What homeowners actually ask ChatGPT about contractors

Contractors picture a homeowner typing "best roofer near me" into ChatGPT. That happens, but it's the minority case. The typical prompt already has a company name in it. The homeowner has a quote on the kitchen table and wants a second opinion.

Picture a plumber in Garland with one truck and 22 Google reviews, the newest from June. He bids $9,800 to repipe a 1978 ranch house.

At 10pm the homeowner types the company name into ChatGPT and asks two things: is this outfit real, and is $9,800 a rip-off. The plumber sleeps through the whole thing. He finds out how it went when the phone rings, or doesn't.

To see the pattern, WorkBrag read roughly 200 public prompts and Reddit threads where people described asking ChatGPT, Gemini or Claude about a contractor. It's a qualitative sample sorted by hand, not a survey. Four buckets covered nearly everything.

What kinds of questions do homeowners ask AI about contractors?

Legitimacy checks, about 38%. "Is this roofing company legit?" "They knocked on my door after the storm, should I trust them?" "Can you check if this plumber is licensed in Texas?" The homeowner has a name and wants a verdict.

Price checks, about 27%. "Is $14,000 fair for a 28-square roof in Fort Worth?" "My HVAC quote is $9,800, is that high?" These look like money questions. Read the whole thread and most are trust questions in disguise: am I being taken?

Recommendations, about 21%. "Who does good pool work in Frisco?" This is the search-style query contractors expect. One in five.

Comparisons, about 14%. "I have three bids, here they are, which one would you pick?" The bot is being asked to break a tie.

Sorting these by hand surprised us. We expected recommendations to be the big bucket, since that's the one every "AI search" pitch is built on, and it came in third. We'll pull another 200 prompts each month, sort them the same way, and update the shares here.

The typical prompt already has a company name in it.

Why are most of these trust questions?

The homeowner is scared of getting burned, and the numbers say that fear is earned. Roofing Contractor magazine's 2026 homeowner journey study lists unclear costs as the top pain point at 33%, with poor communication and delays at 23% each.

Fake reviews feed the same fear. A 2026 compilation by Testimonial Star puts the share of online reviews estimated to be fake at roughly 30%. A homeowner who suspects a third of what they read is planted wants a referee, and the chatbot is free at 10pm.

Nobody treats it as gospel. BrightLocal's 2026 Local Consumer Review Survey found 40% of consumers trust AI recommendations, well under the 49% who trust online reviews as much as a personal recommendation.

But 82% read AI-generated review summaries, and consumers now check an average of six review sources. AI is one more source in the stack.

Our take, and you can argue with it: the license question isn't really about the license. Almost nobody checks the number with the state. They want a neutral voice to say "this company looks normal," and the bot only says that when it finds recent, boring proof that other people hired you and lived.

What does ChatGPT read when it answers?

The same things a careful homeowner would read, only faster. Orbit Media's 2026 study of 13,184 citations found ChatGPT cites about 4.5 sources per answer and leans on primary sources and vendor pages: the Google Business Profile, review sites, and the contractor's own website.

So a legitimacy check is a review check. Ask "is this company legit" and the bot pulls up review count, review recency, the response pattern and whether the business looks alive online.

Omnibound's 2026 roundup reports that 76% of ChatGPT's most-cited pages were updated in the last 30 days. Stale proof reads as a warning. Our Garland plumber, last review in June, is getting a shrug at best.

27% asking about price are also asking about you. When the bot says a quote "looks in range," it often adds a line about the company's reviews. A fair number attached to a business with 60 recent reviews lands differently from the same number attached to one with nine.

What this means for a contractor

You'll never see these prompts. The homeowner asks the bot, gets an answer and either calls you or doesn't. Your job is to make the answer easy to give:

• Answer the legitimacy question before it is asked. License number, insurance, physical address and a live phone number on your Google Business Profile and website.
• Keep the review stream current. A review from this month tells the bot you are still working. TrueReview's 2026 benchmarks put 6-12 new reviews a month as competitive for home services.
• Put a homeowner on camera. The same Testimonial Star compilation found 88% of people trust a video testimonial as much as a personal recommendation. A face is hard to fake.
• Make asking automatic. This is the job WorkBrag does for you: one text after the job, the review and the video come back, and the clip is posted for you.

Methodology

Third-party figures come from BrightLocal's Local Consumer Review Survey 2026, Roofing Contractor magazine's "The Homeowner's Roofing Journey in 2026," Orbit Media's 2026 analysis of 13,184 AI citations, Omnibound's 2026 generative engine optimization roundup, TrueReview's 2026 review-request benchmarks and a 2026 compilation by Testimonial Star.

The question-type shares come from WorkBrag's own reading of about 200 public prompts and Reddit threads, sorted by hand into four buckets; it is a qualitative sample, not a survey.

Figures marked as WorkBrag modeled estimates are projections built from those public benchmarks and our own launch assumptions, not measured results. We will replace them with measured data from live accounts and update this page.

Frequently asked questions

How do people use ChatGPT to find a contractor?

Mostly to check a contractor they already found, not to discover one. In WorkBrag's qualitative reading of about 200 public prompts and Reddit threads, 38% asked whether a company was legitimate, 27% whether a price was fair, 21% for a recommendation and 14% for a comparison between two bids.

Do homeowners trust what ChatGPT says about a contractor?

Partly. BrightLocal's 2026 survey found 40% of consumers trust AI recommendations and 82% read AI-generated review summaries, but they still check an average of six review sources.

What does ChatGPT read when a homeowner asks if a contractor is legit?

Public proof it can fetch: the Google Business Profile, review sites like Yelp, the contractor's own website and recent social posts. Orbit Media's 2026 study found ChatGPT leans on primary sources and cites about 4.5 of them per answer.