AI Overviews and Local Business Traffic: What Contractors Should Actually Do

AI Overviews answer the question before the links. Here's what that changes for local contractors, how to check whether AI assistants recommend you, and the work that actually feeds those answers.

Listen to this article6 min · read by Monzurul Hasan
6:00

Something changed at the top of the results page, and most contractors haven’t looked at it closely yet. Ask Google a question now, something like “do foundation cracks always mean structural damage,” and there’s a good chance the answer appears written out above the links. The homeowner reads it, gets what she came for, and never opens anybody’s website.

If you’ve been buying visibility through blue links for years, that deserves a careful look before it shows up in your numbers as a mystery.

I want to make two arguments in this post. The first is that this matters much less for a contractor than the panic suggests. The second is that the work it rewards is work you should have been doing anyway, which is either comforting or annoying depending on whether you’ve been doing it.

The part that’s overstated

Think about which of your searches an AI summary can actually satisfy. “What causes slab cracks” is a reading question; a paragraph answers it completely, and the reader was probably never going to hire anyone this month. “Foundation repair near me” is a hiring question, and no paragraph satisfies it. She needs a company, a rating, a phone number, and the page still gives her the green box, the ads, and the map for exactly that reason.

The answer now appears above the links: AI Overview first, then ads and map pack, then organic results
The reading question gets answered on the page. The hiring question still needs a company.

So the honest summary is that top-of-funnel reading traffic thins while bottom-of-funnel calling traffic holds. A publisher whose business is pageviews should worry a lot. A contractor whose business is booked jobs mostly lost visits from readers, and readers were never the pipeline. There’s a real but second-order cost, since a reader today sometimes becomes a caller in two years when their crack widened, and that slow familiarity is worth something. It just isn’t worth panic.

Before you accept my word for it, check your own numbers. Search Console, last 90 days against the same 90 last year. If impressions are steady while clicks fall on question-type queries, that’s the summary answering in your place. Then look separately at queries with a city, a “near me,” or a service in them. If those held, your pipeline is intact and what changed is reading behaviour, not buying behaviour.

The ten-minute test almost nobody runs

Open ChatGPT or Gemini and ask it what a homeowner would ask. Who are the best foundation repair companies in your city. Who should I call about a crack in my basement wall. And then the uncomfortable one: is your company, by name, any good.

Three prompts to test whether AI assistants recommend your business
Run these once a quarter. Write down what you see.

If you’re named, notice what the assistant cited, because that source is doing work for you and deserves feeding. If competitors are named and you’re not, study what they have that you don’t. In my experience the gap is nearly always review volume and recency, sometimes a clearer site. And if the assistant hedges and names nobody, your whole market is unclaimed, which is the best news on this page, because the work below has no competition yet.

What feeds the answer

Here’s the annoying-or-comforting part. Assistants recommending local businesses lean on the same signals the map pack always leaned on.

Your Business Profile is the most structured, most trusted description of your business in existence, and it’s free, and if it’s stale you are invisible to the layer that reads it. Review volume and recency carry enormous weight when an assistant is asked to recommend, the same way they decide map pack position. Pages that ask one question in a heading and answer it plainly in the first two sentences get quoted; essays with the answer buried in paragraph nine don’t. Consistent name, address, and phone everywhere you appear makes you safe to recommend, and contradictions make you a risk. Named towns and real job photos give the machine something to attach you to a place.

What feeds AI answers about local businesses: business profile, recent reviews, plain question-and-answer pages, consistent listings and local specificity
The same work that wins the map pack, done properly.

Notice what’s absent from that list. Volume. Five hundred thin posts feed nothing, and a “GEO optimization” package can guarantee nothing, because nobody outside these companies knows precisely how the answers assemble. Anyone selling certainty here is reselling the “guaranteed rankings” pitch from a few years back with new letters on it. The honest work is the review system, the profile, the plain answers, and patience.

What I can’t tell you yet is how the citation economics settle. Right now assistants sometimes cite sources and sometimes don’t, the rules shift quarterly, and whether “being the cited source” becomes a durable traffic channel or a consolation prize is genuinely unresolved. I’d rather admit that than pretend a roadmap exists. What I do know is that every hedge in that sentence still points at the same to-do list.

So, this quarter

Run the three prompts and write down what you see. Fix the profile if it’s stale. Get the review habit breathing. Rewrite your five most-asked customer questions as headings with direct answers. Check your name and number match everywhere. Then re-run the prompts in three months and see if anything moved.

That work is what our SEO service does for foundation and waterproofing contractors. If you want the map first, the free market audit includes what the assistants currently say about your city and who they name instead of you. Recorded against your real market, yours to keep.

LET'S TALK
ABOUT YOUR
DIGITAL SUCCESS

Get your digital marketing plan, tailored made for your business. Share your story, vision and goals and let us do the rest.

Book A Call
Monzurul Hasan — Marketing Team Leader
Monzurul HasanBusiness Lead
  • 10x Meta Blueprint Certified.
  • HubSpot Certified Inbound Marketer.
  • 8x Google Ads & Microsoft Ads Certified.
  • Conversion Rate Optimization Professional.
  • Google Analytics & Google Tag Manager Certified.