arrow_backBack to Blog
AI for TradesAugust 2026·7 min read
AI for HVAC Businesses: Three Automations You Can Set Up Before Next Monday
Three AI automations every HVAC owner can set up this week, with real tools and real setup times: (1) an after-hours missed-call text-back using a tool like Missed Call Text Back or your CRM's built-in SMS, takes about 90 minutes to configure; (2) a review request sequence triggered automatically after a job is marked complete in ServiceTitan or Jobber, about two hours to build; and (3) an AI-powered estimate follow-up email that writes a custom note for each prospect based on the job type, maybe one hour once you have a solid prompt. None of these require a developer. None cost more than $50 a month in tool fees. All three can run while you're on a job site Friday afternoon.
HVAC owners get pitched AI tools constantly. Most of that pitch is about chatbots that "handle your entire customer journey" or dashboards that "give you insights." That is not this post.
These three automations were chosen because they solve the three problems that cost HVAC shops the most money right now: lost leads from missed calls, low review counts that hurt search rankings, and follow-up that never happens because the tech is already running the next call. Each one is specific, testable, and reversible if it doesn't fit your operation.
Someone calls at 9 PM about a no-cool. You're not answering. They call the next guy on Google. You lose the job without ever knowing you had a shot at it.
A missed-call text-back fires an SMS within two minutes of a missed call. Something like: "Hey, this is [Company]. We saw you called and we don't want you left hanging. We'll have someone reach out first thing in the morning, or you can reply here with what's going on." That message keeps the lead warm. A lot of the time, people will reply with their address and system details before you even call them back.
Tools to use: If you're on GoHighLevel, this is native. If you're on Jobber, check their automation tab. If neither, Make.com can connect your phone system to Twilio for SMS in an afternoon. The Make.com automation template library has a Twilio flow that takes about 45 minutes to configure if you've never used it before.
According to Josh Cowen at Applied AI Works, HVAC shops that implement this single automation typically recover two to four after-hours leads per week that would have gone cold. At an average ticket of $400 to $600, that math closes fast.
Most HVAC owners know they should ask for reviews. Almost none of them do it consistently. The tech is tired at the end of a long day. The office person forgets. And the window for a good review is short: about an hour after the job closes, the customer is still warm.
Here's the setup: in Jobber or ServiceTitan, create an automation trigger that fires when a job status changes to "complete." That trigger sends a two-message SMS sequence. Message one goes out within 30 minutes of job close: a short thank-you with a direct Google review link. If no response in 48 hours, message two sends: a softer nudge. No more than two messages. Ever.
The direct Google review link is the part people miss. You can get your specific review shortlink from your Google Business Profile dashboard under "Ask for reviews." Drop that link in the message. Don't make the customer search for where to leave a review. Every extra click drops response rate significantly.
If you're not on Jobber or ServiceTitan, you can build this in Make.com with a Google Sheets trigger: technician marks the job done in a shared sheet, Make.com fires the SMS via Twilio. Clunkier, but it works.
You send an estimate. You hear nothing. You mean to follow up but three days go by and it feels awkward. The lead goes cold.
This automation uses AI to write a short, specific follow-up email for each estimate, based on the job type and the customer's situation. You're not blasting a generic "just checking in" email. You're sending something that references their actual system or the specific work quoted.
The setup: write an AI prompt that takes four inputs: customer name, job type (new install, repair, tune-up), system type (if known), and days since estimate was sent. The prompt produces a three-sentence follow-up email in plain, conversational language. You review it, adjust if needed, send. Takes about 90 seconds per estimate instead of five minutes of staring at a blank compose window.
If you want to automate it further, Make.com can pull open estimates from Jobber on a daily schedule, pass the details to your AI via the API, and draft the emails into a shared doc for your office person to review and send. Full version takes about three hours to build. The manual version, just copying job info into ChatGPT, Claude, Gemini, or your favorite AI, takes one hour to set up and runs on discipline after that.
A sample prompt starter: "Write a short, friendly follow-up email from an HVAC company to a homeowner who received an estimate [X] days ago for [job type] on a [system type] system. The tone should sound like a real person, not a corporation. Three sentences max. Do not start with 'I hope this message finds you well.'"
If you're completely new to this, start with the missed-call text-back. It is the fastest to set up, has the clearest ROI, and touches a problem you feel every single week. Get that running. Then come back for the review automation.
If you want someone to look at your whole operation and tell you exactly where AI will save you the most time and money, that's what the AI Audit at Applied AI Works is built for. It's a $1,500 engagement. You get a specific roadmap for your business, not a generic checklist. Or if you just want to talk it through first, start with a free call.
Rooting for you,
Josh
The tools that actually get used in HVAC shops right now are Make.com for automation workflows, ChatGPT for writing tasks like follow-up emails and estimate summaries, Jobber or ServiceTitan for triggering automations off job status changes, and Twilio for SMS when your CRM doesn't have native texting. Otter.ai is also useful for recording and summarizing tech debriefs in the field. None of these require a tech background to start.
For the three automations in this post, expect $40 to $80 per month in tool costs: Make.com at roughly $10 to $16 per month depending on usage, Twilio SMS costs based on volume (typically $5 to $20 for a small shop), and ChatGPT Plus or Claude at $20 per month if you're using it manually. ServiceTitan and Jobber have their own subscription costs but likely already include automation features you're not using.
AI phone agents exist and are improving. Tools like Air.ai and Synthflow are being tested by some service companies. For most HVAC shops with five to fifty employees, a fully AI-answered phone line is not the right starting point. A missed-call text-back covers the most important gap, after-hours and overflow calls, without the risk of an AI mishandling a customer with an urgent situation.
No, and that is not what these automations are built to do. The goal is to remove repetitive tasks from their plate so they can focus on things that require judgment: handling upset customers, scheduling complex jobs, keeping techs on track. Applied AI Works is explicit about this: AI handles the predictable work, people handle the unpredictable.
The missed-call text-back can show results within the first week. Review volume typically picks up within two to four weeks of a consistent automated request sequence. Estimate follow-up results depend on your close rate and lead volume, but most shops notice the difference within a month. These are operational fixes that work quickly or tell you something useful about your process.
The AI audit gives you a specific roadmap for your operation. Not general ideas.
Book Your AI Audit — $1,500arrow_forward