2026 Industry Study
The 2026 Home Service AI Visibility Report: Who AI Recommends
An audit of 50 home service contractors across 50 competitive markets. Who Google AI Overviews cite, why traditional map rankings fall short, and how contractors can claim Share of AI Voice.
Executive Study by Brian Childers, Founder of Foxxr Digital Marketing, with research methodology led by Matt Chauhan.
tracked
100 scores
zero
Key Takeaways
Appearing in Google's AI Overview and being recommended by it are two different things. 86% of tracked businesses showed up. Only 74% held enough coverage to matter.
Seven businesses scored zero for their own core service. Three distinct failure patterns explain why: entity confusion, a missing geographic anchor, and a trust layer too thin for AI systems to work with.
A 4.8-star rating and 181 reviews did not guarantee AI visibility. One moving company proved that reputation without structured entity information leaves the AI with nothing to confirm.
Businesses that named specific places (cities, neighborhoods, counties) consistently outscored those claiming broad regional coverage. The AI matches to the most specific, verifiable thing it can find.
AI Overviews source from the full weight of a business's entity footprint, not just its website. GBP completeness, NAP consistency, third-party mentions, and service page specificity all feed the same signal.
Introduction
Home service contractors have been asking whether Google's AI recommends them. The harder question is whether it recommends them consistently across their full service area, and that gap is what this study set out to measure.
In July 2026, Foxxr tracked 50 home service businesses across 12 states and 15 trades using Local Falcon's geospatial grid tracking system. Each business was tested with one natural-language prompt written for its specific trade and location. The resulting metric, Share of AI Voice, scores AI Overview presence from 0 to 100 across the full local grid. The full methodology is at the end of this report.
AI search was changing fast, and we ran this study to get ahead of it. The result that surprised us most was a business with 181 reviews and a 4.8-star rating scoring zero on AI visibility.
Brian Childers
Founder, Foxxr Digital Marketing
The Winners and Losers of AI Search
We tracked 50 home service businesses across 12 states and 15 trades, measuring where each one appeared when Google's AI answered questions about their service and location.
A business with a score of 100 appeared every time the AI answered the tracked query. A business scoring 0 never did. The median score across the sample was 71, which looks healthy until you examine what sits on either end of that number.
The Full-Grid Problem
AI Overviews produce different answers depending on where a searcher is located within a service area. A business that appears in an AI Overview for a homeowner asking from one neighborhood may not appear for the same homeowner asking from ten miles away.
Share of AI Voice captures that variance. A score below 40 means the AI is recommending that business across less than 40% of its own service area. At that level, the commercial outcome is nearly identical to scoring zero.
The reverse is just as striking.
Most businesses in this study are findable if a homeowner already knows their name. For homeowners who don't, the AI is already deciding who gets recommended, and for 26% of this sample, that decision consistently goes to a competitor.
Full Study Results: Share of AI Voice by Business
| 1 | Custom Embroidery and Apparel Shop, St. Petersburg, FL | 100 |
| 2 | Luxury Coastal Home Builder, Tampa Bay, FL | 100 |
| 3 | Certified Mold Removal and Testing Company, Wall Township, NJ | 100 |
| 4 | Asbestos Abatement and Demolition Contractor, Howell, NJ | 100 |
| 5 | Oxygen Rental and Delivery Service, Colorado and Wyoming | 100 |
| 6 | Basement Waterproofing Contractor, Freehold Township, NJ | 100 |
| 7 | Custom Pool Builder, Boca Raton, FL | 100 |
| 8 | Landscape Installation and Garden Center, Sunbury, OH | 100 |
| 9 | Tree Service Company, Leander, TX | 100 |
| 10 | Independent Mold Inspection Company, Sacramento, CA | 97 |
| 11 | HVAC Repair and Installation Contractor, Blue Ridge, GA | 97 |
| 12 | Outdoor Teak Furniture and Wood Deck Restoration Specialist, Carlsbad, CA | 96 |
| 13 | Locally Owned Plumbing Contractor, Pinellas County, FL | 96 |
| 14 | Dog Boarding, Daycare, and Grooming Facility, Tulsa, OK | 96 |
| 15 | Exterior Cleaning and Soft Washing Company, Santa Cruz, CA | 95 |
| 16 | Children's Music School, Santa Cruz and Santa Clara Counties, CA | 92 |
| 17 | Home and Office Spring Water Delivery Company, Bay Area, CA | 91 |
| 18 | Landscape Supply Yard, Scotts Valley, CA | 91 |
| 19 | Custom Landscape Lighting Design and Installation Company, Metro Atlanta, GA | 89 |
| 20 | HVAC and Electrical Services Contractor, Pensacola, FL | 85 |
| 21 | Custom Exhibit and Retail Display Fabricator, Santa Cruz, CA | 84 |
| 22 | Irrigation and Sprinkler System Contractor, Wall Township, NJ | 83 |
| 23 | Commercial and Residential Appliance Repair Service, College Station, TX | 83 |
| 24 | Flooring Contractor, Freehold, NJ | 82 |
| 25 | Acupuncture Clinic, Santa Cruz, CA | 71 |
| 26 | Dental Practice Specializing in Cosmetic Dentistry, Buford, GA | 71 |
| 27 | Junk Removal Service, Brick Township, NJ | 71 |
| 28 | Air Duct Cleaning Company, Tampa, FL | 70 |
| 29 | Promotional Products Supplier, St. Petersburg, FL | 68 |
| 30 | HVAC and Plumbing Contractor, Norcross, GA | 64 |
| 31 | Decorative Concrete and Hardscaping Contractor, Location Withheld | 62 |
| 32 | Certified Arborist and Tree Care Company, Santa Cruz, CA | 56 |
| 33 | Plumbing Contractor, Santa Cruz, CA | 55 |
| 34 | Pool and Spa Builder, Tampa, FL | 53 |
| 35 | Custom Luxury Pool and Outdoor Living Builder, Houston, TX | 51 |
| 36 | Mold Testing and Environmental Services Company, St. Petersburg, FL | 46 |
| 37 | Pool Maintenance and Cleaning Service, St. Petersburg, FL | 44 |
| 38 | Functional Medicine and Integrative Health Clinic, Santa Rosa, CA | 39 |
| 39 | Full-Service Remodeling Contractor, San Francisco, CA | 36 |
| 40 | Roofing Contractor, Naples, FL | 31 |
| 41 | Commercial Pool Management Company, Maryland | 11 |
| 42 | AC Installation and Repair Company, St. Petersburg, FL | 3 |
| 43 | Financial Planning Firm, Walnut Creek, CA | 1 |
| 44 | Home Services Contractor, Florida | 0 |
| 45 | Hardscaping and Landscape Contractor, Dallas, TX | 0 |
| 46 | Moving Company, New Jersey | 0 |
| 47 | Fencing Contractor, Santa Cruz, CA | 0 |
| 48 | Remodeling Contractor, North Carolina | 0 |
| 49 | Residential Electrician, Location Withheld | 0 |
| 50 | Kitchen and Bath Remodeler, St. Louis, MO | 0 |
What the AI Can Confirm, It Will Recommend
The question this data raises is not whether content matters. The question is what kind of content the AI actually uses.
A blog post about mold exists on dozens of contractor websites. A service page titled "Mold Removal and Remediation Services in Wall Township, NJ" exists on one. The AI does not treat those two pieces of content as equivalent. One tells it that a business knows about mold. The other tells it that a specific business performs a specific service in a specific town, and that Google can verify it against a GBP, a review profile, and a set of third-party mentions that all say the same thing. That second business gets recommended. The first depends on whether its content happens to be specific enough for the AI to make the same connection.
That distinction runs through every high-scoring business in this study.
What the AI Needs to Confirm a Business
A mold inspection and remediation company in Wall Township, New Jersey, scored 100 in this study. Their top organic pages show exactly how that happens.
Their fourth most visited page is titled "Mold Removal and Remediation Services in Wall Township, NJ." It names the service, the town, and the state. It carries the highest engagement ratio in their top ten pages at 1.61 views per active user, meaning the people who find it stay and read it.
That page gives the AI a specific service attached to a specific location attached to a business Google can confirm through GBP, reviews, and third-party sources. When a homeowner in Wall Township asks Google's AI who handles mold remediation, that company is the answer because every signal points to the same entity.
Views per active user on the Wall Township service page, the highest in their top ten
Clicks driven by "cause of musty smell in bedroom"
Clicks from "what does black mold smell like" and "how to check for mold in an apartment"
The informational content surrounding that page does a different job. Posts like "what does black mold smell like," "how to check for mold in an apartment," and "cause of musty smell in bedroom" pull in homeowners before they are ready to call.
Those posts drive 149, 129, and 564 clicks, respectively. Each one answers a question a specific homeowner typed into Google at a specific point in their decision process. The content is written around a real query, and our client is the source.
Specificity Is the Signal
The businesses that scored zero mostly had content. What they were missing was content the AI could attach to a confirmed entity.
Examples of AI-Confirmable Content
Ten illustrative examples of the kind of specific, locatable content that gives Google's AI something to anchor a recommendation to.
| Mold Inspection · NJ | Mold inspection and air quality testing in Wall Township, completed March 2026 |
| Pool Builder · FL | Custom pool build with sunshelf and screen enclosure in Bonita Springs, completed January 2026 |
| HVAC · CA | Mitsubishi mini-split installation in a 1950s ranch home in San Diego, completed February 2026 |
| Waterproofing · NJ | Interior drain tile system and sump pump installation in a Hoboken brownstone, completed April 2026 |
| Landscape Lighting · GA | Low-voltage pathway and uplighting installation on a 3-acre property in Alpharetta, completed November 2025 |
| Tree Service · FL | Emergency oak removal after hurricane damage in St. Petersburg, completed October 2025 |
| Flooring · CA | LVP installation throughout a 2,200 sq ft home in Irvine, completed March 2026 |
| Plumbing · NJ | Tankless water heater conversion in a 1970s colonial in Cherry Hill, completed December 2025 |
| Appliance Repair · TX | Commercial refrigeration repair for a Dallas restaurant, same-day service, completed February 2026 |
| Environmental Testing · NJ | Post-remediation air quality clearance testing in Toms River, completed January 2026 |
What Kept Seven Businesses Out of AI Search Entirely
Seven businesses in this study were invisible to Google's AI for their own core service. A homeowner asking the right question in the right city got an answer that never included them. We looked at each zero-score business individually across their GBP status, review volume and quality, website structure, service page specificity, and entity consistency across indexed sources. Three patterns came into focus.
The Three Failure Patterns
Entity Confusion
One business lost to a competitor carrying a near-identical name in the same market. When two entities share a name and operate in the same geography, AI systems resolve the ambiguity toward the one with the stronger, cleaner signal profile. This business had reasonable on-page SEO. The AI could not determine with confidence which entity to recommend and defaulted to the competitor whose signals were less ambiguous.
Fixing entity confusion requires differentiating the business across every indexed source: the GBP name, the website, directory listings, review profiles, and any third-party mentions. Every source has to point to the same distinct entity before the AI can recommend it with confidence.
Missing Geographic Anchor
Three businesses had no findable Google Business Profile. Two had no GBP at all. One had a GBP that was not indexed in Google's local entity graph. The result in each case was the same: the AI had no confirmed geographic anchor to attach the business to. AI Overviews source heavily from Google's local entity graph when constructing recommendations for service queries. Without a confirmed location in that graph, the AI defaults to competitors that have one.
A claimed, verified, and completely filled GBP with the correct primary category is the minimum requirement for geographic anchoring.
Thin Trust Layer
Three businesses had GBPs, but the trust layer surrounding them was too thin for the AI to work with confidently. One had zero reviews and an incomplete GBP. One had 24 reviews but no findable GBP in search results. One had a well-designed website with no reviews at all and incomplete GBP information.
The AI cross-references what a business claims about itself against what third-party sources confirm. A business with no reviews, no third-party mentions, and incomplete GBP information gives the AI nothing to cross-reference. It picks competitors it can verify.
The Reputation Paradox
The most instructive zero-score result in the study sits outside all three patterns.
A moving company in this study had 181 Google reviews and a 4.8-star rating. The AI did not recommend them once for their core service. The explanation sits on the website. The About page was too short to convey meaningful entity information.
The service pages lacked the structured specificity the AI needs to match a business to a query. The copy was written to impress visitors rather than to give a search engine factual, confirmable information about what the business does, where it operates, and for whom.
The AI knew the business was liked. Without a website that told it what the business does and where, that reputation had nowhere to attach.
Google reviews
Average rating (stars)
Share of AI Voice
What the AI Needs Beyond Your Website
The AI does not take a business at its own word. It looks for confirmation from sources that have no reason to say favorable things about a business unless they are warranted.
A chamber listing says the business is a real, participating member of the local economy. A manufacturer contractor finder says the business is certified to install that brand. A local news mention says the business did something noteworthy enough for a journalist to write about. A directory listing says the business submitted its own information to a form. The AI weighs those sources differently.
Why Local Authority Works Differently for AI Search
Traditional SEO treated backlinks primarily as authority signals. More links from higher-authority domains produced higher rankings. That logic still applies in organic search, but AI Overviews operate on a different layer. They are constructing answers. To construct a confident answer, the AI needs sources it can cross-reference.
A business with 40 directory citations and no local authority sources is a business the AI can confirm exists. A business with 10 directory citations, a chamber listing, a manufacturer page, and two local press mentions is a business the AI can confirm exists, operates in a specific geography, performs a specific service, and has been recognized by sources outside its own control. That second profile produces a more confident recommendation.
Three things local authority sources provide that directories do not:
Geographic context. A listing in the Pinellas County Chamber of Commerce tells the AI this business operates in Pinellas County, Florida. A listing on Yelp tells the AI the business submitted a form. Local pack and AI Overview results weight geographic confirmation signals. The chamber carries that weight. The directory does not.
Category verification. A listing on a manufacturer's contractor finder, a Bradford White water heater dealer page, or a GAF certified roofer directory confirms the business installs that brand and has been vetted to do so. A generic directory listing is a self-claimed category. The manufacturer link is third-party verified.
Earned credibility. Chambers, trade associations, and local press cover real businesses through real relationships and editorial judgment. Google's systems have spent years learning to distinguish sources that require something real from sources that accept any submission. The AI inherits that distinction.
Examples of Legitimate Local Authority Sources
The right sources vary by trade and market. These are the categories that carry the most weight across the trades and states represented in this study.
| Chamber/business association | Local and county chambers, regional business councils, BBB accreditation |
| Manufacturer/vendor | Bradford White, Carrier, GAF, Mitsubishi Electric, Hunter Industries, Pentair, Hayward |
| Trade association | PHCC, NARI, ACCA, ABC, NALP, NRCA, IEC |
| Local and regional media | City business journals, local news outlets, neighborhood publications, hyperlocal blogs |
| Community and nonprofit | Sponsor listings, charity partner pages, local sports organization pages, neighborhood association sites |
| Government and civic | City vendor lists, licensed contractor registries, permit holder directories |
What This Section Does Not Claim
Directories alone are not enough to build confirmable authority that drives full-grid AI visibility. A complete and consistent citation profile across Google, Yelp, Bing Places, Apple Maps, and relevant trade directories is still a baseline requirement for local entity consistency. The distinction is between a floor and a ceiling. Directories get a business to the floor. Local authority sources are what push the score higher.
Participation matters more than presence. A chamber listing from a business that attends events, sponsors programs, and maintains an active profile carries more signal weight than a listing from a business that paid the membership fee and never engaged.
Will AI Replace Local SEO?
Ahrefs research found that Google's AI Overviews now result in a 58% lower average click-through rate for top-ranking pages, up from 34.5% just eight months prior. That number gets cited often as evidence that AI is dismantling local search. The more precise reading is that clicks to ranked pages are down, but being named in an AI Overview tells a different story.
Seer Interactive research from November 2025 found that brands cited in AI Overviews earn 35% more organic clicks and 91% more paid clicks compared to brands that are not cited. The businesses losing traffic are the ones ranked but not cited. The businesses gaining ground are the ones the AI recommends by name.
Lower average click-through rate for top-ranking pages, up from 34.5% eight months prior
More organic clicks for brands cited in AI Overviews
More paid clicks for brands cited in AI Overviews
What Changes When Search Becomes an Answer
The most important shift in AI search is not which signals matter. It is how many get checked simultaneously. When a homeowner asks Google's AI who handles mold remediation in their city, the system constructs an answer by cross-referencing the business's GBP, website, reviews, third-party mentions, and any indexed sources that confirm or contradict what the website claims about itself.
Ahrefs describes this process as query fan-out: one prompt becomes multiple hidden searches checking services, locations, costs, credentials, reviews, and risks before the AI produces a response. A business with gaps in any of those layers gives the AI less to work with. A competitor with complete coverage gets cited instead.
| One question becomes many hidden searches | Your content needs to cover services, locations, costs, credentials, and common customer questions. A single homepage with a contact form addresses none of them. |
| AI systems retrieve short passages | Every section of every page should state the useful thing directly. The answer goes first. The detail follows. |
| Being cited matters beyond the click | Seer Interactive found brands cited in AI Overviews earn 35% more organic clicks and 91% more paid clicks. The recommendation carries commercial weight before a homeowner clicks anything. |
GEO Is Entity Consistency at Scale
Generative engine optimization, or GEO, is the practice of making a business easier for AI systems to understand, retrieve, and cite. For a local home service contractor, it comes down to one question: does every source that indexes your business agree on who you are, what you do, and where you do it?
A business whose website claims one service area, whose GBP lists a different phone number, whose directory listings use an old address, and whose service pages describe the work in generic terms is a blurry entity. The AI has to make inferences rather than confirmations. It defaults to a competitor whose signals are consistent.
In this study, the nine businesses that scored 100 all shared one characteristic beyond their individual signals: their information was consistent and confirmable across every indexed source. The AI confirmed them and recommended them.
A Note from Matt, Foxxr's Research Lead
My job was to work out what people actually ask when they need one of these services, then check whether each business shows up when an AI answers. I pulled the seed phrases from Local Falcon's Local Keyword Research tool and ran them through Gemini to see how a real person's question gets worded.
Nobody types the tidy keyword we would put in a title tag. They ask full questions. Then I tracked every business in Local Falcon's grid so I could see where they landed in the AI Overview across the whole map, not just at a single pin.
What got me was the gap between showing up and showing up everywhere. On paper, 86% appeared, but once I looked at how much of the map each one actually held, it dropped to 74% with any real coverage. A couple only surfaced at a single point. Even in a set we watch closely, one in seven did not come up at all for their own core service. For a contractor, that is the whole thing: AI answers are already picking who gets recommended, and it is not the same lineup as the old map pack. If you are not showing up when someone asks Google's AI for the best in your trade, you are not in the running. Most owners have no clue which side of that line they are on.
Matt Chauhan
Research Lead, Foxxr Digital Marketing
Conclusion
The pattern in this study is straightforward. Every business we tracked had a website. Most had content. Several had strong review profiles. Fourteen percent were invisible to Google's AI for their own core service anyway.
The gap is not about effort. It is about confirmability. The AI does not reward the businesses that worked hardest on their websites. It rewards the businesses whose full entity, website, GBP, reviews, service pages, and third-party mentions all point to the same thing and give it something concrete to match against a real homeowner query.
The businesses that built that profile, whether deliberately or not, scored 100. The businesses that left gaps in it scored zero.
The Main Takeaways
AI visibility is concentrated. Nine businesses out of 50 held a perfect Share of AI Voice score of 100. Seven scored zero for their own core service. The middle is large, but the distance between the top and the bottom is not gradual.
Appearing is not the same as being recommended. Eighty-six percent of tracked businesses appeared in an AI Overview. Only 74% held enough coverage to matter commercially. Three businesses scored between 1 and 9, present in isolated corners of their service area and nowhere else.
The failures follow three patterns. Entity confusion, a missing geographic anchor, and a thin trust layer account for every zero-score business in the study. Each pattern has a different fix, and treating them as one problem produces the wrong corrective action for most of them.
Strong reputation signals alone are not enough. One business in this study had 181 reviews and a 4.8-star rating and scored zero. Reviews tell the AI a business is well regarded. The website and entity footprint have to tell it what the business does and where.
The standard got higher. The game did not change.
The practical conclusion is this: winning AI search means making a business easier to find, easier to confirm, and easier to recommend than the competitors around it.
For a home service contractor, that means a complete GBP, service pages built around specific entities, reviews that carry real information, and third-party sources that confirm the business exists beyond its own website. The contractors who build those things now hold the strongest positions when a homeowner asks Google's AI who to call.
Methodology, Limits, and Source Notes
This study tracked 50 home service businesses across 12 states and 15 trades. All scans were collected in July 2026. The sample consists almost entirely of current Foxxr clients, with a small number of former prospects included to reach 50.
Geographic distribution is weighted: roughly two-thirds of the sample sits across California, Florida, and New Jersey, with Georgia and Texas next and a handful of single-state entries rounding out the rest.
Trades are spread thin, with most represented by one or two businesses. The "12 states, 15 trades" description is accurate but reads more national in scope than the sample's actual distribution warrants.
| Businesses tracked | 50 |
| Collection window | July 2026 |
| Sample composition | Foxxr clients and former prospects |
| States represented | 12, weighted toward California, Florida, and New Jersey |
| Trades represented | 15, most with one to two businesses each |
| Tracking tool | Local Falcon geospatial grid tracking |
| Grid size | Mostly 9x9, some 7x7 |
| Grid radius | 2 to 9 miles per point, varied by market density |
| Prompts | One custom natural-language prompt per business, no shared template |
| Prompt source | Local Falcon seed keywords rewritten via Gemini |
| Scan method | Server-side, session-independent, no personal browser or login |
| Share of AI Voice | Calculated automatically by Local Falcon |
| Appeared definition | Named directly in AI Overview summary, source-link appearances excluded |
| Meaningful coverage threshold | Score of 40 or higher, judgment call not an external standard |
| Exclusions | None, every scanned business included |
| Named businesses | High scorers only, zero scorers anonymized |
Prompt Construction
Each business received one custom natural-language prompt written for its specific trade and location. Seed keywords came from Local Falcon's Local Keyword Research tool. Those keywords were then rewritten using Gemini into the full conversational question a homeowner would type or speak.
AI Overviews trigger most reliably on longer natural queries rather than clean keyword phrases. Every prompt across the 50 businesses is different. No shared template was used.
Grid-Based Tracking
AI Overview presence was measured using Local Falcon's geospatial grid tracking system. Grid sizes varied by business, with most running a 9x9 grid and at least one running 7x7. Grid radius per point ranged from 2 to 9 miles depending on the market's population density and competitive landscape. Scans were run server-side by Local Falcon, not from a personal browser or logged-in Google account, eliminating personalization bias from the results.
Share of AI Voice
Share of AI Voice is calculated automatically by Local Falcon. Of the grid points where an AI Overview containing brand names appeared, it measures the percentage where the tracked business was named directly. Grid points with no AI Overview, or an Overview listing no business names, are excluded.
A business appearing only as a source link does not count toward its score. Scores were not adjusted post-collection.
Meaningful Coverage Threshold
A business was classified as holding meaningful AI visibility at a Share of AI Voice score of 40 or higher. That threshold is a judgment call, not a standard defined by Local Falcon or any external body. It represents the point at which a business is appearing consistently enough across its service area that AI recommendation becomes a plausible lead source.
Important Caveats
Source Files
The working dataset includes 50-business scan data with Share of AI Voice scores, Local Falcon grid tracking results, custom prompts for all 50 businesses, and Local Falcon Local Keyword Research seed terms. Foxxr retains the underlying data. Reporters, trade press, and industry publications evaluating this report can request access to verify specific findings by reaching out at foxxr.com.
Who Built This Study
This research was led by Brian Childers, Founder of Foxxr Digital Marketing. Foxxr is a digital marketing agency serving home service contractors across the United States, with services including AI search optimization, GEO, local SEO, web design, and paid advertising.
The research and grid-based tracking methodology was designed and executed by Matt, Foxxr's research lead, using Local Falcon's geospatial AI Overview tracking system and Gemini for prompt construction.
If this report raises questions about where your business stands in AI search, visit foxxr.com or reach out directly. We run the same grid-based AI visibility audit for contractor markets across the country and show you exactly where you appear, what your Share of AI Voice score is, and what is driving it.
At Foxxr, we believe the contractors who understand this shift earliest will hold the strongest positions when AI search finishes reshaping how homeowners find local services. This report is our contribution to that understanding. More research is coming.