Build an HVAC Missed Call Text-Back Agent with n8n, Twilio, and ChatGPT

When a homeowner's AC dies in July or their furnace quits in January, they call the first HVAC company they find. If nobody picks up, they call the next one. The company that responds first wins the job. It is that simple. Studies show that 80% of callers who reach voicemail will not leave a message and will not call back. They just move on.

I built a missed call text-back agent for HVAC companies that solves this problem completely. When a call goes unanswered, the system sends an instant text within seconds, continues the conversation over SMS using ChatGPT to qualify the issue, checks Google Calendar for availability, and books the appointment. Everything gets logged to Google Sheets. The entire system runs on n8n, Twilio, and ChatGPT with no custom code required.

In this post I walk through exactly how the system works, why HVAC is the perfect niche for this build, and how you can deliver it as a service for a strong monthly retainer.

Why HVAC Leads Are the Most Time-Sensitive in Local Services

HVAC is not like a landscaping quote or a kitchen remodel inquiry where the homeowner is comparison shopping over a few weeks. When someone calls an HVAC company, they usually have an urgent problem right now. The AC stopped blowing cold air. The furnace is making a grinding noise. There is water pooling under the unit. These are not "I'll think about it" situations. They are "I need someone here today" situations.

That urgency is exactly why missed calls are so expensive for HVAC contractors. A single missed call during peak season can represent $500 to $5,000 in lost revenue depending on whether it was a simple repair or a full system replacement. Over a summer or winter season, a busy HVAC company missing just three to five calls per day can lose $30,000 to $75,000 in potential revenue. The math is brutal and it makes the ROI case for a text-back agent incredibly easy to sell.

The respond-first dynamic also means that speed is more important than perfection. A text that arrives 15 seconds after a missed call saying "Hey, sorry we missed your call. What HVAC issue are you dealing with?" keeps that lead engaged. They are still standing next to their broken unit. They have not called the next company yet. That 15-second window is where this entire system earns its money.

How the Text-Back Agent Works End to End

The system has six stages, each handled by a different piece of the n8n workflow. Here is how the full flow operates from missed call to booked appointment.

Stage 1: Catch the Missed Call

Twilio monitors the HVAC company's business phone number. When a call goes unanswered or hits voicemail, Twilio fires a webhook to n8n with the caller's phone number, the time of the call, and the call status. The n8n workflow triggers instantly. No delay, no polling, no cron job. The webhook fires and the workflow runs.

Stage 2: Send the Instant Text-Back

Within seconds of the missed call, the system sends an SMS back to the caller through Twilio. The message is friendly and direct. Something like: "Hi, this is [Company Name]. Sorry we missed your call. What HVAC issue can we help with today?" The key is that this message goes out before the homeowner has time to search for and call another company. Speed is everything.

Stage 3: Qualify the Issue Over SMS

When the caller replies, ChatGPT takes over the conversation. The AI asks targeted questions to understand the problem: What type of system do you have? Is it cooling, heating, or both? When did the issue start? Is anyone in the home who needs immediate attention (elderly, infants)? The AI is prompted with HVAC-specific context so it asks the right questions and can triage between emergencies and routine service requests.

Stage 4: Check Calendar Availability

Once the AI has enough information to classify the job, it checks Google Calendar for available appointment slots. The system knows the company's service hours, existing bookings, and can account for drive time between jobs. It presents the caller with two or three available time options via text.

Stage 5: Book the Appointment

When the caller confirms a time, the workflow creates the calendar event in Google Calendar with all the details from the conversation: customer name, phone number, address, system type, problem description, and urgency level. The HVAC company's dispatcher sees the appointment appear in their calendar with all the context they need.

Stage 6: Log Everything to Google Sheets

Every interaction gets logged to a Google Sheet: the original missed call timestamp, the text-back send time, the full SMS conversation, the qualification details, and the appointment information. This gives the HVAC company a clear record of every lead and also gives you as the service provider a way to report on results and prove ROI.

Handling Reschedules and Follow-Ups

The system does not stop at the initial booking. If a customer needs to reschedule, they can text back and the AI handles the entire process. It checks for new available slots, confirms the change, updates the Google Calendar event, and logs the reschedule in the sheet. No human intervention required.

For leads that go cold mid-conversation (they replied once but never confirmed an appointment), the workflow includes a timed follow-up. If there has been no response for a set period, the system sends a gentle nudge: "Still dealing with that AC issue? We have a tech available tomorrow morning if you'd like to get it taken care of." This single follow-up recovers a meaningful percentage of leads that would otherwise be lost.

The Tech Stack

The entire build uses four tools. n8n handles the workflow orchestration and all the logic. Twilio handles the phone number, missed call detection via webhook, and SMS sending and receiving. ChatGPT (via the OpenAI API) powers the conversational AI that qualifies leads and manages the back-and-forth. Google Calendar handles scheduling, and Google Sheets handles logging. If you are already comfortable with n8n, you can build the full system in a day. If you are new to n8n, budget a weekend.

The running costs are minimal. Twilio SMS costs about $0.0079 per message. OpenAI API usage for the conversational piece runs a few cents per lead. Google Calendar and Sheets are free. n8n self-hosted is free, or the cloud version starts at $20/month. For a typical HVAC company handling 50 to 100 missed calls per month, the total tool cost is under $30/month. That margin makes this an incredibly profitable service to deliver.

How to Sell This as a Service

This is one of the easiest AI automation services to demo and close. The pitch is dead simple: "How many calls does your team miss every week? What if every single one of those got a text back in 15 seconds and an AI booked the appointment for you?" Every HVAC owner knows they miss calls. They know it costs them money. They just have not seen a solution that actually works.

For pricing, I recommend $500 to $1,000/month as a managed service retainer. You handle the n8n hosting, Twilio account, prompt tuning, and monthly reporting. The HVAC company sees booked appointments show up in their calendar and a monthly report showing how many leads were captured and converted. At $500/month, the service pays for itself if it books just one additional job per month. Most companies will see five to fifteen additional bookings.

The demo is what closes the deal. Call the Twilio number, let it ring, and show the HVAC owner the text arrive on your phone in real time. Then have the conversation with the AI right in front of them. When they see the appointment appear on the calendar 90 seconds after the missed call, they get it immediately. No slides needed. No ROI spreadsheet needed. The live demo sells itself.

Expanding the System

Once the missed call text-back is running, there are natural upsells. Add a review request workflow that triggers after a completed job. Add seasonal maintenance reminders that go out 60 days before filter change dates. Add a quote follow-up sequence for estimates that have not been accepted. Each addition increases the retainer value and deepens the client relationship. For a full breakdown of every HVAC automation you can offer, see our complete HVAC AI automation guide.

Watch the Full Build

I recorded a full video walkthrough of this build from start to finish. You can watch it here: Build a HVAC Text-Back Agent with n8n. The video covers every node in the n8n workflow, the Twilio configuration, the ChatGPT prompts, and the Google Calendar and Sheets integration.

If you want to connect with other people building and selling AI automation services, join the community on Skool. And if you want to see more of what we build for clients, check out Kingstone Systems.