A 2026 study of contractor AI visibility found a plumbing business with 181 Google reviews and a 4.8-star rating that scored zero on AI visibility. Zero. ChatGPT, Gemini, and Perplexity never recommended it once across dozens of test queries. The business had done everything traditional local SEO told it to do. It had the reviews, the rating, the Google Business Profile. None of that moved the needle for AI search.

The problem is not unique. In most local contractor markets, the top 10 businesses on Google Maps all have 4.5 stars or higher. Some have 200 reviews. Some have 500. When every competitor has the same review profile, reviews stop being a differentiating signal. AI platforms cannot use them to pick one business over another. They look elsewhere.

Why Reviews Stopped Being a Tiebreaker

Review count and star rating worked as signals when Google's map pack was the main discovery surface. The algorithm could rank businesses by proximity, review volume, and rating. It was a formula with three variables, and businesses could game each one.

AI search does not work that way. When a homeowner asks ChatGPT for a plumber in Bergen County, the AI does not sort businesses by review count. It reads the available information about each business across multiple sources, builds a profile, and generates a recommendation. If every business in the pool has similar reviews, that signal cancels out. The AI needs something else to make a decision.

According to a 2026 field test by researcher Ryan Doser, only 22.8% of 250 AI contractor recommendations matched Google's top 3 map results across two test cities. The AI was not copying Google's ranking. It was building its own list using different signals.

What AI Reads Instead of Reviews

When review signals cancel out, AI platforms fall back on six other signals to decide which contractor to recommend. These are the signals that separated the zero-visibility business from the businesses that actually got cited.

1. Service specificity on the business website

A plumbing company whose website says "we handle all your plumbing needs" gives AI nothing specific to cite. A plumbing company whose website has a page titled "tankless water heater installation in Hackensack" gives AI a concrete service, a concrete location, and a reason to recommend that business for that specific query. AI rewards specificity because specificity is what makes a recommendation useful to a homeowner.

2. Third-party mentions beyond Google

AI platforms read mentions across the web, not just Google. A contractor mentioned in a local newspaper article, listed on a trade association directory, or reviewed on Angi gets cited by AI more often than a contractor whose entire web presence lives on Google Business Profile and a single website. The AI is looking for consensus. If it finds the same business name across three or four independent sources, it gains confidence. One source is not enough.

3. Business description consistency

If a contractor's Google Business Profile says "HVAC repair and installation" but their website says "heating and cooling services" and their Yelp listing says "air conditioning and heating," the AI sees three slightly different businesses. Inconsistent descriptions reduce AI confidence. Businesses that use the same service description across every platform get recommended more often because AI can confirm what they do without ambiguity.

4. Recency of content and information

AI platforms factor in how recently a business's information was updated. A Google Business Profile with photos posted last week and a blog post from this month signals an active business. A profile with photos from 2023 and no website updates since last year signals a business that might not even be operating anymore. The AI has no way to call and check. It reads recency as a proxy for whether the business is still alive.

5. Local area code and address specificity

A contractor with a (201) area code and a street address in Paramus reads as genuinely local to an AI parsing location signals. A contractor with a toll-free number and a PO box reads as a lead-gen operation or a regional franchise. AI platforms are learning to distinguish between real local businesses and aggregators that pretend to be local. The businesses with verifiable local details win.

6. Structured data on the website

Schema markup gives AI platforms machine-readable information about a business: name, address, phone, services, service area, hours, and pricing. Without schema, the AI has to parse the website's HTML and guess. With schema, it gets clean structured data it can verify instantly. Businesses with LocalBusiness schema on their website get cited by AI more often because the information is easier to confirm.

How to Build the Signals That Replace Reviews

A contractor with 200 five-star reviews and zero AI visibility does not need more reviews. That well is dry as a differentiator. The contractor needs the signals AI is actually reading.

Start with the website. Create a specific page for each service in each service area. "Boiler repair in Bergen County" is better than "our services." "Emergency drain cleaning in Paramus" is better than "plumbing." Each page should answer a specific question a homeowner would ask: what the service costs, how long it takes, what areas you cover, and how to schedule.

Update the Google Business Profile weekly. Post a photo. Add a service update. Answer a review. These actions signal to AI platforms that the business is active and responsive.

Get listed on at least three directories beyond Google. Angi, HomeAdvisor, the local chamber of commerce, the trade association for the specific trade. The goal is not to get leads from these directories. The goal is to create the multi-source consensus that AI platforms need before they will recommend a business.

Add LocalBusiness schema to the website. This takes 15 minutes if the site runs on WordPress. It tells AI platforms exactly what the business does, where it is, and how to contact it, in a format they can read without guessing.

Make sure the business name, address, phone number, and service description are identical across every platform. One typo in an address on Yelp can create enough inconsistency for AI to hesitate.

The Contractor Who Got Recommended After Fixing One Thing

One HVAC contractor in the Foxxr study went from zero AI visibility to appearing in 64% of test queries after doing one thing: adding service-specific pages to the website. The business already had 90 reviews and a 4.7 rating. It already had a complete Google Business Profile. The missing piece was service specificity. The website had one generic services page. After splitting it into eight pages, each targeting a specific service and location, the AI started citing the business for queries that matched those pages.

That is the gap. Reviews tell AI a business is well regarded. Service specificity tells AI what the business does and when to recommend it. A contractor can have 500 five-star reviews and still get zero AI recommendations if the AI cannot figure out what the business specializes in.

FAQ

Do reviews matter at all for AI search?

Reviews still matter as a baseline signal. AI platforms use them to confirm a business is real and well regarded. But once every competitor has similar reviews, they stop differentiating. Reviews get a business into the pool of candidates. Specificity, consistency, and third-party mentions determine which candidate gets recommended.

How many third-party mentions does AI need before it recommends a business?

There is no fixed number. AI platforms look for consensus, which means finding the same business across multiple independent sources. Three to four sources beyond Google Business Profile is a reasonable target. A mention in a local publication, a listing on a trade directory, and a profile on Angi or HomeAdvisor gives AI enough to confirm the business.

Can a contractor with 50 reviews beat a contractor with 500 reviews in AI search?

Yes. If the contractor with 50 reviews has a more specific website, more consistent business descriptions, and more third-party mentions, AI will recommend them over a contractor with 500 reviews and none of those things. Review volume is one signal. It is not the deciding signal when the playing field is even.

How long does it take for AI visibility to improve after fixing these signals?

Most contractors see movement in AI recommendations within 60 to 90 days of making changes. AI platforms refresh their indices on rolling schedules. Changes to a website, Google Business Profile, or directory listings do not appear instantly. Consistency over a 60-day window gives the best results.

What is the fastest way to improve AI visibility?

Add LocalBusiness schema to the website, create service-specific pages for each trade and service area, and get listed on at least three directories. These three actions address the signals AI platforms weight most heavily when reviews are not a differentiator.