I Started Checking What ChatGPT Says About My Clients. Every Local Business Should.

Earlier this year I added a new step to every client audit I run: before I open Google Search Console, I open ChatGPT, Gemini and Perplexity and ask them the same question a homeowner would ask. "Who's the best electrician near [city]?" "Who should I call for AC repair in [city]?"

The results changed how I do local SEO.

Businesses I'd spent months getting into Google's Map Pack — top 3, strong reviews, steady calls — were nowhere in the AI answers. Not ranked lower. Completely absent. Meanwhile, competitors with weaker Google rankings were getting named, described and recommended.

If you run a home service business, this is the part of search you can't see happening. Nobody screenshots their ChatGPT conversation and sends it to you. The customer simply calls whoever the AI named — and it was probably not you.

The numbers behind the shift

This isn't a future trend. It's already measurable:

• 45% of consumers used an AI tool like ChatGPT, Gemini or Perplexity to find a local business in the past year — up from just 6% in 2025. Adoption is highest (64%) among 30–44 year olds: exactly the homeowners booking electricians, HVAC techs and plumbers.
• SOCi's 2026 Local Visibility Index, which analyzed 350,000+ business locations, found ChatGPT recommends only 1.2% of them — versus 35.9% that appear in Google's local 3-pack. Search Engine Land called AI local visibility "up to 30x harder" than ranking on Google.
• There's only about a 45% overlap between businesses winning Google's Map Pack and businesses AI tools recommend. Read that again: more than half the companies ranking top-3 on Google Maps don't exist as far as AI is concerned.
• Nearly a third of the US population will use generative AI search this year, per eMarketer.

And here's the number that should get every business owner's attention: ChatGPT referral traffic converts at roughly 15.9%, versus 1.76% for Google organic. When an AI names one or two businesses instead of showing ten blue links and three ads, the customer arrives pre-sold.

SEO, AEO, GEO — what actually changed

You'll see three acronyms thrown around. Here's the plain-English version:

• SEO gets you ranked in a list of results. The customer still chooses.
• AEO (answer engine optimization) gets your content selected as the answer — in Google's AI Overviews, featured snippets and voice assistants.
• GEO (generative engine optimization) gets your business cited and recommended inside AI-generated responses in ChatGPT, Gemini, Claude and Perplexity.

The mistake I see agencies make is treating these as separate services. For a local business they're one job: make your business the safest possible answer for a machine to give. Everything below serves that goal.

How ChatGPT picks local businesses

ChatGPT doesn't crawl the web the way Google does. When someone asks it for a plumber, it leans on licensed data and its search partner:

• Foursquare's Places API. Research on ChatGPT's local answers suggests 60–70% of the businesses it surfaces first come from Foursquare data. Most trades businesses have never touched their Foursquare listing — many don't know it exists. Claim it, complete it, keep it current.
• Bing's index. ChatGPT's live web results run through Bing. If you set up Bing Places and verified Bing Webmaster Tools, you're ahead of most competitors already.
• Review breadth. ChatGPT cross-references reviews from multiple platforms — Google alone isn't enough. Yelp, Facebook, BBB and trade directories all feed the picture it forms of you.

The pattern: ChatGPT recommends businesses whose data agrees everywhere it looks. One inconsistent phone number across directories is enough to make a machine skip you for a safer answer.

How Gemini and AI Overviews pick businesses

Google's AI products are the closest to classic local SEO, because they read Google's own data — your Google Business Profile, Maps reviews and the local index. If your GBP is fully built out, you're most of the way there for Gemini.

But there's a twist. Only 38% of AI Overview citations now come from pages ranking in Google's top 10 — down from 76% in mid-2025. AI Overviews increasingly cite pages that directly answer the question, regardless of their ranking. That's why every page I build now answers its core question in the first two sentences, uses real FAQ sections with structured data, and states plainly who the business serves and where.

How Perplexity picks businesses

Perplexity is the most transparent of the engines — it cites its sources inline. For local queries it leans on live web search plus data partnerships (including Yelp for local listings), and it loves community sources like Reddit. If your business gets named in a "who's a good electrician in [city]?" thread, Perplexity will find it and repeat it. Genuine community reputation — never fake posts — is a ranking factor here in a way it never quite was for Google.

How Claude picks businesses

Claude answers local questions through its web search and the open web — there's no business dashboard to claim. What works is entity clarity: a website that states your name, trade, service area and credentials consistently, schema markup that says the same thing in machine-readable form, and an llms.txt file that gives AI models a clean summary of who you are. I publish one for this site and for clients; it takes twenty minutes.

The playbook I now run for every client

This is the exact sequence, in priority order:

1. Fix your data everywhere machines look. Identical name, address, phone and website on Google, Foursquare, Bing Places, Apple Business Connect, Yelp, Facebook and BBB. This is the single highest-leverage hour you can spend.
2. Keep winning Google — it still feeds the machines. Your Google Business Profile powers Gemini and AI Overviews directly. If you're not showing up on Google Maps, fix that first.
3. Build review breadth, not just review count. The businesses AI recommends average 4.3+ stars across multiple platforms. Send every third review request to Yelp or Facebook instead of Google.
4. Answer questions in plain language on your site. FAQ sections with FAQPage schema, service pages that say who you serve and where in the first paragraph, real prices where you can. Machines quote pages that commit to answers.
5. Earn community mentions. Local press, neighborhood groups, genuine Reddit and Quora participation. One honest recommendation in a local thread outweighs a dozen directory backlinks for AI visibility.
6. Add the technical layer. Person/Service/FAQ schema, an llms.txt file, fast pages. Small effort, compounding payoff.

Notice what's not on the list: chasing every new "AI SEO tool," rewriting your site for robots, or abandoning Google. Local SEO for electricians, HVAC companies and plumbers still pays the bills — AI search optimization is the same discipline extended to more places.

What ended up mattering

After running this on client accounts for months, three things held up:

• Consistency beats brilliance. AI engines recommend the business they can verify from five angles, not the one with the cleverest website.
• Reviews are the currency. Breadth across platforms, a 4.3+ average, and recency — this is the closest thing to a "ranking factor" AI recommendations have.
• The winners moved early. Only 1.2% of businesses are getting recommended right now. That's not a wall — it's an open field. The overlap with Google is only 45%, which means the AI leaderboard in your city hasn't been claimed yet.

Five years ago the businesses that took Google Business Profile seriously before their competitors owned the Map Pack. The same window is open again — it just answers in full sentences now.

Frequently Asked Questions