AI visibility tracking:
does ChatGPT recommend you?
Your customers ask ChatGPT, Gemini and Claude what to buy. Glowytics asks the same questions twice a week, with web search on, and shows how often each assistant recommends your brand, which competitor it picks instead and which websites it trusts.
Share of recommendation
Mattresses, unbranded questions
ChatGPT
21%
Gemini
29%
Claude
22%
Illustrative numbers. Brand names are placeholders.
One answer, taken apart
Every answer becomes a row of data.
A question goes to each assistant exactly as a shopper would type it, with no brand name in it. We store the full answer, then pull out every brand it names, the order, the reason given, the tone and the sources it cited.
Which mattress is best for back pain? My budget is mid-range.
Searched the web before answering
For back pain, a medium-firm mattress with zoned support is the usual advice 1. Northwind Sleep Ortho Zone comes up most in reviews for lumbar support 2. Lumen Home is a good mid-range pick with a long trial period, although some buyers find it firm 3. Alder & Co is worth a look if you sleep hot.
| # | Brand | Reason given | Tone | Sources |
|---|---|---|---|---|
| 1 | Northwind Sleep | Zoned lumbar support, strong reviews | Positive | 2 |
| 2 | Lumen HomeYou | Mid-range price, long trial | Mixed | 3 |
| 3 | Alder & Co | Cooling for hot sleepers | Neutral | – |
Illustrative example. Brand names are fictional; real projects show your brand and the competitors the assistants actually name.
What you get
Eleven views, named.
Not a single vague "AI score". These are the screens in the AI Visibility dashboard and the question each one answers.
11 reports included
Visibility
Share of recommendation
Of all the brands the assistants recommend in your category, what share goes to you, per assistant and per period.
Brand comparison
You against the competitors you name, plus the ones the assistants bring up that you did not list.
Trend
Share of recommendation from one measurement to the next, so a drop shows up within days.
Need map
Which customer needs (budget, back pain, cooling, delivery) you own, and where a competitor gets picked instead.
Sources
Source types
How much the answers lean on brand sites, marketplaces, news, forums, review sites and expert content.
Sources × brand
The websites cited when you are recommended, and the ones cited when a competitor wins and you do not.
Pages
The exact URLs behind the answers, so PR and content work has a target list.
Perception
Tone
Positive, mixed, neutral or negative, with the strengths and caveats the assistants repeat and the sentences they use.
Branded questions
What the assistants say when asked about your brand directly, and which alternatives they suggest.
Method
Raw answers
Open the full answer behind every number and read it yourself.
Period comparison
Confidence intervals and paired comparisons separate a real change from model noise.
Answers are stored in full, so past periods can be recalculated when the method improves. A read-only, password-protected link can be shared with your team.
How it works
We agree the questions
Together we list the questions shoppers in your category ask, grouped by need. Brand names never appear in them, in any language or market you sell in.
Assistants answer, we extract
Each question goes to ChatGPT, Gemini and Claude with web search on, from your market's location, twice a week. Every brand, rank, reason, tone and source is captured.
You read the dashboard
Pick a period, compare it with the previous one and open any answer. You see where you are recommended, who wins instead and which sites to work on.
AI visibility, explained
What is AI visibility?
AI visibility is how often, and how favourably, AI assistants such as ChatGPT, Gemini and Claude mention or recommend your brand when people ask them questions about your category. It is measured with unbranded questions, the way a shopper asks before they know which brand to pick.
How is this different from SEO or generative engine optimization (GEO)?
SEO and GEO are the work: improving your site, content and presence on the sources AI assistants read. AI visibility tracking is the measurement that tells you whether that work changes the answers, and which sources and needs to focus on next.
Which AI assistants do you track?
ChatGPT, Gemini and Claude, each with web search enabled so the answers reflect what a real user sees today, not only the model's training data.
Why not just ask ChatGPT myself?
One answer is one sample. The same question gives different brands from run to run and from assistant to assistant. Tracking means many questions, three assistants and repeated measurements, with confidence intervals so you can tell a real change from noise.
How often is it measured?
Twice a week by default. Each run is stored in full, and the dashboard pools runs into the period you choose.
Do you include our competitors?
Yes. You name your competitors, and every other brand the assistants mention is captured too, so new rivals show up without anyone adding them.
Does it work outside Turkey and in other languages?
Yes. Questions are written in the language your customers use and the assistants are queried with your market's location, so the answers match what local shoppers see.
Find out what ChatGPT tells your customers.
We build the question list with you. After that, the assistants answer twice a week and the dashboard fills with your own data.