The owner of a Novi-based HVAC company came to us in April 2026 with a single question: "How many jobs am I losing to missed calls?"
He did not know the answer. He knew he was missing calls — his voicemail had messages he could not always get to during the day — but the revenue impact was invisible. Missed calls do not show up in QuickBooks.
After 30 days of running a voice agent and lead response automation, the answer was 11 confirmed jobs. Revenue recovered: approximately $7,400.
This is the account of how it happened.
The Starting Point
The company runs 2 full-time technicians and the owner, who splits his time between the office, the field, and managing dispatch. Annual revenue is in the $600K–$800K range — a well-run small operation serving Oakland County and surrounding areas.
The problems were not unusual:
Calls going unanswered. During peak service windows — particularly mid-morning when both technicians were on-site and the owner was occupied — inbound calls hit voicemail. Some callers left messages. Some did not.
Web leads responding slowly. The company ran Google Local Services Ads that generated consistent form submissions. The owner's practice was to call back within the hour. In practice, during busy periods, callbacks happened 2–4 hours later. By then, a portion of those leads had booked with a competitor.
No visibility into the loss. There was no way to quantify what was being lost because the lost calls and cold leads left no record. The owner's intuition told him it was significant. He could not prove it.
The Audit
The $297 AI Readiness Audit mapped the owner's two highest-friction points:
- Inbound phone calls during peak hours — volume highest during 9 AM–2 PM on weekdays, exactly when field coverage was busiest.
- Web form submissions from Google LSA — averaging 8–12 per week, with a callback window that was inconsistent.
The audit identified two deployments as the highest-ROI starting points: a voice agent for inbound calls and an automated lead response system for web forms. Combined build estimate: $5,500. Estimated payback period: 30–45 days.
The owner approved both in the same conversation.
The Build
Voice agent (10 days to deploy):
The voice agent was configured with the company's specific intake flow: service address, system type (furnace, AC, heat pump, water heater), issue description, and preferred service window. The agent books directly into the owner's dispatch calendar — a simple Google Calendar integration — and sends a structured SMS brief to the owner and the dispatcher the moment a call is completed.
The agent handles after-hours calls with a modified flow: collects the same intake information, assesses urgency (no heat in winter, no AC during heat advisory), routes genuine emergencies to the owner's cell via SMS immediately, and books non-urgent calls into the next morning's dispatch window.
Lead response automation (5 days to deploy):
Google LSA form submissions trigger an immediate SMS to the prospective customer — within 60 seconds of the form submission — that acknowledges the request, confirms what was submitted, and asks one qualifying question: "Is this an urgent issue or can we schedule a service window?" The reply routes to either an emergency callback or a standard booking flow.
Total build time: 15 days. Both systems went live on May 1, 2026.
The Results: First 30 Days
| Metric | Before | After |
|---|---|---|
| Calls answered without owner/dispatcher | 0% | 94% |
| Average web lead response time | 2–4 hours | < 60 seconds |
| Confirmed missed-call jobs recovered | — | 11 |
| Estimated revenue from recovered jobs | — | ~$7,400 |
| After-hours emergency escalations handled | Manual (missed) | 3 routed correctly |
| Owner's dispatch coordination time | ~8 hrs/week | ~3 hrs/week |
The 11 recovered jobs were identified by cross-referencing voice agent call logs against bookings — callers who reached the agent, booked a service window, and completed a job that would not have existed under the old system. This is a conservative count: it does not include web leads that may have gone cold in prior months and are now converting.
The owner's comment after week four: "I used to assume I was losing some calls. I did not realize it was this many."
What Changed Operationally
Beyond the revenue recovery, two operational shifts are worth noting.
The owner's dispatch time dropped by roughly five hours per week. The voice agent handles intake, the calendar integration handles booking, and the structured briefs mean the owner is reviewing confirmed jobs rather than managing a phone queue. That time is now going to field work and customer relationship calls — higher-value uses.
After-hours calls are handled correctly for the first time. Prior to the deployment, after-hours calls went to a voicemail the owner checked the next morning. Three after-hours emergency calls in the first 30 days — two no-heat calls during an April cold snap, one water heater failure — were routed to the owner's cell immediately via SMS. All three became emergency service calls.
The Build Cost vs. the Return
| Item | Amount |
|---|---|
| Voice agent build | $3,000 |
| Lead response automation build | $2,500 |
| Platform fees (month 1) | $180 |
| Total month-one cost | $5,680 |
| Revenue from recovered jobs (month 1) | ~$7,400 |
| Net month-one return | +$1,720 |
The system paid for itself in the first month. Month two and beyond: the build cost is sunk, and the platform fees run $180/month against continued job recovery.
The Broader Point
This is not a story about AI replacing anyone. The owner still runs dispatch. The technicians still do the work. What changed is that calls and leads that were previously invisible — lost before they could even enter the pipeline — now get captured and handled.
For a service business where every inbound contact is a potential job, that capture rate is the business. The margin difference between 70% capture and 94% capture at $600K revenue is not a rounding error.
Next Step
If you run a home services business in Metro Detroit and want to understand what your missed-call rate actually looks like before committing to a build, the $297 AI Readiness Audit maps your call volume and lead sources against the same framework used here.
For businesses ready to deploy, the Lead Response Automation and Voice Agent service pages cover the full build and configuration process.
