HVAC Company Case Study: From Six AI Referred Calls a Month to Twenty Three
A six person HVAC company had six AI referred enquiries in a month. Three specific changes later, the next month brought twenty three, and nine became booked jobs.
A six person HVAC and heating company in a coastal town had been installing and servicing furnaces, heat pumps, and air conditioning units for eleven years without ever losing a customer to word of mouth. Nobody at the company had asked how many of those jobs started with someone typing a question into ChatGPT instead of calling a neighbor for a recommendation. Once someone did ask, the answer for one full month was six.
The Month Nobody Was Counting
The company started logging where every enquiry came from, call by call and form submission by form submission, for one full month. Fifty four enquiries came in during that period. Only six people mentioned an AI assistant as the reason they called, and three of those six said it almost as an afterthought.
1. What The Website Was Actually Saying
The company's site listed services as one long paragraph, furnace repair, heat pump installation, air conditioning tune ups, ductwork, general HVAC work. Every service led to the same line, contact us for a free estimate. The company's EPA certification and the brand of equipment it installed sat on an about page that had barely been opened in two years. No emergency response time appeared anywhere on the site. To a returning customer who already trusted the company, none of this mattered. To an AI assistant comparing five HVAC companies in the same town with nearly identical service pages, there was nothing to set one apart from the rest.
2. Three Changes, Made In One Week
Nothing about the company's pricing or the installations themselves changed. Three specific facts were added to the existing site over the course of one week.
- A price range for the three most requested jobs, a furnace replacement, a heat pump installation, and a seasonal tune up, stated as a range instead of contact us for a free estimate
- The brand and model line of the equipment the company installs, along with its EPA certification number, placed on the homepage instead of only on the about page
- A stated average response time for no heat emergencies in winter, broken down by the three towns the company actually services
3. The Following Month Told A Different Story
The company kept logging every enquiry the same way. Total enquiries rose to sixty eight the following month. Twenty three of them, roughly a third, specifically named ChatGPT or Google's AI Overview as the reason they contacted this company instead of one of the other HVAC businesses nearby.
| One month before the changes | One month after the changes |
|---|---|
| 54 total enquiries | 68 total enquiries |
| 6 mentioned an AI assistant | 23 mentioned ChatGPT or an AI Overview |
| No price range, no visible certification number, no stated response time | Price ranges, a visible EPA certification number, and a stated emergency response time on every service page |
4. Calls Into Booked Jobs
Not every enquiry becomes a job, and this company was no exception. Of the twenty three AI referred enquiries, nine were booked as paying jobs within the same month, four heat pump installations, three furnace replacements, and two seasonal tune ups. None of the nine mentioned comparing quotes from other HVAC companies first.
An AI assistant cannot recommend a price it was never told and a certification it was never shown.
What This Means For HVAC and Heating Companies
None of the three changes required a new website or a marketing budget. They required stating, in plain text, the facts an AI assistant needs to name one HVAC company instead of five identical looking ones, what a job actually costs, what equipment gets installed, and how fast someone answers during an emergency. HVAC companies waiting for a slow season to get around to this are the ones still showing up as six on their own version of this measurement.