Short answer: nobody outside these companies knows the full recipe, and they don't publish it. What they do publish is enough to act on. Here's what's documented, what's a reasonable guess, and what's unknown.
What the companies say
Google (AI Overviews and AI Mode). A page has to be indexed and eligible to show in search with a snippet. Google says there are no additional requirements, and no special optimization, markup or schema is needed. It also says meeting every requirement doesn't guarantee a page gets crawled, indexed or shown.
OpenAI (ChatGPT search). OpenAI runs a crawler called OAI-SearchBot for its search feature. Its documentation says sites that opt out of it won't be shown in ChatGPT search answers. A separate crawler, GPTBot, is used for training, and blocking it doesn't affect search.
Perplexity. It says it searches the web in real time and puts numbered footnotes in each answer that link to the original sources. It describes favoring authoritative sources and picking the material that best fits the question, without explaining how it ranks them.
What that adds up to
Two things have to be true before an AI tool can recommend you.
- It can reach your pages. They're indexed, not blocked, and not hidden behind scripts or logins.
- Your pages answer the question. That means a page that says what you do, for whom, and where, in plain text.
Everything after that is judgment by the tool, and the tool doesn't tell us how it judges.
What research suggests
One academic study of GEO, presented at KDD 2024, tested different ways of changing page content. Adding statistics and quotations from relevant sources improved how often content was cited. Stuffing keywords made it worse than doing nothing. It's one study on a general benchmark, not on IT firms, so treat it as a hint about what these systems like: specific and sourced beats repetitive.
What's a reasonable guess
AI answers are built from what the tool can find about you, and not only on your own site. If a directory, a partner page or an article describes your firm, that's another source it might use. I couldn't find a published study I'd trust that measures how much weight each type of source gets, so I'd treat this as likely, not proven.
What nobody knows yet
- Why the answer changes. A 2026 SparkToro study found that the odds of getting the same list of brands twice were under 1 in 100. Some brands still appeared in 60% to 90% of responses for a given kind of question, even while their position moved. Smaller markets gave steadier results, which may help niche IT firms, though the study didn't test that.
- How much of this comes from the AI's memory and how much from live search. It differs by tool and changes with updates.
What you can do about it
- Make sure the crawlers can reach you. Check robots.txt for OAI-SearchBot and the other search crawlers.
- Write pages that answer real buyer questions in plain text.
- Keep how you're described the same across your site, LinkedIn and directories.
- Check how often you're named across many questions, not just once.
Sources
- Google Search Central, AI features and your website.
- OpenAI, Overview of OpenAI crawlers.
- Perplexity, How does Perplexity work?
- Aggarwal et al., GEO: Generative Engine Optimization, KDD 2024.
- Search Engine Land, AI recommendation lists rarely repeat, SparkToro study, January 2026.
Common questions
How does Perplexity choose its sources?
It says it searches the web in real time and links numbered footnotes to the original sources. It does not explain how it ranks them.
Do I need special markup to appear in Google AI Overviews?
Google says no. A page has to be indexed and eligible to show in search with a snippet, and no special optimization, markup or schema is needed.
How do businesses get mentioned in AI answers?
Make the site reachable, write pages that answer real buyer questions, and keep your description consistent across your site, LinkedIn and directories. No one can guarantee a mention.
Kryston's AI visibility audit gives you a dated baseline of this for your own firm: what ChatGPT, Perplexity, Gemini, AI Overviews and Google say about it today, and what to fix first.
Kryston guides
Analysis for IT services and software development firms on search, AI visibility and buyer trust.
