AI Visibility for Home Services and Contractors

Homeowners ask AI assistants for a shortlist, not a page of links. The prompts sound like a call to a friend who knows the trades. Here are hypothetical examples for a mid-sized metro:

• "Who's a good emergency plumber near me that's open right now?"
• "Which roofing companies in Tulsa are licensed and insured?"
• "Recommend an HVAC contractor in Phoenix that services heat pumps."
• "What does a sewer line replacement cost in Denver, and who does trenchless?"
• "Electrician near me who can install an EV charger this week."
• "Which pest control companies in Raleigh do a one-time treatment without a contract?"
• "I need three quotes for a kitchen remodel. Who should I call first?"

The engine rarely searches your exact words. OpenAI says ChatGPT search typically rewrites a prompt into one or more targeted queries, and Google may fan a question out across subtopics. Writing a page for every phrasing is a losing game.

What these prompts share is a short list of deciding facts: location, licence, insurance, availability, service type and rough price. Those are what you need stated clearly, on your site and on the sites that list you.

Next move: write 10 to 15 prompts your own customers would ask. Cover emergencies, planned projects and price questions. This becomes your test set.

As of October 2026, a homeowner can get a contractor recommendation from at least six surfaces, and each behaves differently.

• ChatGPT searches the web at the model's discretion. Its search index is fed by OAI-SearchBot.
• Google AI Overviews and AI Mode only cite pages that are indexed and eligible to show in Search with a snippet. Google says there are no extra technical requirements beyond that.
• Gemini, Perplexity and Claude each run their own retrieval. Claude's web search is a user toggle in the consumer app.
• Microsoft Copilot grounds its answers in the Bing index.

When an engine answers without searching, it draws on training data frozen at a cutoff date. Nothing you publish this month reaches that answer. So a recent move, a new licence or a new service line can be missing or wrong in some answers for a long time.

Engines also diverge sharply. In OtterlyAI's 2026 experiment, Microsoft Copilot produced most of one site's tracked citations while ChatGPT, Gemini and Google's AI surfaces almost never cited it. A single blended score would have hidden that.

Next move: test your prompts on each engine separately, logged out, and note whether the answer shows sources. An answer with no citations may not have searched at all.

For "who should I hire" prompts, engines often cite other sites rather than the contractor's own. The evidence shows a strong association between being named on those cited sites and appearing in the answer. It does not promise an effect. Still, these are the pages an engine may be reading about you.

Check these categories:

• Home services directories and lead marketplaces. These often carry service areas and categories you set years ago.
• Review platforms. Check the business name, phone, hours and listed services, not just the star rating.
• Map and local listings, including Google Business Profile and Bing Places, since Bing feeds Copilot.
• State licensing board lookups. An engine asked "is this company licensed" may find the board record or nothing.
• Manufacturer certified-installer lists and trade association member directories.
• Local news, neighbourhood guides and "best of" roundups that name contractors.

Look for stale data: an old address, a former brand name, a county you stopped serving, an emergency line that no longer exists. Fix what you can edit and request corrections where you can't.

Next move: find out which of these sites engines cite for your prompts. A free way to see which competitors AI recommends also shows you which sources appear next to them.

AI platforms get specific facts wrong about specific businesses. For contractors, the costly errors are the ones a homeowner acts on.

Stating a specific fact on a page that crawlers can fetch gives the engine a sentence it can quote. That is the only positive experimental signal for on-page content.

Say you run an HVAC company in Columbus that stopped serving a neighbouring county last year. If your old service area page still lists that county, and three directories do too, an engine has four sources saying you serve it. Fix your page first, then the listings.

The same pattern shows up for other licensed local businesses, such as dental clinics, where credentials and locations drive the choice.

The levers with real evidence behind them remove blocks. Run these checks before you change any content.

1. JavaScript rendering. Vercel found that ChatGPT and Claude crawlers fetch JavaScript files but don't execute them. If your service areas or hours load through a widget or script, those crawlers may see nothing. Google's crawler does render JavaScript.
2. Per-bot robots.txt rules. OpenAI says each setting is independent. GPTBot governs training. OAI-SearchBot governs ChatGPT search. A blanket AI block can remove you from ChatGPT search without anyone noticing. Read more on whether to block GPTBot.
3. Firewall and CDN. A firewall can return 403 to a crawler that robots.txt allows. From 15 September 2026, Cloudflare blocks Training and Agent crawlers by default on ad-displaying pages for new domains and free-tier customers, while Search stays allowed. Check what your plan now does.
4. Snippet controls. A stray noindex, nosnippet, max-snippet:0 or data-nosnippet on a service page keeps it out of AI answers.

A minimal robots.txt that allows search crawlers while opting out of OpenAI training looks like this:

User-agent: OAI-SearchBot
Allow: /

User-agent: PerplexityBot
Allow: /

User-agent: Claude-SearchBot
Allow: /

User-agent: GPTBot
Disallow: /

The AI Crawler Checker requests a page as each AI crawler and shows which ones your firewall turns away.

The common contractor shortcut is one templated page per town: "Plumber in Westfield," "Plumber in Eastfield," and so on, with only the town name changed. AI writing tools make it easy to generate hundreds.

Google's spam policy defines scaled content abuse as "when many pages are generated for the primary purpose of manipulating search rankings and not helping users." Its first example is "Using generative AI tools or other similar tools to generate many pages without adding value for users." The policy targets volume without value, not the tool.

OtterlyAI tested the pattern in 2026 with two new domains, each loaded with roughly 1,000 AI-generated posts. Google removed one site after about 14 days and the other after about 90. Otterly reports no manual action in Search Console on either. Once Google dropped the pages, Google's AI surfaces almost never cited them, which fits Google's rule that AI Overviews only cite indexed pages. It's one small test on a low-stakes topic, but the removal was silent.

What to do instead:

• Publish one strong service area page that lists every town you serve.
• Create a separate location page only where you have something distinct to say: a real office, a crew based there, local permit rules, or completed projects.
• Delete or merge thin town pages you already have.

Week 1: access. Fetch your key pages with JavaScript off and confirm the service area, hours and licence text appear. Check robots.txt bot by bot. Ask your host or CDN whether AI search crawlers get a 200 response. Remove any accidental noindex or nosnippet.

Week 2: facts on your site. Put licence numbers, insurance, service areas, emergency terms and pricing in plain text. Make each fact match across the About, contact and footer. Merge thin city pages.

Week 3: third-party listings. Update directories, review platforms, map listings and association profiles. Make the name, phone, hours and service area match your site exactly.

Week 4: test and record. Run your 10 to 15 prompts on each engine. Log whether you're named, which facts are wrong, and which sources were cited. The free checker on this site gives a snapshot for up to three buyer questions. Repeat monthly, since answers change.

None of this guarantees a recommendation. It removes the reasons an engine can't find you or describes you wrongly. If you're weighing outside help, read whether a small business needs an AI search service and whether you need an expert or can do it in-house.

Will LocalBusiness schema get my company recommended by AI? There is no evidence it will. Google says structured data isn't required for generative AI search. Ahrefs' matched test of 1,885 pages found AI Overview citations fell 4.6% after schema was added, though those pages were already heavily cited. Existing markup does no harm, so there's no reason to remove it. Just don't treat it as a lever. Plain, fetchable text stating your facts is the better use of your time.

Should I block GPTBot on my contractor website? That depends on whether you want your content used for OpenAI model training. GPTBot only governs training. OAI-SearchBot governs whether you appear in ChatGPT search answers, and OpenAI says each setting is independent. You can block GPTBot and still allow OAI-SearchBot. The mistake to avoid is a blanket block on every AI crawler, which quietly removes you from ChatGPT search.

Can I pay to be the contractor ChatGPT recommends? Nobody can sell a guaranteed AI recommendation. The engines decide what to search and cite, and Google exposes impressions only, with no position data to verify a ranking claim. Be wary of packages that promise a share of answers. What a provider can do is fix access, facts and listings. See [whether AI share of voice packages are worth buying](/questions/ai-share-of-voice-packages) before you sign.

Do online reviews decide which contractors AI names? No published study measures that reliably. The most quoted review-profile study never defines its citation rate, so it can't carry the claim. What the evidence does show is an association between being named on the sites engines cite and appearing in answers. Reviews matter to homeowners anyway, so keep your profiles accurate and current, especially the hours, services and service area.

How often should a contractor check AI answers? Monthly is a sensible rhythm for most contractors, plus a check after any change to your service area, licence, hours or website. Engines change models and behaviour often, and answers vary from one run to the next. Treat each check as a snapshot. Run the same prompts logged out each time so results stay comparable, and record which facts were wrong.