What a Good AI Visibility Score for Home Services Actually Looks Like

There hasn't been a practical answer to the question "what AI visibility score should I be aiming for?" -- at least not for home services. You could see you were invisible. You could see that chains dominated. But "better than zero" isn't a target.

In our September 28, 2026 investigation (Scout session 152), we documented the first published percentile benchmark data for home-services AI visibility. The source is Cheers.tech, a Y Combinator-backed AI visibility tool that published panel data from 119 home-services organizations, running approximately 680,000 query passes per AI engine across 31 buying prompts over 28 days ending September 2, 2026. The benchmark metric is the percentage of tracked buying prompts in which the business appeared in an AI answer.

For the first time, there is a concrete answer to what the competitive distribution looks like.

The Benchmark Numbers

The Cheers panel breaks out AI visibility scores by percentile:

| Percentile | Visibility score | |---|---| | Top decile (90th+) | 47.4%+ | | Top quartile (75th+) | 37.3%+ | | Median (50th) | ~27.4% | | Bottom quartile (below 25th) | <15.1% |

Panel mean: 27.0%.

To be in the top quartile of home-services businesses actively tracking AI visibility, you need to appear in at least 37% of relevant buying prompts. The median -- the midpoint of the 119-organization panel -- is about 27%, or roughly one in four queries.

These numbers describe the distribution among businesses that are measured and competing. They do not describe the market at large.

Where Most of the Market Still Sits

The Cheers benchmark panel covers organizations that have invested enough in AI visibility to be using a monitoring tool and being benchmarked against peers. That is a specific slice.

In our June 12, 2026 investigation of AI citation share across HVAC and plumbing verticals (Scout session 45, `knowledge/home-services-ai-citation-data.md`), we documented the broader picture: 87% of independent HVAC and plumbing contractors in the US have effectively zero measurable AI citation share, even those with 800+ Google reviews and decades of established customer history. The US market for those two trades alone is approximately 230,000 independent operators.

The same investigation found a 1.2% ChatGPT citation rate for local business locations on contractor category queries. For the typical independent plumber or HVAC operator, ChatGPT recommends their business in approximately 1 in 100 relevant queries.

The Cheers benchmarks and the 87% invisible stat are measuring different populations. The benchmarks describe the distribution inside the group that has started optimizing. The 87% figure is the share of the overall market that hasn't started at all. Together they define the opportunity: most of your local competition is at zero, and the median among businesses that are actively working on AI visibility is 27%. Getting from zero to median is the accessible target.

Why the Gap Exists

The 87% invisible pattern doesn't come from lack of effort or service quality. Our June 12, 2026 investigation documented one of the more persistent counterintuitive findings in AI visibility research: Google review count has almost no relationship to AI citation share. Contractors with 800+ verified reviews show the same AI citation rate as contractors with fewer than 50.

The gap comes from entity infrastructure. AI platforms don't retrieve your Google review count and convert it to trust. They look for a different set of signals: is this business confirmed across multiple third-party sources as a verified, operating entity with a consistent identity? A contractor with 850 reviews and a single-source online presence is less citable than one with 30 reviews and consistent verified listings across Google Business Profile, Bing Places, BBB, and two trade-specific directories.

The mental model that wins in Google local ranking -- accumulate reviews, maintain your GBP, build domain authority -- is the wrong model for AI citation. The underlying signal is entity infrastructure, not social proof volume.

What Moves the Score by Platform

Closing the gap from zero toward the 27% median requires different work on each AI platform, because each platform weighs different signals.

**For Gemini**, the primary citation signal for local services is the business's own website. The Steady Demand AI Citation Ledger -- which tracked 14,472 citations across 50 US metros and 10 local service categories, filed in our September 23, 2026 investigation (Scout session 147, `knowledge/home-services-ai-citation-data.md`) -- found that 59.9% of Gemini's local service citations came from the business's own website. That is 6x the citation share of any single directory type. General business directories and local-service directories (Angi, Thumbtack) each accounted for roughly 10%. A contractor whose Gemini score is low should start with site structure and semantic clarity -- well-organized service pages, LocalBusiness schema, explicit geographic specification -- before investing in directory expansion.

**For ChatGPT**, the mechanism is different. Our documentation of the Yext 2026 AI Visibility Study (Scout session 14, `knowledge/platform-citation-behaviors.md`, created 2026-05-12) -- which analyzed 17.2 million citations across ChatGPT, Perplexity, and Gemini -- found that ChatGPT pulls 49% of its citations from third-party directories: Yelp, BBB, Angi, Thumbtack, HomeAdvisor, and industry registries. A contractor with a strong website but thin directory coverage will score reasonably on Gemini and poorly on ChatGPT. The fix for ChatGPT is directory infrastructure. These are independent problems requiring independent solutions.

**For Perplexity**, the primary lever is niche expert directories specific to the category -- not the general directories that work for ChatGPT. Angi and Thumbtack listing completeness matters more than broad directory volume for most contractor categories. Perplexity performs real-time web searches per query, which means freshness and crawlability affect results faster than on ChatGPT. A site with AI crawler blocks in robots.txt is not getting indexed.

The Timing Dimension

The Cheers panel reports a top decile at 47.4% as of September 2026. That ceiling is not fixed. As more contractors build entity infrastructure, competition for AI citation slots increases and the benchmark numbers shift upward.

Our June 2026 investigation of citation patterns in the roofing vertical -- which confirmed that the same 1.2% ChatGPT citation rate applies across home services trades, not just HVAC and plumbing -- documented a dynamic that applies across the market: citation patterns, once formed, compound. AI platforms return to the businesses they've cited before. Early movers establish positions that become structurally harder to displace as citation history deepens.

The contractors at 37%+ visibility today built that position while the competition was still at zero. The window to establish position before the broader market catches up exists now in a way it won't in 2028.

What to Do With the Benchmark

The Cheers benchmarks give useful framing, but they're based on a 31-prompt buying panel with a specific methodology and a 119-organization sample. The relevant number for any individual contractor is where they stand on the queries that matter for their market, service category, and geography -- not against a panel average.

A Signal Check at sourcepull.ca gives you a current citation score across ChatGPT, Perplexity, Gemini, and Claude for your specific business and location. If you're at zero on most platforms, the Cheers data gives you a concrete target: the median achieved by businesses that have started is 27%. The gap from zero to median is the Phase 1 entity infrastructure work -- not content, not schema refinement, not reviews. It is claiming and completing your listings, establishing consistent NAP data across directories, and connecting your site to the external profiles that AI platforms use to confirm your entity exists.