Launch Bonus: new accounts get 30% off their first Amplifier Engine. Your 7 days start the moment you create your free account.

How ChatGPT, Perplexity and Google's AI pick who to mention

Ask ChatGPT for a good accountant for freelancers, or ask Perplexity which project management tools suit a small agency, and you'll get names. Google's AI Overviews do the same at the top of many search results. If you run a business, the natural next question is: how do these tools decide who gets mentioned?

The honest starting point is that nobody outside these companies knows exactly. The systems are complex, they change often and the companies don't publish a full recipe. Anyone who claims to know the precise formula is guessing. What we can do is describe what is reasonably understood, and that turns out to be useful, because it points toward habits that also make sense for ordinary search.

Two ways an AI tool can know about you

It helps to separate two different sources of knowledge.

The first is what a model learned during training. Large language models are trained on very large amounts of text, much of it published on the open web. If your business, products or ideas were written about widely and clearly before that training happened, the model may have absorbed some of it. That knowledge has a cutoff date and can be out of date.

The second is what the tool looks up at the moment you ask. Perplexity, Google's AI Overviews and ChatGPT when it searches the web all retrieve pages in real time, read them and write an answer from what they found, often with links to the sources. Here, the question is closer to classic search: which pages did the tool find, and which did it trust enough to use?

Most of what you can influence day to day sits in that second category.

What search-connected answers tend to draw on

When an AI tool searches the web before answering, it has to pick pages quickly. While the exact methods aren't public, a few qualities make a page easier to find and easier to use.

Being findable in the first place. If a tool relies on a search index, a page that isn't indexed or ranks nowhere is unlikely to be retrieved. Ordinary SEO groundwork still matters.

Answering the question directly. A page that states plainly what a business does, who it serves and where it operates is simple to summarize. A page full of vague slogans gives a tool little to work with.

Clear structure. Descriptive headings, short paragraphs and specific facts make it easier for any reader, human or machine, to pull out the relevant part.

Being consistent across the web. If your business name, services and location appear the same way on your site, in directories and in articles about you, the picture is easier to piece together.

Why being mentioned in many places helps

A single page saying you're a good choice is one voice. Several independent pages saying similar things are more convincing, to people and very likely to AI tools as well.

Think of how you'd research a contractor yourself. You'd check their website, then look for reviews, local news mentions and recommendations in community forums. When those sources line up, you feel confident. AI tools that read across many pages face a similar situation: content that is widely referenced and consistent is simply more available to draw from.

This is a reasonable inference, not a published rule. But it fits what we know about how both training data and live retrieval work: the more clear, accurate places your business appears, the more chances a tool has to come across it.

A realistic example

Picture an independent bike shop in Madison that specializes in fitting commuter bikes. Its website has a short, clear page explaining its fitting service, its neighborhood and its hours. It has written two helpful articles: one on choosing a commuter bike for winter, one on what a professional bike fit involves.

Those articles have been republished on a few local and cycling-focused sites, each copy crediting and linking back to the shop. The shop is also listed with the same details in the main map and directory listings.

When someone asks an AI tool about bike fitting in Madison, there's a fair chance the tool finds something useful: a clear service page, a couple of on-topic articles and consistent mentions elsewhere. Nothing guarantees the shop gets named. But it has made itself easy to find, easy to understand and easy to cite, which is about as much as any business can do.

What we don't know, and why that's fine

There's no public dashboard showing why one business was named and another wasn't. Answers can vary from one person to the next, and from one day to the next, even for the same question. Each company also runs its systems differently, so a habit that helps with one tool may matter less for another.

That uncertainty is a good reason to focus on fundamentals rather than tricks. Clear pages, helpful articles, consistent business details and genuine mentions across the web are useful for customers, useful for search engines and very likely useful for AI tools too. If the systems change, those habits still hold up.

Where to put your effort first

If you're starting from scratch, three steps go a long way. Make sure your main pages say plainly what you do and for whom. Publish articles that answer the real questions your customers ask. Then get those articles in front of more readers on more sites, so your business shows up in more places than just your own.

Launch Pages publishes your article and republishes it across a network of news and blog sites, each copy crediting you with a link back, so your work appears in more places that people and tools can find. Learn more

Your next article could be live on 135+ sites.

Create Free Account