When you ask ChatGPT, Perplexity, or Google AI for a local business recommendation, it gives you names. Not ads. Not links. Names. The businesses it recommends are not paying for that placement. Their websites are just structured so AI can read them, trust them, and cite them. Yours might not be.
Want to see it in action?
Open ChatGPT right now. Type something like "best piano teacher in [your city]" or "top-rated plumber near [your town]." You will get a short list of business names. Sometimes three. Sometimes five. Each one presented as a confident recommendation.
Now try the same query on Perplexity. You will get a similar list, but with source links. Perplexity shows its work. It pulls from review sites, directories, and business websites to build its answer.
Now try Google. Not the traditional search results. The AI Overview at the top. That box that appears before the blue links. Google's AI gives you a synthesized answer with names, descriptions, and sometimes a brief reason why each business made the list.
Three different AI platforms. Three different answers. But the pattern is the same. AI gives names, not links. It recommends, not ranks. And the businesses it names got there without paying a dime for that specific placement.
So how does AI decide who to recommend?
AI does not have a favorites list. It does not accept payments. It builds its answers from the information it can find and verify about your business. That information comes from your website, your Google Business Profile, review platforms, directories, and any other structured data it can access.
The key word is "structured." AI is not browsing your site the way a human does. It is not admiring your design or reading your About page like a story. It is scanning for specific signals.
Does your site have schema markup that clearly states your business name, location, services, and hours? Does your robots.txt file allow AI crawlers to access your pages? Do you have FAQ content structured so AI can pull direct answers? Is your content written in clear, quotable statements that AI can use in its responses?
If the answer to those questions is yes, you are a candidate for citation. If not, AI skips you. Not because your business is bad. Because your website does not speak AI's language.
What happened when we tested a real piano studio?
We ran an audit for a local piano studio called Lively Keys. The owner had a good website. Clean design. Solid content. Good reviews on Google. But when we queried AI platforms for piano lesson recommendations in that area, Lively Keys was nowhere.
The businesses AI did recommend? AllStar Music Academy. Elizabeth Farrell Music. These were the names that came up consistently across ChatGPT and Google AI. Not because they were better businesses. Because their web presence gave AI the structured signals it needed.
Lively Keys scored a 39 out of 100 on our AEO (Answer Engine Optimization) readiness assessment. The issues were clear. No schema markup. No FAQ structure. The robots.txt file was blocking AI crawlers. No llms.txt file. The content, while well-written for humans, was not structured for AI to parse and quote.
We fixed the infrastructure. Added LocalBusiness schema. Built FAQ architecture. Updated the robots.txt. Created an llms.txt file. Restructured headings and key paragraphs so AI could extract clean answers.
Two days later, Lively Keys appeared in Google AI Overviews and Gemini responses. The AEO readiness score jumped from 39 to 65. The business went from invisible to cited on two AI platforms in 48 hours.
Are the businesses AI recommends paying for that placement?
This is the part most business owners miss. There is no ad auction for AI citations. You cannot buy your way into a ChatGPT recommendation. The businesses that appear there got there because their digital presence is structured correctly.
That is both good news and bad news. Good news: you do not need a big ad budget to compete. Bad news: if your competitors have the right structure and you do not, they get recommended and you get nothing. No amount of ad spend fixes that.
What determines whether AI picks you or skips you?
It comes down to five things.
Schema markup. This is the code that tells AI exactly what your business is, where it is, what it does, and when it is open. Without it, AI has to guess. And when AI guesses, it picks someone who made it easy.
robots.txt configuration. Many websites accidentally block AI crawlers like GPTBot and ClaudeBot. If AI cannot read your site, it cannot recommend you.
FAQ structure. AI loves question-and-answer formats. When your site has clear questions with direct answers, AI can pull those answers into its responses and credit you as the source.
Content clarity. AI needs quotable sentences. Short, factual statements about what you do, who you serve, and where you are located. Long paragraphs with vague language get skipped.
Consistency across platforms. Your business name, address, phone number, and services need to match across your website, Google Business Profile, and directories. Inconsistency creates doubt. AI avoids doubt.
Is the shift to AI search already happening?
60% of Google searches now end without a click. The user gets their answer from AI. If your business is not the one AI cites, you are losing customers you never knew existed. They asked. AI answered. And your name was not in the answer.
This is not a future problem. This is happening right now, every day, in every market. The businesses that structure their web presence for AI citation will be the ones AI recommends. The ones that do not will keep wondering why the phone stopped ringing.