Does ChatGPT name you when your customer asks?
The difference between being readable and being recommended, why the answer changes every time, and how to measure where you stand instead of guessing.
When your customer asks ChatGPT for the best provider in your field, an answer comes back with names in it. This page explains why one name appears and another does not, why being readable is not the same as being recommended, and how to find out where you stand by measuring rather than guessing.
Indexed is not the same as recommended
A crawler being able to read your site is necessary but nowhere near sufficient. Thousands of perfectly readable sites are never named by an assistant, because being readable only makes you available. Being named requires that your page is a clear answer to the question asked, and that something outside your own site corroborates you.
Why an assistant names one company over another
Three factors recur in our measurements among the names that do appear:
A page that says what you sell, in words. The companies that get named usually have a page that states the service explicitly — not a general marketing message, but text saying what the service is and who it is for. A page that tells a beautiful story without ever naming the service is not read as a service.
Presence outside your own site. The assistant builds its answer from sources. If your name is in none of them, it has nothing to lean on in order to name you.
A clear tie to the market. These questions are usually asked locally — "in Amman", "in Riyadh", "in Dubai" — and answers lean towards names whose connection to that market is visible.
The answer changes every time
Ask ChatGPT the same question three times and you may get three different lists. That is how generative models work; it is not a fault. The practical consequence is blunt: a single ask measures nothing. You might appear once and be missing twice, and walk away either reassured or discouraged on a sample of one.
This is why every question is asked three times on every engine, and the count is taken across everything that came back.
Why it matters who was named instead
"I wasn't mentioned" is half a number. Knowing who was, and in how many answers, tells you whether your category is settled around one name or wide open. In our measurements of young Arabic-language markets the usual finding is that nobody dominates — which is a completely different piece of news from "I am absent".
Why cited sources matter
An answer is not an opinion out of thin air; there are pages behind it. Collecting the links the answers leaned on and checking them tells you where the assistant reads about your category, and whether your name is on those pages. That is a far more practical list than any general advice.
Why crawlability matters
Where a site blocks one of the AI crawlers, or its text lives inside images, or nothing appears until JavaScript has run, what everything else is built on may not arrive at all. This is a thing the report checks and reports; it is not something anyone can tell you without fetching the site first.
How to find out where you stand
The free check — twelve questions about your category and city, put to ChatGPT and Gemini, showing how many answers named you and who was named instead. No account, no card.
The full report — 72 questions across three engines and three rounds, with a crawl of your site, the list of cited sources and a plan in priority order. Details on the methodology page.