How ChatGPT Decides Which Plumber or Roofer to Recommend

A homeowner in Salinas has a roof leak. Instead of opening Google, she opens ChatGPT and types: “Who’s the best roofing contractor near me?” ChatGPT answers her — with a name. Maybe yours. Probably not.

This used to be a niche behavior. It isn’t anymore. BrightLocal’s 2026 Local Consumer Review Survey found that 45% of consumers now use AI tools to find local services — up from just 6% the year before. A separate Scorpion national study found that 22% of homeowners now go to ChatGPT first, before Google, when they need a contractor. For a plumber in Salinas or an HVAC company on the Monterey Peninsula, that’s a real share of potential calls being routed through a system most contractors have never thought about.

The question I get from contractors all the time is: how does ChatGPT actually pick who it recommends? The answer isn’t mysterious, but it does require understanding a few things that most marketing agencies won’t bother to explain.

ChatGPT Doesn’t Show Ten Results — It Names One or Two

Google gives homeowners a list. ChatGPT gives them an answer.

That’s a meaningful difference. SOCi’s 2026 Local Visibility Index found that ChatGPT recommended only 1.2% of businesses in its study set, compared with 35.9% appearing in Google’s local 3-pack. Getting into that 1.2% isn’t a separate strategy you layer on top of everything else — it comes from the same foundations that drive strong local SEO.

But the stakes are different. If you’re not in Google’s top three, you still show up on page one. If ChatGPT doesn’t name you, you don’t exist in that conversation at all. The homeowner gets a recommendation and moves on. She doesn’t scroll.

What ChatGPT is doing, under the hood, is looking for businesses it can confidently recommend. It needs enough trust signals to clear its own internal threshold before it’ll attach a real contractor’s name to an answer. When there isn’t enough signal, it either gives a vague answer or names a competitor who left more of a digital trail. I’ve written more about this in Why ChatGPT Recommends Your Competitor and Not You if you want to go deeper on the competitive angle.

The practical takeaway: being partially visible online isn’t enough anymore. You either clear the confidence threshold or you don’t get named.

Where ChatGPT Actually Pulls Its Data

This is the part most contractors miss. Different AI platforms pull from different review sources — and they don’t all weight Google the same way.

Google AI Overviews lean heavily on Google Business Profile reviews. That makes sense; Google trusts its own data. But ChatGPT draws significantly from Yelp, BBB, and Angi. Perplexity pulls from a broader mix that includes Reddit threads and local news coverage.

For a roofing contractor in Santa Cruz or an electrician in Seaside, that means:

• A thin Yelp profile with two reviews from 2019 is a gap in the data ChatGPT uses to evaluate you
• An unverified BBB listing signals to the system that your business data hasn’t been maintained
• No Angi presence at all — even if you’ve never paid for leads there — means one more platform where you’re invisible to the AI layer

Research suggests businesses recommended by ChatGPT average around 4.3 stars across these platforms. That number matters less as a target and more as context: the businesses getting named aren’t the ones gaming the system. They’re the ones who’ve consistently asked happy customers to leave reviews, across multiple platforms, over time.

One pattern I see repeatedly with contractors in Monterey County: they’ve built a solid Google review profile but completely ignored Yelp because they don’t like the platform. That’s a legitimate frustration — Yelp’s filtering algorithm is notoriously aggressive — but ignoring it creates a real blind spot in what ChatGPT can verify about you. I touched on how AI tools source different types of online content in Reddit Shows Up in AI Search — But Not the Way You Think, which is worth reading if this platform-diversity angle is new to you.

Where AI Search Platforms Get Their Contractor Data

Different AI platforms pull from different sources. Here’s how the major ones compare for home service contractors on the Central Coast.

Business Data Completeness Is the Part Nobody Talks About

Before ChatGPT will recommend a contractor, it has to be reasonably confident that contractor is real and actively serving the area. It does that by cross-referencing your business data across platforms.

Google says businesses with complete and accurate Business Profile information are more likely to appear in local search results — and the same logic applies to the AI recommendation layer sitting on top of it. Incomplete profiles, phone numbers that don’t match across directories, or a service-area description that conflicts with what your website says all give AI systems a reason to skip you.

This matters even more for service-area businesses. A landscaper in Watsonville or a pest control company in Hollister doesn’t have a storefront address to anchor the listing. The only thing proving you serve that area is your business data — your GBP service area settings, your website’s location language, and whether your citations across platforms tell a consistent story.

Common data problems I see on contractor profiles across the Monterey Bay Area:

• Phone number on Google doesn’t match the number on Yelp or the HVAC company’s website
• GBP service area set to a single city when the contractor works across three counties
• Business hours listed as “by appointment” with no further detail, which reads as incomplete to both users and AI systems
• NAP (name, address, phone) inconsistencies left behind from old directory listings that were never updated

Fixing these isn’t glamorous work. But it’s the foundation that everything else sits on. If an AI platform can’t verify your basic information, the quality of your reviews doesn’t matter — you won’t clear the threshold to begin with. This is a core part of what we address through AI Search Sync, our approach to building local visibility across both traditional and AI-powered search.

Your Google Business Profile Is Now an AI Feed, Not Just a Listing goes into more detail on how GBP specifically feeds into AI recommendations — it’s one of the more practical reads on this topic if you manage your own profile.

What AI Platforms Need to Recommend a Local Contractor

These are the signals that AI recommendation systems use to decide whether a contractor clears the confidence threshold. Contractors strong in all three areas are the ones getting named.

Reviews Still Drive This — But the Platform Mix Has Changed

I want to be specific here because “get more reviews” is advice every agency gives and almost no one explains properly.

The review strategy that worked well in 2022 — focus everything on Google, ignore the rest — is no longer enough. ChatGPT doesn’t pull primarily from Google reviews. It cross-references Yelp, BBB, and Angi. If a plumber in Marina has 80 Google reviews and four Yelp reviews, ChatGPT sees a very incomplete picture.

The customer language in those reviews also matters more than most people realize. AI systems aren’t just looking at star ratings — they’re reading the text. A review that says “fixed our water heater same day, very professional” gives an AI system something to work with when a homeowner asks about water heater repair in Salinas. Generic reviews like “great company” contribute less signal.

This is why the content of reviews matters alongside the quantity. Your Customers Already Wrote Your Best SEO Content covers this in more detail — how the specific language customers use in reviews gets picked up by search and AI systems alike.

For contractors in Monterey County, the practical move is a simple ask: when a job goes well, tell the customer you’d appreciate a review, and mention one or two platforms — not just Google. Most happy customers will write one review. If you direct them to Google every single time, that’s all you’ll ever build.

Frequently Asked Questions About Getting Found on ChatGPT