The After-Hours On-Call Playbook for Plumbing Shops
After-hours emergencies are the highest-margin work in residential plumbing. A 2 AM burst-pipe call that becomes a $4,800 emergency replacement is a 60β90% margin job. The shops that capture those calls grow. The shops that send them to voicemail stay small.
But running 24/7 service is also the fastest way to burn out a small team. The on-call playbook here is what weβve seen work across shops weβve installed for β a structure that captures the calls without exhausting the people.
Step 1 β Define your true-emergency criteria
Be explicit about what wakes the on-call tech. The default the AI receptionist asks:
- Water shut off and still flooding β yes = emergency
- Gas smell β yes = call utility emergency line first, then us
- Sewage backup into living areas β yes = emergency
- No hot water + below-freezing weather β yes = emergency
- No water at all + customer canβt shut off main β yes = emergency
Everything else (slow drain, dripping faucet, smell from drain, water-heater pilot out in summer) β books for next-day service.
Step 2 β Build your on-call rotation
A typical 4-tech residential shopβs rotation:
- Primary on-call β most senior emergency-qualified tech, this week
- Backup on-call β secondary qualified tech
- Owner β final fallback if both primary and backup miss
Rotation runs weekly, Sunday-to-Sunday. Compensate per-call ($50β$100 per emergency answered) AND per-hour on-site ($25β$40 above standard rate). Most techs accept on-call duty 1 week in 4 if the compensation is fair.
Step 3 β Wire the AI receptionist to qualify
The AI receptionist picks up in under 1 ring. Three qualifying questions in this order:
- βIs the water shut off, or is it still flowing?β
- βDo you smell gas?β
- βIs water actively flooding the property right now?β
Any βyesβ β escalation to on-call. All βnoβ β book into next-day dispatch calendar.
This 3-question check takes 18β25 seconds. Itβs polite. Itβs fast. The homeowner appreciates the speed.
Step 4 β Configure the escalation rotation
The escalation timing:
- Primary on-callβs phone rings for 90 seconds
- No answer β backup on-callβs phone rings for 90 seconds
- No answer β ownerβs phone rings (no timeout β vibrates until answered)
- Still no answer after 5 minutes total β AI tells customer: βAll our technicians are on calls right now. Someone will call you back within 30 minutes. Please stay safe β if water is flooding, do you know where your main shut-off valve is?β Then AI texts every tech in the shop with the job details.
This 4-tier escalation captures ~99% of emergencies. The 1% that fall through get a personal callback when the owner sees the morning report.
Step 5 β Set the customer-facing experience
The customer should feel that you have their emergency under control from the first ring. The AIβs tone matters more than its words. Our default opening:
βHi, youβve reached [Shop Name]βs 24-hour emergency line. Iβm an AI assistant β Iβll get your information and connect you to a plumber as fast as possible. First β is the water shut off?β
After qualifying:
βGot it. Iβm sending a plumber to your address now. Youβll get a text in the next 60 seconds with their ETA and a tracking link. Please stay on the line for one more question β whatβs the best phone number to reach you in case the tech needs to call?β
Confirmation SMS sent within 15 seconds of call end. Tech ETA SMS sent within 60 seconds of dispatch.
Step 6 β Track the metrics
Weekly report we generate for shops we install for:
| Metric | Healthy range |
|---|---|
| Total after-hours calls | 6β18 / week |
| Emergency / non-emergency split | 30% / 70% |
| Dispatch time (call answered β ETA texted) | Under 90 seconds |
| Customer satisfaction (post-job survey) | 4.7+ stars |
| Plan-enrollment offer accept rate | 18β25% |
If any metric drifts out of range, we adjust. Common adjustments:
- High emergency rate β tighten qualifying questions (homeowners over-classify)
- Low dispatch time β tech rotation isnβt picking up β check phone settings + Do Not Disturb exceptions
- Low CSAT β the AIβs tone is off, or after-hours pricing not clearly disclosed
Step 7 β Adjust based on real data
After 30 days, review patterns and refine. Specific adjustments shops typically make:
- Add βHow many fixtures are affected?β β distinguishes a single leak from whole-house flood
- Add after-hours pricing disclosure β most shops add a $75β$150 after-hours dispatch fee, mentioned to the customer before booking
- Tighten qualifying questions for specific seasons β winter adds βis it below freezing where you are?β
- Add a βdo you have a maintenance plan?β lookup β plan members get prioritized routing
Common pitfalls
Pitfall 1 β Letting the AI book any emergency. Some shops let the AI book everything as emergency to be safe. Result: techs get woken up for slow drains at 2 AM, burn out, quit. Solution: trust the qualifying questions.
Pitfall 2 β Not training techs on the SMS workflow. Tech gets the dispatch SMS, doesnβt see the βOn the wayβ button, calls the dispatcher to confirm. Wastes 15 minutes. Solution: 30-minute training, then test-fire emergency dispatches at 7 PM Monday.
Pitfall 3 β Forgetting Do Not Disturb exceptions. On-call techβs phone is on DND for the night, never hears the call. Solution: configure DND to allow calls from the shopβs main dispatch number + the AI receptionistβs outbound number.

