How ChatGPT Chooses Which Businesses to Recommend

ChatGPT chooses which businesses to recommend from two sources: what its underlying model learned during training, and what its live search finds and reads when you ask. When it searches, it favors businesses that several trusted sources describe consistently, such as review platforms, directories, best-of lists, local news and the business's own website. It tends to name the businesses it can find most often, described most clearly, for the service and place in the question.

This matters because people now ask AI tools who to hire. In BrightLocal's 2026 consumer survey, 45% of consumers had used AI tools to find local businesses, up from 6% a year earlier.

Two Ways ChatGPT Knows About a Business

What the model learned in training

ChatGPT's model was trained on a large body of text. If your business was widely written about before its training cut-off, it may mention you from memory. Training data is a snapshot, though: newer businesses, recent changes and small local firms often are not in it, or appear with outdated details.

What its live search finds

For questions about current options, like "best family dentist in Denver" or "who installs heat pumps near Austin", ChatGPT often searches the web, reads several pages and summarizes them, with links to its sources. This is where most businesses can make a difference, because it depends on what is published about you now.

OpenAI's documentation on its crawlers explains that its search crawler, OAI-SearchBot, is what surfaces websites in ChatGPT search, and that sites which opt out of it will not be shown in ChatGPT search answers. It is separate from GPTBot, which collects training data. You can block one and allow the other.

The Signals That Shape a Recommendation

Why Some Businesses Get Named and Others Do Not

Look at a ChatGPT answer that recommends a competitor and check the sources it cites. You will usually see a pattern: the same handful of review sites, one or two local "best of" articles, a directory and sometimes a forum thread. The businesses named are the ones that appear in those pages.

That is good news for smaller firms. You do not need to be the biggest business in your market, but you do need to be present, and described clearly, on the pages AI search reads for your service and place.

How to Check Where You Stand

1. Write down the questions your buyers ask, the way they would ask them: "Who is the best emergency plumber in Tampa?", "Which accounting firms in Leeds work with contractors?"
2. Ask ChatGPT each one in a fresh chat, with search turned on, and note who is named.
3. Record the sources it cites. These are the pages where you need to appear.
4. Repeat monthly. Answers change with wording, location and what the search finds, so track a set of questions over time rather than one answer.

Our free AI Overview checker does the same for Google's AI Overviews, which draw on similar signals.

How to Become Easier to Recommend

• Allow OpenAI's search crawler in your robots.txt, even if you block training crawlers.
• Make your service pages unambiguous: what you do, where, for whom, prices or ranges, and proof.
• Earn reviews steadily on the platforms your customers trust, and reply to them.
• Get listed and mentioned where AI search looks for your field: reputable directories, local press, associations and the "best of" articles that already rank.
• Keep your details identical across your site, your Google Business Profile and every listing.
• Publish specific, original information, such as prices, timelines and data from your own work, that other pages do not have.

This is the core of generative engine optimization. Our plain guide to GEO explains the wider approach, and our guide to showing up in Google AI Overviews covers Google's side.

If you would rather see exactly where you stand first, our AI search optimization service starts with a visibility test across the questions your buyers ask.