How Contractors Get Named in AI Overviews

Homeowners are asking AI who to call. One or two contractors get named in the answer — here is how you become one of them.

A homeowner with a flooding backyard at 9:40 on a Tuesday night may not scroll search results, clicking on multiple websites looking for the best match. Instead, she types the problem into her phone and reads the AI answer. One or two contractors get named in that answer. Everyone else may as well not exist. The answer, is to be part of the “answer”.

That is the whole game behind AI Optimization GEO AEO: getting your company named when a homeowner asks an engine instead of a search box. It is not bought, and it is not a plugin. It is done by sharing the trade knowledge your crews already know.

Here is what we have learned running this for home service contractors — what the AI engines actually cite, which ones name local trades, and how to tell whether any of it is producing booked jobs.

Why AI Skips Generic Content

Answer engines ignore recycled (AI Written) copy because they have already seen it a thousand times. AI doesn’t like to learn from AI. It likes to learn from new information it’s never understood or seen before, A furnace repair page that could be dropped onto any competitor’s site with a find-and-replace on the company name has no unique information in it.

What makes an AI engine take notice is first-hand detail: the build era of the neighborhood, the city code quirk, what the tech actually found in the crawlspace. Those specifics are the difference between a page that matches a real searcher’s situation and one that matches a hypothetical.

So the rule we work by is blunt. If you use a language model to write your blogs, you do not own the answer — you are just adding to the growing pile of website conent that AI engines already skip.

Which AI Engines Cite Contractors

Gemini cites local contractors most. Because Gemini powers Google’s AI Answer engine, it would cite more local businesses and contractors than the others. Optimizing your website, keeping your Google Business Profile accurate, and generating fresh Google reviews all feed that same signal.

ChatGPT is second on the list. It uses Yelp as a primary directory for business reviews and information — before that partnership, Yelp was sliding toward irrelevance. Now Yelp reviews are worth asking for again.

Claude is third. It has no business directory and no review partnership, so it relies on crawling third-party sites. There is no special play there beyond solid website optimization and a current review profile, which is why Review Management Automation belongs in an AEO program, not beside it.

Getting Expertise Out Of Your Head

Everything starts with a recorded conversation. We run 20- to 40-minute interviews with the owner or lead tech about real completed jobs — the diagnostic call, the problem nobody else hit, the local permit headache. The audio becomes a voice-to-text transcript, and that transcript is the only source the article is written from. No outside research. No filler.

“What we’re doing at Triton Commerce, is extracting information from our clients’ heads onto an article with real customer testimonials and hyper-local details, because that’s what’s going to be cited,” Corey says. “With the old ways of SEO, you could build pages, add keywords, write blogs on a topic, and see a direct ranking outcome over time. However, with AI optimization, you need to show evidence, not keywords. Evidence of your work, craftsmanship, and customer testimonials, in order to build a clear picture for AI of your company’s service area, services, and quality of work. It’s much harder to fake it - AI is litterally giving a recommendation and by default will only recommend those with the most evedence and earn it.”

Here is how it works. We talk to the client or their staff, upload files, images and other project details. That information lands in the Authenticity Vault — a private knowledge base, with tons of information about who you are, what you do, and where. SEO Content Creation then pulls from that Vault for every article, FAQ, and case study, so one interview feeds several assets. Schema and entity work make you unambiguous. Cross-web corroboration makes you trusted.

Hyperlocal Proof Wins The Answer

Neighborhood-level specificity is what makes you the only possible answer. Finish an overlay in Boca Raton, send us the before-and-after photos, talk us through the job — and that becomes a case study interlinked to the Boca Raton city page and the overlay service page. There is a good chance nobody else in the area has written a word about Boca Raton pavement.

“When someone asks a question on that subject, you’re the answer because you’re the only one that wrote about it or you’re the one that wrote about it the most.”

The details that carry the weight are the ones only your crew has:

• Soil, drainage, and water table conditions specific to the region
• Build era and common construction for that subdivision
• Common problems, municipal codes, and permit thresholds
• HOA rules that decide what a homeowner can actually do in their yard

Ask an engine whether you can put a retention pond in your backyard under a specific HOA, and the answer has to come from somewhere. It should come from your site. That is the same structural authority we build across every trade on our Contractor Industry Solutions pages.

Tracking Citations And AI Leads

ChatGPT and Gemini now pass UTM data on website referrals, so a lead capture system that reads those parameters is the most accurate way to attribute leads from AI citations. GA4 and Google Search Console both have AI answer engine reports now. They are new, they do not show the full picture yet, and they will get better.

Lead Capture Scoring logs every call, form, chat, and text with its source and a 1-to-5 intent score, so an AI-referred lead shows up with a name beside it instead of hiding inside “direct” traffic. Expect modest volume at first — across the accounts we track, roughly one referral arrives from ChatGPT for every hundred from Google. That is exactly why it is cheap to win now.

Start With One Conversation

We are not going to hand you a strategy deck. We will get you in an interview about a job you finished, turn it into content no competitor can replicate, wrap it in schema, and show you in the dashboard whether it is producing leads. Expect meaningful ranking movement in the six-to-nine-month range, and growth that compounds after that.

If you want to see what that looks like with your own services and service area on it, Triton will build you a plan from our Winona, MN office. Call 651.321.0578 or use the Contact Triton page, and we will map the first interviews and the pages they feed.

Frequently asked questions

Can I just publish AI-written blogs to show up in AI Overviews?

No. Generative engines skip recycled, machine-written copy because they are looking for information they have not already processed. As Corey puts it, "There is absolutely no way you own that answer if you're using AI to write your blogs or write your website content."

Which AI engine cites home service contractors most often?

Gemini, because it powers Google's AI answer engine and leans on local business signals. ChatGPT is second and uses Yelp as a primary directory for business reviews, which makes Yelp reviews worth collecting again. Claude is third and relies on crawling third-party sites.

How much traffic actually comes from AI engines right now?

Across the accounts we track with UTM parameters, roughly one referral comes from ChatGPT for every hundred from Google. It is a small channel today and a growing one, which is why it is cheap to win now.

How long does this take to show results?

We generally see meaningful ranking improvement between six and nine months for an established contractor, with growth continuing after that. The work compounds — each interview makes the next piece of content faster to produce and harder for a competitor to copy.