Home / Guides

AI visibility for HVAC companies

The first hot week of summer, a homeowner's AC quits. They open ChatGPT and type "AC repair near me." The first cold snap, it's "furnace repair near me." Those two weeks bring a big share of your calls. You want your company in that answer.

This guide covers how ChatGPT answers a seasonal HVAC question, the service area schema an HVAC site needs, and five fixes you can make before the next season. It uses what we found when we scanned 21 local HVAC websites on October 23, 2026. No one can promise ChatGPT will name you. These checks remove the reasons it can't.

How ChatGPT answers "AC repair near me"

ChatGPT can search the web when a question needs current facts. OpenAI's help page on ChatGPT search says it can use your approximate location from your IP address and rewrite the question into a search. Their example turns "good restaurants near me" into "top restaurants San Francisco."

So "AC repair near me" becomes something like "AC repair Tampa." ChatGPT reads the pages that answer that search and cites them as sources. OpenAI says it may use partners for "nearby business listings," and that "Placement is not guaranteed."

That gives an HVAC company three jobs:

Let ChatGPT's crawler read your site, in peak season and off season.
Say what you fix (AC, furnace, heat pump) in plain text.
Say where you work, in plain text and in schema.

The service area problem

Most HVAC companies work out of one shop and drive to 10 or 20 towns. A homeowner in a suburb asks for "AC repair in [suburb]." If your site names your shop's city and no other town, you give ChatGPT no reason to match you to that suburb.

Schema fixes part of this. Schema.org has an HVACBusiness type: "A business that provides Heating, Ventilation and Air Conditioning services." It has an areaServed property: "The geographic area where a service or offered item is provided." Google's local business guide lists the fields it reads for name, address, phone, and hours.

Service area schema example

This short example is for a made-up company. Put it in a script tag of type application/ld+json on your home page. Change every value to match your own page.

{
  "@context": "https://schema.org",
  "@type": "HVACBusiness",
  "name": "Example Heating and Air",
  "url": "https://example-hvac.example/",
  "telephone": "+1-555-010-0199",
  "address": {
    "@type": "PostalAddress",
    "streetAddress": "100 Main St",
    "addressLocality": "Tampa",
    "addressRegion": "FL",
    "postalCode": "33601",
    "addressCountry": "US"
  },
  "areaServed": [
    {"@type": "City", "name": "Tampa"},
    {"@type": "City", "name": "Brandon"},
    {"@type": "City", "name": "Riverview"}
  ],
  "openingHours": "Mo-Su 00:00-23:59",
  "makesOffer": [
    {"@type": "Offer", "itemOffered": {"@type": "Service", "name": "AC repair"}},
    {"@type": "Offer", "itemOffered": {"@type": "Service", "name": "Furnace repair"}}
  ]
}

Schema is a hint, not a ranking switch. Match it to what your page says in plain text. A service area page that lists each town, with a line on what you fix in each, does the same job for readers and for AI.

What we found on 21 local HVAC websites

On October 23, 2026 we ran our free check on 21 HVAC company websites. They came from public search results in six US metros: Kansas City, Tampa, Austin, Raleigh, Columbus, and Sacramento. We are not naming them.

Finding	Sites	Share
Our checker could not read the homepage (403 or bot challenge page)	4 of 21	19%
No llms.txt	14 of 17	82%
No LocalBusiness or trade schema type	7 of 17	41%
Schema with errors, or none	6 of 17	35%
Used the HVACBusiness type	10 of 17	59%
Server took over 2.5 seconds to respond	1 of 17	6%
No sitemap found	1 of 17	6%
robots.txt blocks an AI crawler	0 of 17	0%

The 17 sites we could read scored 83 out of 100 on average. The median was 85. One scored 100. Four scored under 80. The low was 55.

Robots.txt was not the problem. No site blocked OAI-SearchBot. The two big gaps were bot shields and llms.txt. Three of the four sites we could not read sent a bot challenge page from Cloudflare or SiteGround. The fourth returned 403. A homeowner sees those sites fine. Our checker got a challenge page with no content. AI crawlers may get the same wall. And 14 of the 17 readable sites had no llms.txt file.

A real scan, name removed

One HVAC company in North Carolina scored 55. The page had 5,448 characters of readable text, and every AI crawler was allowed. But it had no robots.txt, no llms.txt, no JSON-LD schema, and no sitemap. Its first H1 heading was its phone number. A crawler could read the words but had no clean record of the business type, the towns it serves, or its hours. Each of those is a one-time fix.

What to fix before the next season

Run the free scan. It tests OAI-SearchBot and seven other AI crawlers, and tells you when a challenge page blocks the checker.
Check your bot shield. If the scan can't read your homepage, ask your host to allow OpenAI's published IP ranges. OpenAI's crawler docs say sites that opt out of OAI-SearchBot "will not be shown in ChatGPT search answers." Our AI crawlers guide lists the user agents.
Add HVACBusiness schema with areaServed. Use the example above. List the towns you drive to, your phone, and your hours.
Add an llms.txt file. An llms.txt file is a short Markdown file at yoursite.com/llms.txt. It is a proposed standard, and OpenAI does not say ChatGPT uses it. It takes ten minutes. See our llms.txt example.
Write your seasonal services in plain text. "We repair AC units and heat pumps in Tampa, Brandon, and Riverview, 7 days a week." Put it in page text, not in an image or a banner. That matches the search ChatGPT runs.

OpenAI's crawler docs say robots.txt changes take about 24 hours to show in search. Make these fixes in the off season, before the rush. Our ChatGPT search guide covers the full crawler setup.

When to hire it out

If your site runs on a builder you can't edit, or the scan shows a firewall problem, you can pay us to do it. Fix it for me is $499 one time to fix the issues your scan found, or $1,500 to rebuild the site.

If you want to know each month what ChatGPT says when people ask for AC or furnace repair in your area, the Monitor is $29 a month. It asks 25 buyer questions each week and shows who AI names: you or up to 5 competitors.

Other trades

Check your site

Scan your site free. It takes under a minute and needs no signup.

FAQ

Does ChatGPT recommend HVAC companies? Yes. ChatGPT can search the web and use your approximate location to find local results. OpenAI says placement is not guaranteed.

What schema type should an HVAC company use? Use HVACBusiness from schema.org. Add your name, phone, address, hours, and an areaServed list of the towns you serve.

What is areaServed? A schema.org property for "the geographic area where a service or offered item is provided." For an HVAC company, that is the list of towns or counties you drive to.

Why can't the checker read my HVAC site? Most of the time a bot shield (Cloudflare, SiteGround, or a host firewall) sends a challenge page to automated readers. Ask your host to allow OpenAI's published IP ranges for OAI-SearchBot.

When should I make these fixes? Before your busy season. OpenAI says robots.txt changes take about 24 hours to reach ChatGPT search. New schema and text show up after the next crawl of your site.

What score should my HVAC site get on the AI Ready Check? Aim for 85 or higher with no failed crawler checks. The HVAC sites we scanned averaged 83.

Checked 2026-10-23. Prices and policies change. Confirm on the source before you buy.

Sources

More guides

More from us: Trades With AI (AI tools for trades businesses) · SaaS Price Watch (SaaS pricing, tracked weekly) · WrapMockup (Free vehicle wrap mockups for trades) · SkillsPicker (Agent skills picked for your project)