Glotier tells you whether ChatGPT, Gemini and Perplexity name your product when buyers ask what to use in your category. It puts your buyers' real questions to all three as live web queries, records whether you were named and by which of them, which competitors were named instead, and which pages each answer was built from.

How does Glotier check AI visibility?

It runs one live web search per question. ChatGPT and Gemini read the identical results, and Perplexity runs its own retrieval because it does not accept results handed to it. Whether your product is named is decided by a literal presence check on the answer text rather than by asking a model for its opinion, so the number cannot be inflated by a model being agreeable. You can also connect your own ChatGPT or Claude key and have the questions run through your account instead.

Why ask three assistants instead of one?

Because they disagree. In our own study of 40 buying questions, all three named the same product only 30% of the time, and 47% of products were named by exactly one assistant. Checking a single assistant shows you roughly 60% of what is being said about your category, so a founder who checks ChatGPT, sees their name and relaxes has measured one engine on one phrasing.

Is Glotier free to use?

The check is free with no signup. Paid plans start at $39/month (Solo), $99 (Studio) and $249 (Agency), and cover weekly tracking and the content the agent writes.

Does Glotier make up data about my product?

No. Every result comes from observed data: the live search results, the answers the assistants actually returned, and the sources they cited. It does not invent metrics, ratings or engine claims, and when a measurement has not run it says so rather than printing a zero.

Anyone selling something buyers research before choosing, where the category has enough alternatives that an assistant has to pick a shortlist: software, tools, e-commerce, and services. Founders, marketers and agencies all use it, and the work is the same at any size. It is not useful for local or in-person businesses, where maps and reviews decide the outcome instead.

Can I train ChatGPT to recommend my product?

No, and nobody can sell you that. You have no access to the training of a model you do not own, and the conversations you have with ChatGPT do not teach it about your company. What changes an answer is what the model retrieves at the moment it answers, which is a live web search over pages other people control.

Does talking to ChatGPT about my brand teach it anything?

Not for other people. A chat can carry context inside that one conversation, and memory can carry preferences inside your own account, but neither reaches the stranger in another country asking which tool to buy. The answer they get is assembled fresh from a search you were not part of.

What about feeding it my website or an llms.txt file?

Those help a model that has already decided to look at you, and they do nothing to get you into the shortlist. An assistant answering "what is the best X" is reading third-party pages it retrieved for that question. Publishing a beautifully structured file on your own domain does not put you inside somebody else's comparison article.

So what actually changes the answer?

Being present in the pages the assistant retrieves. In the 3 categories we measured end to end, 18 answers were built from 131 cited sources, and the same few domains kept deciding each category. Get into those and the answer changes. Everything else is decoration.

How would I know if it worked?

By asking the same questions again and comparing. One measurement is a snapshot and tells you nothing about direction. This is the honest reason a tool exists in this category at all: not to teach the model anything, but to re-run the same questions on a schedule so a change is visible rather than felt.

Guide

Can you train ChatGPT to recommend your product?

No. Not by chatting to it, not by feeding it your website, not by any service selling you that. The answer does change, though, and the thing that changes it is worth understanding properly.

Why the training idea does not work

You have no access to the training of a model you do not own. That much is obvious once stated. The less obvious part is that your own conversations do not teach it either. A chat can hold context inside that one thread, and account memory can hold your preferences for you, but neither of those reaches the stranger three countries away asking which tool to buy. They get an answer assembled fresh, from a search you were never part of.

This is why “I told ChatGPT about my product and now it knows me” is one of the most expensive misunderstandings in the category. It is true in your session and false everywhere else, which is the worst combination: it feels verified.

What actually happens when someone asks

An assistant answering a buying question runs a live web search and writes from what comes back. So the answer is decided by a handful of pages, retrieved for that specific question, almost none of which you control.

On 26 July 2026 we measured this properly: 3 software categories, six real buying questions each, put to ChatGPT, Gemini and Perplexity, and then we read every source behind the 18 answers. 131 source appearances in total. In each category a small number of domains kept deciding the outcome, and they were third-party pages, not vendor sites.

Two findings worth carrying into any plan. Community forums accounted for 21 of those sources and review directories for 1, which is 0.8% of the total. And the gatekeeper was different in every category: one was decided by a blog most people in that industry have never heard of, another by Reddit outright.

The things people try instead, ranked by how little they do

1
Talking to ChatGPT about your product
Does nothing outside your own session. It is the most common thing people try and the only one that does literally nothing for a stranger's answer.
2
Adding schema markup or an llms.txt to your site
Helps a model that has already arrived at your page decide what you are. Does not get you into the shortlist, because the shortlist is assembled from other people's pages.
3
Buying a directory listing
Worth checking before renewing. Across our 18 answers, directories were 1 of 131 sources. Your category may differ, which is the point: measure it rather than assume.
4
Getting into the pages that actually get retrieved
The one that works, and the slowest. It means being genuinely useful in the specific places your category's answers are built from, and publishing the thing a gatekeeper would want to cite.

How you would know it worked

By asking the same questions again and comparing. That sounds obvious and almost nobody does it, which is why so much of this field runs on feeling. One measurement is a photograph. Direction needs two, taken the same way, far enough apart to mean something.

We hold ourselves to it: across 21 buying questions in our own category, AI names Glotier in 0 of them, and we publish that number because a measurement you only show when it flatters you is not a measurement.

Do this automatically

Get Glotier Solo

You cannot train the model. What you can do is find out exactly which pages decide your category and work on those, then check whether the answer moved. Solo runs that loop daily and writes the work it points to.

The same buyer questions re-asked every day, so you see the moment an answer changes
Every source page behind each answer, which is where the work actually is
An article written from what it measured that day, ready for you to publish
The subreddits and threads worth answering, with a reply drafted for each

Track 3 products and up to 150 buyer questions across ChatGPT, Gemini and Perplexity. $39/month.

Get Solo, $39/month Compare the plans

Cancel any time. Not for you? Email us within 7 days of a charge and we refund it in full. Refund policy

Not ready to pay? The check is free, with no card and no account. Run it on your own product.

Questions people ask next

Can I train ChatGPT to recommend my product?: No, and nobody can sell you that. You have no access to the training of a model you do not own, and the conversations you have with ChatGPT do not teach it about your company. What changes an answer is what the model retrieves at the moment it answers, which is a live web search over pages other people control.
Does talking to ChatGPT about my brand teach it anything?: Not for other people. A chat can carry context inside that one conversation, and memory can carry preferences inside your own account, but neither reaches the stranger in another country asking which tool to buy. The answer they get is assembled fresh from a search you were not part of.
What about feeding it my website or an llms.txt file?: Those help a model that has already decided to look at you, and they do nothing to get you into the shortlist. An assistant answering "what is the best X" is reading third-party pages it retrieved for that question. Publishing a beautifully structured file on your own domain does not put you inside somebody else's comparison article.
So what actually changes the answer?: Being present in the pages the assistant retrieves. In the 3 categories we measured end to end, 18 answers were built from 131 cited sources, and the same few domains kept deciding each category. Get into those and the answer changes. Everything else is decoration.
How would I know if it worked?: By asking the same questions again and comparing. One measurement is a snapshot and tells you nothing about direction. This is the honest reason a tool exists in this category at all: not to teach the model anything, but to re-run the same questions on a schedule so a change is visible rather than felt.

Keep reading

Measured 26 July 2026 against ChatGPT, Gemini and Perplexity on one shared live web search per question. Assistants change their answers over time, so every figure here is a record of a moment.