The Problem
Like most home services businesses, the owner was bleeding money to missed calls — but couldn't see it on the P&L. Industry data shows 40–60% of after-hours service calls go unanswered or unconverted. For this client, the numbers were:
- •~180 inbound after-hours calls/month (5 PM–8 AM and weekends)
- •~62% abandonment rate when routed to voicemail
- •Of the 38% that left a message, only ~30% were called back within 24 hours
- •Average residential service ticket: $487
- •Average emergency/after-hours ticket: $740
Math on the hidden cost: ~111 lost qualified leads per month × ~22% typical close rate × $740 average ticket = ~$18K/month, ~$215K/year in unrecovered revenue. The owner had quoted us “maybe $30K/year” when we asked. The actual exposure was 7x that.
What he had tried: a bilingual answering service ($1,400/month) that took messages but didn't qualify, didn't book, and didn't differentiate between “AC is dead, it's 94 degrees” and “I'd like a quote next week.”
The Logic (what already existed before AI)
The owner had a clear triage logic for emergency vs. routine, a fixed diagnostic-fee structure, a service area boundary, and a dispatcher who handled bookings during the day with ~94% conversion. The logic worked. It just wasn't running 16 hours a day.
What We Built
A voice agent integrated into ServiceTitan and the existing phone system. Picks up after the second ring, 24/7. Uses the dispatcher's exact triage script and pricing logic. Books emergency calls directly into the after-hours tech's calendar, schedules routine calls for next-day, captures lead details for quotes, and texts the owner a summary if anything looks unusual.
The system was live in 5 weeks. We trained it on 60 days of recorded dispatcher calls so it sounded native, not robotic.
The Results (8 months in)
- •After-hours call answer rate: 62% → 98%
- •After-hours booking conversion: ~6% → 31%
- •New emergency jobs captured per month: +24
- •Recovered after-hours revenue: ~$28K/month, ~$340K annualized
- •Answering service eliminated: $16,800/year saved
- •Total Year 1 ROI: ~6.4x on a $52K implementation
The Flywheel
The voice agent surfaced something the owner didn't realize: ~14% of after-hours calls were upsell opportunities (existing customers wanting maintenance plans, equipment upgrades). Phase 2 expands the agent into outbound — recall calls for maintenance plan renewals and post-service NPS follow-up. Modeled additional $90K/year in plan revenue based on industry-standard 38% recall conversion.