I wrote a few weeks ago about whether ChatGPT can find your business. Since then I’ve gone deeper, and the picture is bigger than just ChatGPT. There are now at least four AI systems that people use to find local businesses: ChatGPT, Perplexity, Google’s Gemini, and Google’s AI Overviews, which appear at the top of regular search results. Each one works differently. Each one decides who to recommend using its own logic. And most local business owners have no idea any of this is happening.

This isn’t a future problem. BrightLocal’s 2026 consumer survey found that 45% of consumers now use AI tools to find local businesses, up over 7x from the year before. Google’s AI Overviews now appear on roughly 30% of search results, including local searches. That number has been climbing every quarter.

If you only care about traditional Google rankings, you’re optimizing for a shrinking piece of the pie.

When someone Googles “plumber near me,” they get a Map Pack with three businesses and a list of blue links. They click around, maybe call a couple of places. The searcher does the comparison shopping.

When someone asks ChatGPT or Perplexity “who’s the best plumber in San Antonio,” the AI does the comparison for them. It synthesizes information from multiple sources, picks its favorites, and presents a shortlist with reasons. There’s no page two. There’s no “next 10 results.” If the AI doesn’t mention you, you don’t exist in that search.

Google’s AI Overviews split the difference. They appear at the top of a normal Google search and give a summary answer before the traditional results. A local search might show an AI Overview that says “Here are some highly-rated plumbers in San Antonio based on reviews and services offered” followed by 3-4 recommendations. If you’re not in that overview, you’re pushed below the fold even if you rank on page one.

This is why some business owners tell me their rankings haven’t changed but their call volume has dropped. The calls are going to whoever the AI recommends, and that’s not always the same businesses that rank #1 in traditional search.

Where AI search gets its information

This is the part that matters most, because it tells you what to do about it.

A study tracking 267,000 AI search citations across ChatGPT, Gemini, Perplexity, and Claude found that the most-cited source for local businesses wasn’t Google. It was Yelp, with 28% more citations than Google Business Profiles. Reddit came third. Then Angi, BBB, and Thumbtack.

That probably surprises you if you’ve spent years focused entirely on Google. But it makes sense when you think about how AI models work. They’re trained on the open web. Reddit threads where real people discuss “who’s the best roofer in San Antonio” are exactly the kind of content these models eat up. A Yelp page with 50 detailed reviews gives the AI more to work with than a Google profile with 50 star ratings and no text.

Google’s AI Overviews pull heavily from Google’s own index, obviously, but they also reference content from YouTube, Reddit, and authoritative websites. Perplexity does live web searches and synthesizes results in real time, pulling from whatever it finds. Gemini draws from Google’s full ecosystem including Maps, Reviews, YouTube, and web results.

The takeaway is that your visibility across the internet matters more than it used to. A business that only exists on Google and its own website is invisible to half the AI tools people use.

What AI models actually cite (and what they skip)

I’ve spent real time testing this across all four platforms, running the same queries and studying what gets cited. The pattern is consistent.

AI models cite specific, verifiable facts. “Roof replacement for a 2,000 sq ft home costs $8,000-$15,000 and takes 2-3 days.” That sentence contains three facts the AI can reference: price range, square footage, and timeline. A model can build an answer around that.

AI models skip vague marketing language. “We offer competitive pricing and quality workmanship.” There’s nothing to cite. No fact to reference, no number to pull, no specific claim to use. The AI moves on to the next source that gives it something concrete.

This changes what your website needs to look like. The sites getting cited in AI answers aren’t the ones with the slickest design or the best copywriting. They’re the ones with facts. Pricing ranges. Service timelines. Process descriptions. Specific cities served. Square footage thresholds. Warranty terms. Every concrete detail is something an AI can cite. Every generic sentence is something it ignores.

How to show up in each platform

The four major AI search platforms have overlapping but different source preferences. Here’s what I’ve found works for each.

ChatGPT pulls from its training data plus live web browsing when available. It heavily weights Yelp, BBB, and review platforms. To show up here, you need a presence on multiple review platforms with real, detailed reviews, not just star ratings. The text of your reviews matters because ChatGPT reads them. A review that says “Smith Roofing replaced our roof in 2 days, communicated clearly about the insurance process, and cleaned up perfectly” gives ChatGPT more to work with than a 5-star rating with no text.

Perplexity does live web searches for every query. It’s the most similar to traditional search but presents results conversationally. It loves citing specific pages that directly answer the query. A dedicated page on your website titled “How Much Does Roof Repair Cost in San Antonio” with real pricing data will get cited by Perplexity. A generic “Services” page won’t.

Google Gemini pulls from Google’s full ecosystem: your Business Profile, Maps, Reviews, YouTube, and your website. If you have a YouTube video showing a completed project with a good title and description, Gemini will find it. If your Google Business Profile has detailed service descriptions with pricing, Gemini will reference it. This is where having a complete GBP profile pays double dividends.

Google AI Overviews are generated from Google’s search index and appear right in the search results. They heavily favor content that directly answers the search query in clear, factual language. Pages structured with headers that match common questions (“How much does a new roof cost?” “How long does roof replacement take?”) get pulled into AI Overviews more often than pages with creative or clever headers.

The multi-platform checklist

Here’s the practical version. If you want to show up in AI search across all platforms, these are the things that actually matter, in priority order.

Make sure you exist on the directories AI actually pulls from. Google Business Profile, Yelp, BBB, Apple Business Connect, Angi, Thumbtack, Facebook, and Nextdoor. Apple Business Connect alone feeds Apple Maps, Siri, and Apple’s AI features across 200 countries. It’s free and takes 30 minutes. Most businesses haven’t touched it.

Make your NAP consistent everywhere. Your business name, address, and phone number need to be identical across every platform. Not similar. Identical. “Smith Roofing LLC” on Google and “Smith Roofing” on Yelp counts as inconsistent. AI models use multi-source corroboration as a trust signal. When multiple independent sources describe the same business with the same details, the AI treats that business as more credible.

Put facts on your website. Pricing ranges, service timelines, process descriptions, specific cities you serve, warranty information, licensing details. Every fact is something an AI can cite. Strip out the vague marketing language and replace it with specifics. “We serve Alamo Heights, Stone Oak, Boerne, New Braunfels, and Schertz” is 10x more useful to an AI than “We serve the greater San Antonio area.”

Create content that directly answers customer questions. “How much does roof repair cost in San Antonio?” “How long does it take to replace a roof?” “What should I look for in a roofing contractor?” Each of these is a real query people type into AI tools. A page on your site that answers it clearly, with real numbers, is exactly what AI search needs to cite you.

Get detailed reviews, not just star ratings. Encourage customers to mention specifics: the service performed, the timeline, the experience. “Great job” doesn’t help. “They replaced our roof in 2 days after the hail storm, handled the insurance claim, and were here at 7am every morning” gives AI models content to synthesize and reference.

You don’t need an AI strategy. You need a facts strategy.

Here’s the honest truth about AI search for local businesses. The businesses showing up aren’t doing anything exotic or expensive. They’re not running AI optimization campaigns or hiring consultants. They’re just putting real information on the internet in the places where AI looks for it.

Complete Google profile. Consistent directory listings. Detailed reviews with real words. A website full of specific facts instead of generic copy. That’s the playbook, and it works for traditional Google search, AI search, and whatever search channel comes next. The specifics are the strategy.

The window is open right now because most local businesses haven’t even thought about this yet. The ones that start now will have an established presence across these platforms before their competitors realize AI search matters. And by then, the bar will be much higher.

If you want to see where you stand across both traditional and AI search, I built a free audit tool that checks the things that matter for both. Takes 30 seconds.