A growing share of homeowners no longer Google “plumber near me” and scroll. They ask ChatGPT, Gemini, or Perplexity: “Who’s the best plumber in Charlotte?” or “My water heater is leaking — who should I call?” — and the assistant names a few companies. If you’re not one of them, you never even entered the consideration set. This is generative engine optimization (GEO), also called answer engine optimization, and it’s quickly becoming as important as showing up on the map.
The good news and the catch are the same: you can’t buy your way into an AI’s organic recommendation. Ads are starting to appear around AI answers — Microsoft Copilot, Google’s AI Overviews, and Perplexity all run them now — but no ad buys you into the recommendation itself. What you can do is shape the sources the AI reads to decide who’s worth naming — and that’s the whole game.
How an AI assistant actually picks a local business
An AI doesn’t “know” the best plumber in Charlotte. When asked, it composes an answer from what it can find — its training data plus, increasingly, live web and search results it retrieves in the moment (this is why current information shows up at all). For local services, it leans on a familiar stack of sources:
- Your Google Business Profile — name, categories, reviews, service area.
- Reviews and their volume, freshness, and rating across Google, Yelp, and others.
- Third-party listings and directories — the places that already aggregate “trusted local pros.”
- Community sources like Reddit and local forums — AI assistants and AI Overviews lean heavily on these for “best/who should I call” questions, so a genuine presence in real local threads carries weight.
- “Best [trade] in [city]” articles and roundups.
- Your own website — how clearly it states what you do, where, and for whom, plus your structured data.
The pattern is consistent: businesses that are named the same way, well-reviewed, and clearly described across many trusted sources are the ones the AI feels safe recommending. AI is conservative — it surfaces the entity it has the most consistent evidence for.
What this means you actually have to do
GEO overlaps with good SEO, but it tilts the priorities. Here’s what moves the needle:
1. Be a clear, consistent entity
The assistant has to be confident who you are. That means consistent NAP (name, address, phone) everywhere, a tight category and service description, and structured data — LocalBusiness and FAQPage schema — that spells out your services, area, and answers in a machine-readable way. (Schema demonstrably helps Google’s systems, which feed AI Overviews; its direct effect on ChatGPT or Perplexity is less proven but it never hurts.) The strongest entity signals of all are being represented consistently on Wikidata and Wikipedia where you legitimately qualify — that’s the backbone of the knowledge graph the models lean on. Ambiguity is the enemy; an AI won’t recommend a business it can’t pin down.
2. Show up in the sources AI trusts
You become recommendable by being present and consistent where the AI looks: a strong Google Business Profile, real listings on the major directories, and ideally a place in the “best in city” roundups for your trade. The more independent, trusted sources name you the same way, the more confident the model is.
3. Earn reviews — they’re an AI signal too
Review count, rating, and freshness don’t just drive Maps ranking; they’re a heavy trust signal AI leans on when it decides who to name. A steady flow of recent reviews tells both Google and the model you’re an active, trusted operator.
4. Write answer-ready content
This is the GEO-specific lever. AI composes one answer, so it rewards content it can lift directly — clear questions and answers, plain definitions, real numbers, and structure. A page that directly answers “how much does water heater replacement cost in Charlotte” in clean, extractable language is far more likely to be quoted than a vague brochure. (Every article on this site, including this one, is written that way on purpose.)
5. Make your site reachable by AI in the first place
Before any of this works, the AI has to be allowed to read you. The concrete levers:
- Let the AI crawlers in. Check your
robots.txtisn’t blocking GPTBot (OpenAI), PerplexityBot, and Google-Extended (Google’s AI training/grounding control). Many sites silently block them and wonder why they never appear. - Get indexed in Bing. ChatGPT search and Copilot retrieve heavily through Bing’s index, so submitting your site in Bing Webmaster Tools — not just Google Search Console — is a real, overlooked step.
llms.txt— low effort, unproven. It’s a proposed plain-language summary file for AI crawlers, but as of now no major assistant (OpenAI, Google, Anthropic, Perplexity) has confirmed it reads or honors it — Google’s own search team has compared it to the old keywords meta tag. Treat it as cheap insurance, not a ranking factor.
How to measure it
You can’t see an “AI ranking,” but you can audit it. Ask the assistants the way a customer would — “best HVAC company in [city],” “who should I call for a [problem]” — across ChatGPT, Gemini, Perplexity, and Google AI Overviews, and note whether you appear, how you’re described, and who shows up instead. Do it regularly. If your competitors are named and you aren’t, that’s a measurable gap, and the sources above are how you close it.
What it looks like when it works
One data point from our own business, not a client’s. Our handyman site in Charlotte is a programmatic Astro build: one page per city, per city × service, per ZIP, with LocalBusiness and Service schema on every page. In August 2026 its contact form started recording where each inquiry came from. Over the next 27 days, 3 of 13 genuine leads came from AI assistants — two from ChatGPT, one from Perplexity — and the two September ones landed on the flooring-repair service page rather than the homepage. No AI-specific trick was involved; the assistants simply found and cited the page that answered the question. The full numbers are in the Charlotte handyman case study.
The honest bottom line
There’s no button that makes ChatGPT recommend you, and anyone selling one is selling smoke. What works is unglamorous and durable: be a crystal-clear, well-reviewed entity that shows up consistently across the sources AI reads, with answer-ready content it can quote. That same work strengthens your traditional search and local ranking too — GEO and SEO pull in the same direction.
Building that entity-clear, answer-ready, AI-visible footprint is exactly our AI search optimization and local SEO work. If homeowners are asking AI who to call and your name isn’t coming up, a growth audit shows you exactly where the gap is.
Frequently asked questions
- Can I pay to be recommended by ChatGPT or Gemini?
- Not into the organic recommendation itself. Ad units are starting to appear around AI answers (Copilot, Google AI Overviews, Perplexity), but the names an assistant recommends come from the open web and its connected sources — your Google Business Profile, reviews, directories, Reddit threads, and 'best X in city' pages. You influence the recommendation by shaping those sources, not by buying placement.
- How does an AI assistant decide which local business to recommend?
- It synthesizes what it can find across the web and live search: your Google Business Profile and reviews, third-party listings and directories, 'best plumber in [city]' articles, and your own site's clarity and structured data. Businesses that are consistently named, well-reviewed, and clearly described across many trusted sources are the ones it surfaces.
- Is AI search optimization different from regular SEO?
- It overlaps but isn't identical. Both reward authority, consistency, and good content. AI search (GEO) leans harder on entity clarity, being cited across third-party sources, fresh reviews, and answer-ready content an assistant can lift directly — because it's composing one answer, not listing ten blue links.
- How do I know if AI is recommending my business?
- Ask the assistants directly — 'best [trade] in [city],' 'who should I call for [problem]' — across ChatGPT, Gemini, Perplexity, and Google AI Overviews, and see whether you appear and what they say. Track it over time. If competitors show up and you don't, that's the gap to close.