You show up in ChatGPT and other AI search results mostly the way you show up in Google: by having pages that are easy for a machine to read, clear about who you are and what you do, and good enough that a source worth citing would point to them. There's no separate trick. Google says as much in its own guidance on AI features and your website: the same fundamentals that help a page in regular search apply to its AI features, and there are no extra requirements for appearing in them.
That also means nobody can promise you a place in an AI answer. These tools choose what to cite on their own, and the choice changes. What you can control is whether your site is a reasonable candidate, and that comes down to a few things you can check this week.
Why AI answers are a different surface
A classic search result is a list. The searcher scans ten links and picks one. An AI answer is a short written reply that pulls from a handful of pages, sometimes with links beside it and sometimes with none. If your business isn't one of the pages it pulled from, the searcher never sees you at all. There's no page two to scroll to.
That raises the value of being the clear, specific source on your own subject. A vague page that says a little about everything is easy to skip. A page that answers one question well, in plain words, is easy to quote.
What makes a page easy to cite
Start with the basics, because the AI tools can only use pages they can reach.
- Make sure the page can be crawled (Google's and other systems' bots can open it) and isn't blocked by a setting you forgot about.
- Put the answer in text. If your pricing, service area or process only appears inside an image or a PDF, a machine may never read it.
- Give each page one clear job, with a heading that says what it covers.
- Say who you are and where you work in plain sentences: your business name, your city, what you sell and who it's for. Entity clarity (making it unmistakable which business a page is about and what that business does) is the AI-search version of a good address on a letter.
- Use schema markup (a block of code that labels your address, services and reviews in a form machines read directly) where it matches what's visibly on the page.
None of that is new. It's the technical baseline a good SEO program builds first, and it's the same groundwork we described in how long SEO takes to start working.
What to write about
AI tools answer questions, so pages built around real buyer questions are the ones they have something to lift from. Think about what a customer asks before they call you: what it costs to get started, how the process works, what to check before they hire anyone, what usually goes wrong. Answer those directly, near the top of the page, the way you'd answer them across a desk.
Keyword density matters less here than being the page that actually answers. Repeating a phrase doesn't help a system that's reading for meaning. Being specific does. A page that names the neighborhood, the service and the kind of customer gives a model more to work with than one written in generic marketing language.
What you can't do
You can't pay for a spot in a ChatGPT answer the way you pay for a Google ad, and you can't force a citation. Placement in these answers is decided by the AI companies, so treat anyone who claims to control it with suspicion. You also can't measure it as neatly as rankings yet. Some tools show which sites they cited and some don't, so checking by hand, asking the questions your customers ask and noting who gets named, is still the most honest test available.
Where to start
Open the three pages that matter most to your business and read them as if you'd never heard of you. Is it obvious what you do, where, and for whom? Can the key facts be read as text? Does each page answer a question a customer would really ask?
If the answer to any of those is no, that's your starting list. Our SEO and AI search visibility work begins with exactly this baseline, and an account and site audit will show you where the gaps are before you spend anything on fixes.
