How One Plumbing Company Added $96K in Booked Revenue in 90 Days With an AI Front Office
The setup
A four-truck residential plumbing company in the Pacific Northwest. Founder-operated, $1.8M in 2025 revenue, average ticket of $880, 312 inbound calls in a typical month. They were busy enough that the founder hadn't taken a real weekend in three years — but they were also losing somewhere between 35% and 50% of their inbound calls to voicemail, mostly when the office manager was on another line or after 5pm.
"I knew we were missing calls. I just didn't want to look at the number, because I knew it would be ugly." — Founder
The audit
We ran a Pipeline Audit before the engagement. Three data points came out of it:
- Median first-response time: 22 minutes (against a 5-minute industry benchmark).
- Voicemail rate on inbound calls: 41%.
- After-hours call answer rate: 0% — the line forwarded to voicemail at 5pm.
We modeled the recoverable revenue conservatively: if we could answer 80% of currently-missed calls and convert them at the same rate the office manager converted live calls (52%), we'd add roughly $34,000/month in booked revenue. Their cost of doing nothing was approximately $400,000/year.
The build (weeks 1–3)
We installed three things, in this order:
- Primary AI receptionist on the main business line. Trained on their service menu (emergency, drain, water heater, repipe, fixture install), their pricing, their service area zip codes, and their internal call scripts. It books directly into Housecall Pro and texts the customer a confirmation.
- Missed-call text-back as a safety net. For any call the AI doesn't catch on the first ring (rare, but it happens), Twilio fires an SMS in under 8 seconds offering a calendar link.
- 24/7 after-hours coverage. For emergency calls outside business hours, the AI qualifies, books the job for the next available slot, and pages the on-call tech via SMS if the issue is flagged urgent (no water, sewage backup, gas leak).
Week 3 was parallel-run: the AI answered every call, the office manager listened to recordings, and we tuned scripts based on what she would have said differently. By day 21 the office manager was no longer on the phones; she'd moved to estimate follow-ups and review generation.
The 90-day result
| Metric | Before | After (Day 90) |
|---|---|---|
| Median first-response time | 22 minutes | 11 seconds |
| Voicemail rate | 41% | 2% |
| After-hours calls answered | 0% | 100% |
| Calls booked → job | 52% | 61% |
| Booked revenue (monthly avg) | $138,000 | $170,000 |
Net-new booked revenue across the 90-day period: $96,000.
That number is conservative — it excludes the customer reactivation campaigns we layered in around day 60, which added another ~$14,000. We're tracking those separately because we want a clean attribution number on the front office itself.
What didn't go as planned
Two things worth being honest about:
- Week 2 was rough. The AI booked three customers into the wrong service category (drain cleaning instead of fixture install) because our prompt had a vague qualification step. We caught it on day 9, retrained, and it hasn't happened since. The founder rerouted the three jobs manually and we comped him 30 days of operations.
- The office manager pushed back. Hard, for the first week. She'd been the voice of the business for nine years and watching an AI take over felt threatening. The fix was scope-shifting her role — she's now the head of customer experience, handles every escalation, owns the review program, and her job is more interesting than it was. She's also no longer drowning at 4pm on Fridays.
What you should take from this
Three things:
- The math on AI front office for a 4-truck plumber is not subtle. $96k of net-new booked revenue in 90 days against a $10k build + $2.4k/month operations fee is a 7x ROI in quarter one.
- The bottleneck was never lead generation. They had plenty of leads. They were leaking them.
- Speed compounds. Faster response → more booked jobs → more reviews → better Local Service Ads rank → more leads → which they now also answer fast.
Could you see numbers like this?
Maybe. We don't promise these exact results — we promise $50k in net-new booked revenue in the first 90 days or we keep operating free until we hit that number. The companies that get the biggest wins all share three traits: $1M+ in revenue, 50+ inbound calls per month, and a real missed-call problem you can quantify.
Apply for a Pipeline Audit → and we'll quote you in 48 hours.