Editorial guide · Inspeccia
How to measure your AI visibility (a free, step-by-step method)
You can measure your AI visibility for free with a 30-minute manual method: make a list of 10 to 15 real questions your customer would ask, run them in ChatGPT, Perplexity and Google's AI Overviews (in incognito), and note whether you appear, which competitors show up, and which sources the AI cites. Repeat the same exercise once a month with the same questions and you get a trend. Three metrics matter, in this order: mention rate, share of voice, and cited sources.
Nothing else is required to start. No subscription, no dashboard, no expensive tool you'll stop opening by week two. What follows is that method in detail, the three metrics that actually move the needle, and a free signal almost nobody checks.
The manual method in 4 steps
The idea is to stand in your customer's shoes and watch what the model hands back. Four steps, and the first one decides most of the outcome.
- List 10 to 15 real customer questions. These aren't keywords, they're natural-language questions, the way a person would actually ask. If you sell web audits: "what tool should I use to check my site," "how do I know if my SEO is healthy," "cheaper alternative to Semrush." Think like your customer, not like a marketer. This list is the heart of the method: a good list of fifteen questions is worth more than a script blasting a hundred generic ones.
- Ask in ChatGPT, Perplexity and Google's AI Overviews. Do it in an incognito window and signed out, so personalization doesn't show you results only you can see. With live search on wherever you can, because that's what most users get. Those three cover most of how people search with AI today.
- Record three things per question. Did you appear, yes or no? Which brands showed up instead of or alongside you? Which source did the model cite to answer? A plain spreadsheet with date, question, model and those three columns is plenty. What you don't write down, you don't measure.
- Repeat each month and chart the trend. A single measurement isn't a trend — it's a snapshot. What teaches you is the month-over-month comparison — appearing more, appearing less, holding steady — not the standalone reading. Three monthly measurements are the minimum before drawing conclusions with any confidence.
The three metrics that matter
Of everything you could write down, these three are the ones worth tracking over time. The rest is pretty noise.
1. Mention rate
This is the percentage of your key questions where you appear. Run fifteen questions and show up in six, and your mention rate is 40%. It's the most direct metric: it answers, plainly, "out of the questions that matter, how often does the model name me?" The interesting part isn't the absolute number but how it moves month over month.
2. Share of voice
This measures how much you appear against your competition. Getting named isn't enough: it matters whether they name only you, you plus two others, or you lost among ten. If a question surfaces you and one competitor, your share is half; if five brands show up and you're one, it's far smaller. This metric tells you your relative standing in the category, which is what the model ends up passing on to whoever asked.
3. Cited sources
This is which pages the AI uses to talk about your niche. When the model answers with live search, it usually cites where it pulled the information from. Logging those sources shows you the real terrain: if the same directory, the same outlet or the same forum thread keeps coming up, you now know where you need to be present to get into the conversation. It's the metric that tells you not just where you stand, but what to do about it.
The free signal almost nobody checks: your AI traffic in GA4
Everything above measures what the model says. But there's an even more honest signal, and you have it for free in your account: the traffic already arriving from AI. In Google Analytics 4, referrals from chatgpt.com, perplexity.ai and gemini.google.com are people who didn't just see you in an answer — they clicked and landed on your site. That's a result, not an estimate.
It complements rather than replaces the manual method: the questions tell you whether you appear, GA4 tells you whether that appearance turns into visits. How to isolate and read that traffic is covered in how to track AI traffic in Google Analytics 4. And if you want to go deep on ChatGPT specifically — serious prompt testing, mention monitoring and Brand SERP analysis — that lives in how to measure if ChatGPT recommends you.
The manual method works great at first. The trouble starts as it grows: more questions, more models, comparison against competitors, a monthly cadence. That's when the hand-kept spreadsheet turns painful by month two.
That's exactly what Inspeccia does: the audit checks live how Google and ChatGPT see you and flags your real competitors — the fastest way to start measuring without building spreadsheets. Start a free analysis or request a custom audit.
Frequently asked questions
Can AI visibility be measured precisely?
Not with millimeter precision, and you should be wary of anyone who promises otherwise. A model's answers change from one run to the next: they depend on the exact prompt, on whether live search is on, and on a degree of randomness built into the model. What you can get are reliable trends. If you always use the same questions and repeat them every month, the pattern you see — appearing more, appearing less, holding steady — is solid even though each individual answer varies.
How often should I measure?
Once a month is enough to spot trends for most businesses. Measuring more often tends to add noise, not signal: the natural variation between answers drowns out the real movement. The exception is when you're mid-campaign or have just published new AI-oriented content; there it makes sense to measure every two weeks to catch the before-and-after effect.
Do I need to pay for tools to measure?
Not to get started. The manual method in this article costs nothing but your time, and your AI traffic in Google Analytics 4 is already sitting in your account waiting for you to look at it. Paid tools make sense as you scale: many questions, several models, comparison against competitors, and a cadence that no longer fits a hand-kept spreadsheet. But that's a decision about scale, not an entry requirement.
Measure your AI visibility in 90 seconds
Inspeccia checks live how Google and ChatGPT see you, shows whether you appear and which competitors it positions you against. The fastest starting point, no spreadsheets.