3-Truck HVAC Contractor Cuts Missed Jobs 74% and Recovers $8,400/Month With AI Scheduling Automation
A residential HVAC contractor running 3 service trucks in the Phoenix metro was losing $8,400 per month to no-shows — customers not home during the service window, costing a full truck roll with no revenue. Quote conversion sat at 34%, well below the 52–58% industry benchmark for responsive follow-up. AI automation built on ServiceTitan, n8n, Twilio, and GPT-4 cut missed appointments by 74%, lifted quote conversion to 49%, generated 127 new Google reviews in 90 days, and paid back the full implementation cost in 31 days.
Key Results
74%
No-show reduction
from 22% to 5.7% missed appointment rate
$8,400
Recovered per month
from filled slots and reduced truck rolls
49%
Quote conversion
up from 34% — 44% relative lift
31 days
Full payback period
implementation cost recovered in month 1
The Client
A family-owned residential HVAC contractor — anonymized at client request — operating 3 service trucks in the Chandler/Gilbert corridor of the Phoenix metro. The company: 1 owner-operator, 5 technicians, 1 office admin. Annual revenue: approximately $1.9M across AC repair, AC installation, furnace service, and maintenance agreement plans. Service area: roughly 85,000 residential homes across 4 zip codes. Primary marketing: Google Local Services Ads, Yelp, and Google Business Profile organic placement.
The company had grown steadily for 11 years primarily on referrals and repeat maintenance agreement customers. But the competitive environment had shifted — three large regional HVAC chains had entered the market in the previous 18 months with aggressive pricing, faster booking, and hundreds of Google reviews. The owner was tracking two concerning trends: no-show rates were climbing as more customers booked with multiple contractors simultaneously and went with whoever arrived first, and quote acceptance had been declining as competitors moved faster with follow-up.
The Problem: No-Shows, Cold Quotes, and Zero Review Velocity
Three distinct problems were compounding into a revenue drain. The most immediate was no-shows. Each truck ran 6 service slots per day. A 22% no-show rate meant roughly 1.3 missed slots per truck per day — across 3 trucks, that was 4 empty slots daily. Each wasted slot cost the business $187 in truck roll, tech labor, and lost appointment revenue. Over a month, this accumulated to approximately $8,400 in realized losses after accounting for the slots that got same-day fills.
No-Show Root Cause Analysis (Pre-Automation)
Customer not home, forgot window
Customer booked 3–5 days out and forgot the appointment window — most recoverable with reminder
Booked competitor who arrived first
Customer booked multiple contractors; first arrival got the job — common in emergency AC situations
Problem resolved itself
Intermittent AC issue stopped presenting — customer cancelled mentally but not explicitly
Changed mind or found cheaper quote
Price shopper who received a lower competing quote in the intervening days
The second problem was quote conversion. The office admin was manually following up on open estimates, but with 3 trucks running jobs all day, her time was consumed by dispatch coordination and inbound calls. Quote follow-up happened when she had a spare moment — typically 4–7 days after the estimate was sent. At that point, competitors who followed up within 24 hours had already won many of those jobs. The 34% conversion rate was roughly 18 points below what the company could achieve with same-day follow-up speed.
The third problem was Google reviews. The company had 71 reviews at 4.1 stars — enough to get found, but not enough to win in a market where competitors had 200–400 reviews. Techs asked customers verbally at job close, which worked maybe 30% of the time. There was no systematic post-job outreach.
The Solution: Four Automated Workflows on ServiceTitan
We built four parallel automation workflows that connected to ServiceTitan via API and handled appointment confirmation, quote follow-up, review requests, and maintenance agreement upsell — all without adding headcount or changing how techs and the admin operated day-to-day.
Tech Stack
ServiceTitan
Field service management — appointment scheduling, tech dispatch, job records, quote management, and customer contact data; all workflow triggers pull from ServiceTitan events in real time
n8n (self-hosted)
Workflow orchestration — appointment reminder scheduling, response classification routing, quote follow-up sequencing, waitlist management, and status sync back to ServiceTitan
Twilio
Multi-channel outreach — SMS reminders at 72h/24h/2h, voice call escalation for unconfirmed appointments, two-way reply handling for rescheduling and questions
GPT-4 via API
Message personalization — quote follow-up copy using job type and equipment from ServiceTitan, review request personalization with tech name and equipment, reply classification (confirm/reschedule/question/complaint)
Google My Business API
Review request link generation and monitoring — personalized review links in post-job SMS, review volume and rating tracking
Calendly
Self-service rescheduling — reschedule requests trigger Calendly link scoped to available windows per dispatcher's current schedule
Implementation: 4 Weeks to Full Deployment
ServiceTitan API Integration and Appointment Reminder System (Week 1)
Connected ServiceTitan's appointment API to n8n via webhook. Built the reminder workflow: 72-hour SMS (confirm/reschedule prompt), 24-hour SMS (second confirmation request for unconfirmed appointments + voice call trigger via Twilio for no-response), 2-hour same-day reminder with window time and tech name. Built waitlist logic: when customer reschedules or cancels, n8n pulls the waitlist queue (customers with open quotes or wellness check requests who haven't booked) and sends targeted SMS. First available slot to respond fills the spot. ServiceTitan appointment status auto-updates from confirmed replies.
Quote Follow-Up Automation (Week 2)
Built two-track follow-up logic based on ServiceTitan quote value. Small repairs under $800: Day 1 — 'Any questions about the estimate we sent for your [job type]?' Day 3 — value reinforcement with tech review reference. Day 7 — expiry notice. Large installations ($800+): Day 1 — personalized reference to specific equipment and scope quoted (pulled from ServiceTitan quote line items via GPT-4). Day 3 — financing options with calculated monthly payment. Day 10 — customer testimonial for similar installation type. Day 14 — final follow-up. All sequences pause immediately on any reply. GPT-4 classifies inbound reply intent: ready to book / price objection / not moving forward / no response.
Post-Job Review Request and Maintenance Upsell (Week 3)
Built post-completion trigger: job marked complete in ServiceTitan → 90-minute delay → personalized SMS from tech name: 'Hi [Customer], this is [Tech] from [Company]. Glad we could get your [equipment] running today — if you have a minute, a Google review helps us a lot: [link].' GPT-4 pulls tech name and equipment type from the ServiceTitan job record to build the message. One follow-up at day 5 if no review submitted. For customers who didn't purchase a maintenance agreement during the service call: 7-day follow-up sequence with next-season preparation angle and maintenance plan pricing.
Testing, Edge Cases, and Tech Team Onboarding (Week 4)
End-to-end testing across 40 synthetic appointments covering all edge cases: customer replies 'STOP,' customer replies with question, customer requests rescheduling outside available windows, customer submits emergency call that overlaps with scheduled reminder, quote follow-up sends to customer who already booked. Built escalation rules: any inbound message not classifiable with >0.7 confidence routes to office admin Slack immediately. Tech team training: 5-minute walkthrough — 'Close jobs as normal in ServiceTitan, everything else is automatic.' Admin training: Slack notification handling for escalated replies and flag management.
Results at 30 and 90 Days
5.7%
No-show rate
Down from 22% — 74% reduction. 72-hour reminders alone addressed the 'forgot the window' category (41% of prior no-shows)
$8,400
Monthly recovered revenue
From filled slots via waitlist + eliminated wasted truck rolls — consistent month-over-month improvement after day 45
49%
Quote conversion rate
Up from 34% — 44% relative lift. Large installation follow-up sequence drove most of the improvement (26% → 41%)
$31k
Additional quote revenue (90 days)
Jobs that previously went cold due to slow follow-up, recovered by automated sequence
127 new
Google reviews generated
From 71 to 198 reviews in 90 days. Average rating improved from 4.1 to 4.6 stars
31 days
Full payback period
Implementation cost recovered by end of first full month of operation
Why Home Services Businesses Win or Lose on Speed and Follow-Through
The HVAC market has consolidated dramatically in the past five years — private equity-backed regional chains have entered most major metro markets with technology infrastructure that small independents cannot replicate manually. These chains have automated appointment confirmation, quote follow-up, and review generation as standard operating procedure. An independent contractor competing against them without automation is structurally disadvantaged: the chains will follow up faster, confirm more appointments, and generate more reviews — even with identical service quality and pricing.
The no-show problem is particularly stark in HVAC because of the timing dynamics of AC emergencies. A homeowner whose AC fails on a 110°F Phoenix day will call 3–4 contractors simultaneously and book the first available window from each. They go with whoever confirms fastest and shows up first. A contractor who confirms the appointment via automated SMS 8 minutes after booking is competing against a chain whose system confirms in 3 minutes. The 5-minute difference in confirmation speed is the entire margin between winning and losing that job.
This contractor's 74% no-show reduction was not primarily about better technology — it was about making confirmation effortless. The 72-hour SMS with a single 1-tap confirm took the cognitive effort of remembering and manually calling down to nothing. The resulting no-show rate of 5.7% is better than the national residential HVAC benchmark, which Harvard Business Review data puts at 8–12% even for companies with manual reminder processes.
Frequently Asked Questions
Does this work with ServiceTitan, Jobber, or other field service software?
ServiceTitan is the primary integration platform for this implementation — it has a robust API that exposes appointment data, job status, quote records, and customer contact information in real time. Jobber, Housecall Pro, FieldEdge, and Service Fusion all have API access that supports similar integration architectures. The core n8n workflow (appointment pull → reminder sequence → response routing → CRM status update) is field service software-agnostic; the specific API connector is what changes per platform. For software without direct API access, a webhook or CSV export bridge can serve as a fallback with slightly reduced real-time accuracy.
How does the system handle same-day emergency calls vs. scheduled maintenance appointments?
Emergency calls enter a completely separate workflow. When a customer calls the main number and selects 'emergency,' or submits an emergency form on the website, n8n triggers an immediate dispatcher alert via SMS and phone — no waiting for a reminder sequence. For scheduled appointments, the reminder system activates at 72 hours, 24 hours, and 2 hours before the window. Emergency dispatch notifications go to the on-call tech directly. The two workflows are entirely parallel — a system that's excellent at reminder management for scheduled work shouldn't slow down emergency response.
What happens when a customer replies to a reminder with a question or complaint?
All inbound replies route to a unified inbox in n8n, where GPT-4 classifies the message type: confirm / reschedule request / question / complaint / wrong number. Confirms update the appointment status in ServiceTitan automatically. Reschedule requests trigger a Calendly link for self-service rescheduling. Questions are answered from a company FAQ knowledge base, then escalated to the office if confidence is below threshold. Complaints immediately route to the office manager's Slack with the full conversation context. No reply sits unacknowledged for more than 4 minutes.
How does quote follow-up differ for large installation jobs vs. small repairs?
ServiceTitan job types and quote values drive the follow-up logic. Small repairs under $800: 3-touch sequence over 7 days. Large installations: slower, higher-value sequence — personalized follow-up referencing the specific equipment quoted, financing options with specific monthly payment estimate, and a customer testimonial for similar installation. The large installation conversion rate went from 26% to 41% in this implementation, primarily because the financing message (sent on day 3) addressed the most common objection at the exact moment customers were weighing options.
What's the realistic timeline to reach a competitive Google review count?
This contractor went from 71 to 198 Google reviews in 90 days — adding 127 reviews by sending a personalized, tech-specific review request to every completed job. At 18 completed jobs per day across 3 trucks, even a 12% review request conversion rate yields 2 new reviews per day. Competing against companies with 400+ reviews typically requires 12–18 months of consistent post-job outreach — or 3–4 months with an automated system hitting every job completion.
Related
AI Automation for Home Services: HVAC, Plumbing & Contractors
AI Appointment Scheduling Automation: Cut No-Shows and Fill Every Slot
Case Study: Dental Practice Reduces No-Shows 68% With AI Scheduling
Case Study: Restaurant Group Saves 42 Hrs/Week With AI Automation
Case Study: Vet Practice Reduces No-Shows 66% With AI Reminders
See All Case Studies →
Swift Headway AI Team
Engineers and automation specialists building AI systems for SMBs across home services, healthcare, professional services, and e-commerce. This case study reflects a real client engagement; company details anonymized at client request.
Your HVAC Business
See What AI Scheduling Looks Like for Your Operation
Book a free Operations Audit. We analyze your current no-show rate, quote conversion, and review velocity to calculate exactly what automation would recover — and design the system.
Get Free Operations Audit →