AI visibility tracking

AI visibility tracking that tells you what to do next.

AI visibility tracking measures how often answer engines name or cite your brand when buyers ask about your market, against the brands they pick instead. Vercatus tracks it across 10 engines, then shows why they choose your competitors and what to change.

Get started See the engines

In short

Track your market's questions grouped into topics, not a short list of invented prompts.
Count named and cited apart, per engine, with a range. A share without a range cannot tell a real move from a lucky week.
Ask many different questions rather than the same few again: a repeated question shared 68% of its brands with the first answer in our tests.
Read engines as people use them. ChatGPT's app and its API named only 30% of the same brands and cited 4% of the same sites.
A count shows where you stand. It cannot say why, or what to change. That is the part most trackers leave to you.

What AI visibility tracking measures

When a buyer asks ChatGPT for the best payroll software for a small business, the answer names a few brands and links a few pages. AI visibility tracking records that, for every question your market asks and every engine that answers, so you can see how often you are in the answer and who is there instead.

Four measures do most of the work. Keep them apart; adding them together hides the difference that matters.

Measure	What it tells you	What it cannot tell you
Named	The answer mentions your brand, with or without a link.	Whether the engine trusts your pages, or only knows your name.
Cited	The answer links one of your pages as a source.	Whether the reader saw the link: most citations get no click.
Top pick	You are the first or only recommendation.	How stable that is: top picks change more than mentions.
Share against competitors	Your share of answers next to each rival's, on one scale.	Why the leader leads. That needs the pages and searches behind the answer.

Vercatus calls the last one Answer Share. Answer Share is the share of AI answers on your topics that name or cite you, shown with its range and next to your competitors on the same scale. It is the number to report upward, because it compares you with the brands you actually lose to, topic by topic. Other tools fold these measures into one AI visibility score; read one with care.

Six things that decide whether the numbers mean anything

Every tracker produces a chart. Whether the chart means anything depends on how the answers were collected. These six checks come from our own experiments of 1 October 2026: 252 answers with web search on, across three markets and six engines, with every cited page fetched and read. The method and the full results are published.

1. The questions come from your market

Invented prompts measure invented demand. Start from what buyers ask: category questions ("best CRM for a ten-person agency"), comparisons ("HubSpot vs Pipedrive"), alternatives, integrations and pricing. In our run, 45% of the searches engines ran contained "best", 11% "vs" and 10% "alternative". Those are the shortlist questions, and the ones to track first.

2. Variety beats repetition: many different questions, not the same few again

Asking one question five times feels thorough. It is not: two answers to the same question shared 68% of their brands, while answers to different questions on the same topic shared 21%. For the same 100 asks, 20 questions repeated 5 times gave about 26 independent answers; 100 different questions gave 100. Ask a tracker how many distinct questions sit behind each number.

3. The engine is read the way people use it

Many tools query a model's API because it is cheap. Buyers use the app. Over the same questions, ChatGPT's app and its API named 30% of the same brands and cited 4% of the same sites. A share measured on the API describes an engine your buyers never see.

4. Web search is on

With search off, an engine answers from what it learned in training; with search on, it reads today's pages. They recommend different brands: ChatGPT's two modes overlapped by only 32%. Track the mode your buyers use, which for shortlist questions is search.

5. Every share carries its range

Answers vary from one ask to the next, so a share from a few dozen answers can move ten points without anything changing. A tracker should show how sure each figure is. If it cannot, treat week-to-week moves as noise until they last.

6. The cited pages are recorded, not just the brands

The pages an engine reads are where its picks come from. A brand named on six or more of the pages other answers cited was named in 34.9% of answers, against 11.9% for a brand on none of them. A tracker that keeps the cited pages tells you which reviews, lists and forums to be on; one that keeps only brand counts does not.

Which engines to track

Track the engines your buyers use, read the way they use them. Vercatus reads up to 10, depending on the plan:

ChatGPT: the engine's own web search.
Google AI Overviews: Google's live results page.
Perplexity: the engine's own web search.
Gemini: its app, as people see it.
Claude: the engine's own web search.
Grok: the engine's own web search.
Le Chat: the engine's own web search.
You.com: the engine's own web search.
Google AI Mode: Google's live AI Mode results.
Copilot Search (Bing): Bing's live results page.

For most B2B companies, start with ChatGPT, Google AI Overviews, AI Mode and Perplexity, then add Gemini and Copilot Search. Google's two surfaces differ: AI Overviews sit on the results page, while AI Mode runs many searches of its own before it answers. Copilot answers from Bing's index, so a page Bing has not indexed will not appear there however well it ranks on Google. How to track your brand in AI Overviews and AI Mode covers Google's two surfaces, including what Search Console shows for free.

How Vercatus tracks AI visibility

Topics from your market. Vercatus groups the questions people ask about your category into topics, each with an estimate of how often it is asked of AI.
Answers through the day. Every topic's questions are asked of every engine in hourly batches, with web search on, from the engine's own app or live results page.
Named, cited and top pick, per engine. Each answer is read for the brands it names, the pages it cites and its first recommendation.
Answer Share with its range. Your share sits next to each competitor's on one scale, by topic and by engine, with the cited domains behind it.

Everything is kept as history, so when you change a page you can see what happened to the answers afterwards.

Where tracking stops

A tracker answers one question: how visible am I? The next three are the ones that decide what your team does on Monday.

Why does AI pick them? Vercatus compares the leader with you on what the engine remembers, the pages it reads and the searches it runs. How the three lenses work.
What should I change? Recommendations: the exact change, on your site or off it, ranked by the gain expected on your answers, with a draft on request. How recommendations work.
Did it work? Every shipped recommendation is judged against topics you left alone. Every recommendation you ship gets a verdict within 45 days. How verdicts are read.

If you only need a count of mentions for a report, a lighter tracker may be all you need, and the comparison says which kind fits which job. If you have to show that the work changed something, tracking is where it starts, not where it ends.

How to start

Create an account and add your domain. Vercatus runs a free audit on ChatGPT, Google AI Overviews, Google AI Mode and Perplexity: your share of their answers on your topics, the brands they name instead and the sites they cite. Tracking over time, recommendations and verdicts start with a plan. Before anything else, the free check shows in a minute whether AI crawlers can read your site at all.

Questions about AI visibility tracking.

Anything else, ask. We reply within one business day.

Ask a question

What is an AI visibility tracker?

Software that asks answer engines such as ChatGPT, Perplexity and Google AI Overviews the questions your buyers ask, records which brands each answer names and which pages it cites, and reports how often you appear against your competitors over time.

How often should AI visibility be tracked?

Continuously, and read weekly. Answers change from one ask to the next, so a single daily check says little. Vercatus asks every topic's questions through the day in hourly batches and shows each share with its range, so a week's movement can be told apart from noise.

How many prompts do I need to track?

Fewer repeats and more distinct questions. In our 1 October 2026 experiments the same question asked again shared 68% of its brands with the first answer, so repeating a short prompt list mostly re-measures the same answer. Vercatus groups your market's questions into topics and spreads its answers across many questions per topic.

Can I track AI visibility for free?

By hand, yes: ask the engines your top questions every week and log who they name. It works for a handful of questions and breaks down beyond that. Vercatus gives every new domain one free audit on ChatGPT, Google AI Overviews, Google AI Mode and Perplexity, and the free check at /check/ shows whether AI crawlers can read your site at all.

Which AI engines does Vercatus track?

ChatGPT, Google AI Overviews, Perplexity, Gemini, Claude, Grok, Le Chat, You.com, Google AI Mode and Copilot Search (Bing). Each is read the way people use it, with web search on; the Google and Bing answers are read from their live results pages.

Is AI visibility the same as ranking on Google?

Related, not the same. In our 1 October 2026 experiments, 66.3% of the brands an answer recommended ranked in the top 10 for one of the searches the engine ran, against 30.7% of the brands it left out. Ranking helps, but the pages an engine reads and what it already knows decide much of the rest.

Where to go from here

Evidence, not mentions.

Create your account, add your site, and see your first answers in a free audit before you choose a plan.

Get started

Cancel any time · Data stored in the EU