A homeowner asks ChatGPT for a plumber, an HVAC tech, or an electrician. The model usually returns one name, maybe two, and a short reason. If that name is not yours, the call goes to someone else. Ranking on Google does not fix this. SOCi's 2026 Local Visibility Index, covering more than 350,000 locations, found ChatGPT recommended only 1.2% of them.
That is the whole problem in one number. Most shops are invisible in AI answers even when the trucks are busy and the Google listing looks fine.
ChatGPT is not a search results page
Google still shows a map pack and a list. The homeowner can scroll. ChatGPT does not work that way. It synthesizes an answer and names a business. BrightLocal's 2026 Local Consumer Review Survey found 45% of consumers now use ChatGPT or other generative AI tools for local business recommendations, up from 6% the year before. Gartner predicted traditional search engine volume would drop 25% by 2026 as people move those questions into chatbots.
For a trades shop, that means the "who do I call" question is leaving the ten-blue-links world. If the model cannot confidently match your shop as a real, consistent local business, it will not risk naming you. It will name a franchise, a directory, or a competitor whose data is cleaner. For the plain-English version of this shift, see the FirstCall guide to AI visibility.
Do AI chatbots use Google rankings?
Partly. They do not copy the local pack.
SOCi reported only about 45% overlap between businesses that rank well in Google local search and businesses that show up in AI recommendations. Scatterbranch Marketing has published similar overlap numbers in that range. A shop can sit in the top three pins on Maps and still never appear when someone asks ChatGPT "who should I call for a water heater in [town]."
The model pulls from Google Business Profile data, review text, the business website, structured data, and third-party mentions. Google rank is one input. It is not the decision. FirstCall covered that split in why a Google ranking does not mean ChatGPT recommends you. The short version: SEO still gets phone calls from people who search Google. It does not automatically put the shop inside the AI answer.
SOCi also found ChatGPT recommended locations averaging 4.3 stars. Gemini sat closer to 3.9. If the Google listing is at 3.7 with unanswered complaints about no-shows, ChatGPT has an easy reason to skip the shop.
The usual reasons a trades shop gets skipped
These are the patterns that show up again and again when an independent plumber, HVAC company, or electrician tests ChatGPT and gets someone else's name.
The Google Business Profile is often thin or messy. Hours are wrong. The primary category is "Contractor" instead of "Plumber" or "HVAC contractor." Service areas are blank. Photos are from 2019. SOCi found business profile accuracy on ChatGPT at 68%, versus 100% on Gemini in the same study. If ChatGPT's copy of the profile is stale, the model treats the shop as incomplete.
Name, address, and phone often do not match. The website says "Joe's Plumbing LLC." Google says "Joes Plumbing." Yelp has last year's cell number. AI engines resolve entities by matching those strings. When they do not match, the shop looks like two or three weak businesses instead of one strong one.
A lot of trades websites still have no LocalBusiness schema. Scatterbranch reported businesses with schema markup earning about 3.2 times more AI citations than businesses without it. A site that is just a phone number on a homepage gives the model almost nothing structured to parse. Plumber, HVACBusiness, and Electrician are valid schema types. Most independent shops still ship none of them.
Reviews also sit under the bar the model uses. ChatGPT's recommended locations in the SOCi index averaged 4.3 stars. Scatterbranch has flagged shops near 3.4 stars with low response rates as getting excluded. Volume matters less than a clean recent record and owner replies.
Franchises win citation share because they show up in trade press, manufacturer locators, and dozens of identical landing pages. 5WPR's 2026 HVAC and Plumbing AI Visibility Index found 87% of independent contractors had zero AI citation share on the prompts it tested. Independents who never get named on a local news site, a supplier page, or a how-to article give the model no third-party proof.
That does not mean the work is bad. It means the model cannot prove the shop is the same business, in the same town, doing the job the homeowner asked about.
What to check this week
Open ChatGPT, Gemini, and Perplexity in three tabs. Ask the same prompts a homeowner would actually type:
- "Who should I call for [trade] in [city]?"
- "Best [trade] near [neighborhood] who can come today"
- "Emergency [trade] in [county]"
Write down every name that comes back. If the shop is missing, the next checks are mechanical:
- Google Business Profile: correct category, current hours, service area, recent photos, review responses.
- NAP match: website footer, Google, Bing Places, Yelp, Angi. Same spelling, same suite number, same phone.
- Schema: LocalBusiness plus the trade subtype on the homepage and service pages.
- Review floor: recent 4-star-and-up reviews, replies to the bad ones.
- One page that answers the job in plain language: "water heater replacement in [city]," "AC not cooling in [city]," "panel upgrade in [city]."
Then run the prompts again in two weeks. AI recommendations move. A 2026 F9XR study put volatility in AI local recommendations around 85%. A one-time check is a snapshot, not a system.
What will not fix this by itself
Adding an llms.txt file will not, by itself, make ChatGPT name a plumbing company. Buying another directory listing will not either. Paying for Google ads does not buy a ChatGPT citation. The model is matching an entity, then scoring trust. A clean profile, consistent NAP, schema, and reviews still do the work.
FAQ
Why doesn't ChatGPT recommend my business?
Because it names very few businesses at all. SOCi's 2026 index found a 1.2% recommendation rate on ChatGPT. The usual trades causes are a thin Google profile, mismatched NAP, missing schema, a rating under the 4.3-star range ChatGPT tends to cite, and no mentions outside directories.
Do AI chatbots use Google rankings?
They read some of the same local data. They do not reprint the map pack. About 45% of the businesses that rank in Google local also show up in AI recommendations, per SOCi. The rest of the AI list is a different set of shops.
Why is a shop missing from AI answers if it already ranks on Google?
Google gives people a list. ChatGPT gives people a name. If the model cannot stitch the website, the Google profile, and third-party listings into one entity, it will not risk that name.
How long does it take to show up after fixing the listing?
There is no guaranteed clock. Profile edits can flow in days. Schema and new reviews take longer to get picked up. Retest the same prompts every month rather than waiting for a ranking report.
Can an independent shop beat a franchise in ChatGPT?
Yes, on local, high-intent prompts, if the entity is cleaner. 5WPR found franchises taking a disproportionate share of "near me" answers, but that is a data problem, not a law of physics. Consistent NAP, trade-specific schema, and recent reviews are the independent's path. A franchise fee is not required.
If the shop is missing from those prompts today, the next step is a punch list, not another SEO retainer. Run your free AI visibility audit.