work.
My verdict: start free, then add a tool
Switch on the Search Console Generative AI report and Bing AI Performance today, because they cost nothing and come from the engines themselves. Then add a 20-prompt manual log for ChatGPT, Gemini and Perplexity. Only buy a tool once you outgrow the sheet. Nobody can guarantee you a mention, but you can always measure where you stand.
Ready to see where you stand in AI answers right now? Book a free SEO strategy session.
FAQ
Is AI visibility tracking the same as rank tracking?
No. Rank tracking records a position on a results page. AI answers have no fixed positions, so you record presence instead: mentioned or not, cited or not, and how you're described.
How often should I run my prompts?
Monthly works for a manual sheet. Weekly or daily runs through a tool make sense once you have dozens of prompts or several competitors to watch.
Can Search Console show my ChatGPT visibility?
No. The Generative AI performance report covers Google's AI Overviews and AI Mode only. For ChatGPT, Perplexity and Claude you need your own prompt runs or a third-party tool.
Why does my tracking tool disagree with what I see?
Both are samples. Location, account history and the day you run a prompt all change the answer, which is why trends across several runs matter more than one result.
Still unsure which numbers matter for your business? Claim a free strategy session, or join the AI Profit Boardroom to learn AI SEO yourself. Either way, talk to Goldie Agency or book a free SEO strategy session and start AI visibility tracking with a plan.











