AI visibility tracking: the complete guide (what it measures and how to read it)

Measure how often AI assistants name your business when buyers ask. What AI visibility tracking is, the four numbers that matter (visibility, position, citations, share of voice), how it works across six engines, and how to read the results without a marketing degree.

AI visibility tracking is the practice of measuring how often AI assistants name your business when people ask them for a recommendation. You pick the questions your buyers actually ask, run them against the AI engines on a schedule, and record whether you were named, where, and who was named instead. That is the whole idea. This guide covers what it measures, how it works, and how to read the numbers without a marketing degree.

Why this suddenly matters

A growing share of your customers no longer Google a problem and click around. They ask ChatGPT, Perplexity, or Google's AI Overview "what's the best [thing] for me?" and take the short list it reads back. If your name is in that list, you get the customer. If it is not, you never existed to them, and here is the part that stings: nobody tells you. There is no bounce in your analytics for an answer you were left out of. The loss is silent, which is exactly why measuring it is the first move.

We saw how invisible this can be when we ran the test on our own category. Across 48 AI answers and six engines, the market leader in our space was named in 29% of answers and we were named in 0%. We did not know that until we measured it. Most businesses are in the same spot: losing recommendations they cannot see.

What "AI visibility" actually measures

There are a few numbers, and each answers a different question. In plain terms:

Visibility. Out of every 100 AI answers about your category, how many name you at all. This is the headline number: are you in the conversation or not.
Position. When you are named, where you land in the list. Being named first is worth far more than being named fifth, because readers stop early.
Citations. Whether the engine actually linked to your website as a source, not just mentioned your name. A citation is a stronger signal than a mention.
Share of voice. Your slice of all the brand mentions in your category, versus the specific competitors you track. It tells you how you stack up, not just whether you show up.

You do not need all of them on day one. Start with visibility, are you named, yes or no, and add the rest as you get serious. Two are worth a closer look once you are: AI share of voice (your slice versus competitors) and what counts as a good visibility score.

How AI visibility tracking works

The mechanics are the same whether you do it by hand or with a tool:

Write the buying questions. Not "tell me about my company," which no customer types. The real ones: "best [category] for [situation]," "who should I hire for [job] in [city]," "top [product] brands for [need]."
Ask them on every engine, repeatedly. Buyers do not all use the same assistant, so a real picture spans ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude and Grok. And because any single answer is close to a coin flip, engines reword themselves run to run, you ask on a schedule and track the rate, not one roll.
Record who was named, and what the engine read. For each answer: were you named, in what position, and which sources did the engine pull before deciding. That last part is the one that turns "we're invisible" into "here is the page we're missing from."

Doing it yourself vs using a tool

You can start free. Ask your top questions in ChatGPT once a week and log the results in a spreadsheet across a few engines. It works, it just gets tedious fast and it will not show you the sources behind each answer. We walk through the free-and-paid options in detail in how to check if ChatGPT recommends your business.

A dedicated tracker does the weekly asking across every engine for you and adds the piece a spreadsheet cannot: the sources. When a competitor is named and you are not, it shows the exact pages the AI trusted, almost always a third-party review or list post rather than the competitor's own site. That is what SearchVis does, and there is a free tier so you can see your own number before paying anyone.

How to read your numbers without overthinking

A few rules of thumb once the data comes in:

A single answer means little. A rate over time means everything. "Named in 2 of 10 answers, always behind the same two competitors" is a real finding. "ChatGPT named me once" is noise.
Watch the trend, not the daily wobble. Answers vary run to run. What matters is whether your visibility is climbing over weeks as you make changes.
Sources beat scores. The number tells you that you are losing. The sources view tells you why, and why is the only thing you can act on. If a competitor keeps getting named instead of you, start with why ChatGPT recommends your competitors.

Common mistakes

Tracking only broad head questions ("best CRM software") where you compete with giants and always lose. Add specific, winnable questions ("best CRM for a two-person real estate team"). Winning a narrow answer at position 1 beats being absent from a broad one.
Checking one engine. Your buyers are split across six. A great score in ChatGPT and a zero in Google's AI Overview is a half-true picture.
Measuring and stopping. Tracking that you are invisible without acting on the sources is just anxiety with a chart. The point is the fix.

Frequently asked questions

What is AI visibility tracking? It is measuring how often AI assistants (ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude, Grok) name your business when buyers ask for a recommendation, tracked over time across engines rather than from a single answer.

Is it the same as SEO rank tracking? No. A rank tracker measures your position in a list of links. AI visibility tracking measures whether you are named inside a written answer, which is a different question with a different answer. More in AEO vs GEO vs SEO.

Can I do it for free? Yes, roughly. Ask your buying questions in ChatGPT weekly and log them in a spreadsheet. You lose the multi-engine coverage and the source data, but it costs nothing to start. SearchVis also has a free tier.

How often should I check? Weekly is plenty for most businesses. Daily during a push (a new page, a PR mention) so you can see whether it moved the needle.

Why am I named on some engines and not others? Because only about 11% of pages are cited by both ChatGPT and Google AI Overviews for the same query. The engines retrieve different sources, so you can win one and lose another. That is why you track all of them.

If you are a small business or an agency

The approach shifts a little depending on who you are. If you run a local or small business, start with AI visibility for small businesses, which focuses on the specific and local questions you can actually win. If you serve clients, AI visibility for agencies covers how to report it and turn it into a service.

The bottom line

You cannot improve what you cannot see, and AI recommendations are invisible by default. Start by measuring, even roughly, so you know which questions you win, which competitors beat you, and which sources you are missing from. Then go fix the sources. For the how-to-fix half, see our plain-English guide to answer engine optimization.