← All work

Home Services · AI Voice

Sample writeup

Every after-hours call answered, every job booked.

A multi-location HVAC group was missing 40% of inbound calls outside business hours and on busy days. We embedded with their dispatch team, deployed a voice agent across 6 locations, and turned the leak into bookings inside 5 weeks.

Client
Anonymized · Multi-Location HVAC Group
Industry
Home Services
Region
Ontario, Canada
Duration
5 weeks

The challenge

Six locations, one shared phone tree, and a dispatch team that could only answer two lines at once. After-hours calls went to voicemail; 62% of callers never called back. The group was spending $9k/month on ads to drive calls, and missing 4 in 10 of the calls they paid for.

The approach

Identify. Build. Adopt.

01

Identify

Week one: we sat with dispatch, listened to 50 real calls across the 6 locations, and tagged every missed-call pattern by time-of-day, technician load, and call type. We costed the leak with their CRM data: $34k/mo in bookable-but-missed work.

02

Build

Voice agent deployed on a parallel number first, then ported. Trained on their actual call corpus and their dispatch SOPs. Books emergency jobs straight into the calendar with auto deposit-collection. SMS follow-up sequences for callers who don't book on the call.

03

Adopt

Two-week side-by-side rollout. Agent runs alongside dispatch, every booking reviewed for the first 5 days. We embedded daily in their Slack until the dispatch team trusted the agent enough to let it own the overflow line.

Outcomes

What changed, measurably.

$28k/mo
Recovered after-hours revenue
100%
Calls answered, 24/7
<4s
Average answer time
5 wks
From embed to live across 6 locations
Services used
AI Voice AgentsAI Workflow Automation
Every call we used to miss now becomes a booking or a follow-up. The ad spend finally has somewhere to land.
Owner
Multi-Location HVAC Group

Got a similar leak?

30 minutes with the founders. Bring us the workflow that hurts most. We'll cost the leak live and tell you whether the same playbook deploys here.