You asked ChatGPT a question about your industry. The answer lists three or four brands, and yours is not one of them. That is frustrating, but it is too early to draw a conclusion from a single answer. In this article we first explain how to check whether this is really the case, then cover the most common causes and what you can do about each one.
Why is one answer not enough?
AI assistants do not give the same answer to the same question every time. In a study SparkToro ran in late 2025, 600 volunteers ran 12 different prompts through ChatGPT, Claude and Google's AI features a total of 2,961 times. The result: the chance of two answers containing the same list of brands was under 1%. The chance of the same list appearing in the same order was roughly one in a thousand. The details of the study are in SparkToro's article.
So if you asked once and did not see your brand, that does not mean you are never mentioned. Seeing it once and relaxing can be just as misleading. The same researchers say that a mention rate measured across many prompts and repeated runs can be a reasonable indicator.
How do you check the situation first?
The right question is not "Did ChatGPT recommend me?" It is "How often am I mentioned in the questions my buyers ask?" To find out:
- Write down the questions your buyers would actually type. Your brand name should not appear in them. For example: "What are the best X companies?" or "Recommend a company that offers X service in Istanbul."
- Ask these questions to ChatGPT, Gemini and Perplexity separately, not to just one engine. The engines draw on different sources, so their results differ too.
- For each answer, note whether your brand appears, in what position, and which brands are mentioned instead.
- Do this at regular intervals, not just once.
You can do this by hand. As the number of questions grows, though, it quickly becomes time-consuming. AIShortlist does it automatically: it suggests questions from your site, does not start measuring until you approve them, then asks each question to three engines and shows the results on one screen.
If the measurement really does show low visibility, the next step is to find the cause. The five causes below are the most common ones.
1. Can AI bots access your site?
The first place to look is your robots.txt file. Some sites block all AI bots at once so their content is not used for AI training. That setting can also keep you out of search answers.
According to OpenAI's crawler documentation, OAI-SearchBot is used to show sites in ChatGPT's search features. Sites that block this bot are not shown in ChatGPT search answers. The bot used for model training is a separate one: GPTBot. So you can stay closed to training and still be open to search. Perplexity makes a similar distinction; its own documentation recommends allowing PerplexityBot if you want to appear in search results.
One thing to watch: you may not have changed robots.txt yourself. A security plugin, your hosting company or a CDN setting can also block bots.
2. Does your site clearly say what you do?
To recommend a brand, an AI has to understand what that brand sells, who it sells to and where it operates. Sites that have only a slogan on the home page, explain their services inside images, or never mention their city anywhere fall behind here.
To check, read your own service page as if you were a stranger. Does the first paragraph answer these three questions: what do you do, for whom, and where? If not, that is the first thing to fix.
3. Are you on the pages the engines read?
AI answers are often compiled from other sites. In an Ahrefs study on ChatGPT, 750 prompts in software, product and agency categories were analyzed. 43.8% of the pages cited in the answers were "best X" style blog lists.
The takeaway is clear: if your competitors are on these lists and you are not, ChatGPT will mention them. No matter how much you improve your own site, that gap does not close. Finding out which list pages are cited and reaching out to their owners is a separate task.
4. Is there enough information about your brand elsewhere?
Engines do not read only your site. Articles about you, industry directories, marketplaces, forums and review sites also shape the answer. If your name does not appear in these places at all, or appears with a different service description in each place, the engine may not be able to match you to the right category.
The work here is to describe your brand's basic information (what you do, where you are, which category you are in) in the same sentences on your site and on your external profiles.
5. Do you have a page that answers the question being asked?
If a buyer asks "What affects the price of X service?" and your site does not have a single paragraph that answers it, the engine has no reason to mention you in that answer. Writing subheadings as questions and putting a short, clear answer right below them is a simple change that helps.
A word of caution: adding numbers you do not have or claims you cannot verify will backfire. When you revise your text, use only the services you really offer and the information you actually have.
Are you a new brand?
If your brand is a few years old or newer, things are a bit different. AI models learn from text up to the point they were trained, and that knowledge is frozen at a certain date. If there was not enough written about you when the model was trained, the engine does not know you from its own knowledge.
In that case, the main way the engine can mention you is web search. When answering a question, the engine searches the web and reads the pages it finds. For a new brand this is good news: you do not have to wait for the model to be retrained. But there are two conditions. First, search bots must be able to access your site. Second, your name must appear on the pages the engine reads, meaning the list and guide pages in your category.
That is why visibility on external pages matters even more for new brands than for established ones. A sensible place to start is finding which pages are cited in your own category.
How do you tell which cause applies to you?
These five items are possible causes, not a definitive diagnosis. Why an engine does not name a brand cannot be known for certain from the outside; what you can do is test each possibility in turn. Not all five may apply at once. This order works well:
- Measure first. See which questions and which engines you are not mentioned in.
- If one of the engines never mentions you, check the access setting for that engine's bot.
- Look at the pages cited in the answers. If there are list pages and you are not on them, the work is external.
- If your own pages are never cited, go back to your page text.
- After making changes, measure again with the same question list. AI answers do not change overnight, which is why weekly tracking makes sense. A rise after a change is a good sign, not proof that the change caused it; answers also change on their own.
The AIShortlist dashboard shows these steps on separate screens: the overview shows the measurement results, the Sources screen shows the pages the engines read and the lists your brand is missing from, and the Site screen shows the technical check and page text suggestions.
Is there a way to guarantee a recommendation?
No. No tool or agency can guarantee that ChatGPT will recommend you. Answers depend on how the engine works, the sources it reads and third parties. What you can do is measure the situation honestly and take the steps that raise your chances of being recommended, one by one.
If you want to see where you stand as a first step, you can run a 10-question measurement with the free trial. No credit card required.