A 4-truck plumbing shop in Dallas ran our snapshot for 14 days and recovered 11 emergency jobs that would have hit voicemail under their old setup. Average ticket on those 11: $612. Total recovered revenue in two weeks: $6,732. The snapshot cost $997. That’s the math we want every operator to run before they buy.
The seven automations below are the ones doing that work. They are not theoretical. They are the seven workflows we look at first when an operator says “I don’t know if I’m missing calls.”
1. AI receptionist with on-call rotation awareness
The single biggest revenue leak in a plumbing shop is the after-hours call that rolls to voicemail. A burst pipe at 11:47 PM does not leave a voicemail. It calls the next shop on the Google map and books with whoever picks up first.
The snapshot’s AI receptionist answers on ring one — 24/7/365. It does four things in sequence:
- Asks if this is a plumbing emergency (yes / no branch).
- If yes: collects address, ZIP, callback number, brief issue description.
- Checks the on-call rotation calendar (we configure this during install) and texts the on-call tech with the lead summary.
- Sends the homeowner an SMS confirmation with the tech’s name and an ETA window.
If the call is non-emergency, the AI offers to book a next-business-day appointment directly on the calendar. No tech wakes up unless they need to.
2. 60-second callback for form submissions
When someone fills out the contact form on your website at 9:14 AM, what happens? In most shops we audited, the form emails an inbox that the owner checks twice a day. By the time they call back, the homeowner has booked elsewhere.
This automation triggers on form submission and does the following inside 60 seconds:
- Texts the homeowner: “Hi [Name], this is [Shop] — got your message about [issue]. Quickest way to lock in a slot is to reply YES to this text and we’ll send you booking options.”
- Notifies the dispatcher in the GHL mobile app with a tap-to-call button.
- Creates a contact, tags it with the form source, and drops it into the “New lead” pipeline.
Reply rate to that first text in the shops running it: 42%. Booking rate on those replies: 61%. The customer is texting back inside their first three minutes of being a lead.
3. Voicemail-to-text rescue
For the calls the AI receptionist doesn’t catch — usually because a homeowner specifically asks to talk to “a person, not a robot” — voicemails get transcribed via Twilio and pushed into a dedicated “Voicemail” pipeline with the transcription pinned to the contact.
The dispatcher sees a one-line preview: “Burst pipe in basement, 30314, call me back, this is John, 404-555-0102” and can return the call from inside GHL with a single tap. The whole pipeline gets cleared by 7:45 AM most mornings.
4. Emergency-tag SLA timer
Not every plumbing call is an emergency, but the ones that are have a 12-minute clock on them. If you don’t respond inside 12 minutes, the homeowner has called two more shops and the first one to confirm a truck gets the job.
The snapshot tags inbound leads with a severity label (emergency / urgent / standard) based on keyword detection in the call transcript or form text. Emergency-tagged leads:
- Get the dispatcher notified on three channels (push, SMS, email) simultaneously.
- Start a visible countdown timer in the pipeline view.
- Auto-escalate to a backup dispatcher number at the 8-minute mark if no action.
- Page the owner’s phone at the 11-minute mark.
We have not seen an emergency lead miss the 12-minute window in any shop running this workflow correctly.
5. Photo-quote turnaround inside 20 minutes
This isn’t an emergency workflow per se, but it’s an emergency-context one — because the moment a tech is standing in front of a burst pipe or a corroded water heater, the homeowner needs a price now or they’re going to call someone else.
The tech opens the GHL mobile app, snaps three photos of the unit, picks the job template (“Water heater — 50 gal gas”), and the workflow:
- Auto-generates a branded PDF with the photos, the price (pulled from your saved price book), the warranty details, and the financing options.
- Emails it to the homeowner.
- Texts them: “Quote for the work we just looked at — link is in your email. Reply YES to lock in install for Friday.”
- Drops the contact into the “Estimate sent” pipeline with a 7-day drip nurture if it doesn’t close on day one.
Close rate on photo-backed quotes in the shops running this: 2.1× baseline.
6. After-hours holiday-surge mode
Christmas Eve, the Sunday of a 4th of July weekend, the night of an ice storm — these are when emergency call volume spikes 4–7× normal and when most shops are critically understaffed.
The snapshot has a “Surge” toggle that, when flipped:
- Activates a second AI receptionist instance trained on triage-only mode (it doesn’t book — it captures and assigns a callback priority).
- Texts every customer who filed an emergency in the last 90 days a proactive “we’re staffed up tonight if you need us” message.
- Pages all four levels of the on-call rotation simultaneously instead of sequentially.
- Switches the booking calendar to 60-minute slots instead of 90 (so dispatch can squeeze more in).
A Phoenix shop running this through last summer’s heatwave booked 184 emergency calls over three days. They normally would have booked 40.
7. Win-back for unbooked emergencies
This is the one most shops never build because it’s emotionally hard. When a homeowner calls at 11 PM, you can’t get a truck out, they call someone else, and they book — that customer is not dead. They are 90 days from another problem.
The win-back sequence:
- Day 3: “Hey, we couldn’t get to you the other night — hope you got squared away. If anything else comes up, here’s our direct line.”
- Day 30: A 10% off coupon for a non-emergency drain inspection or water-heater flush.
- Day 90: A maintenance plan offer.
Recovery rate (one of those three messages converts to a booking): 18% across the shops running it. Which means roughly 1 in 5 of the homeowners who booked your competitor on a Sunday night come back to you within 90 days — for a job worth 3–7× what the emergency would have been.
The math, again, plainly
Snapshot: $997 one-time.
Shop median for the seven automations above, measured at day 14:
- 8–14 recovered emergency jobs.
- Average ticket: $480–$650.
- Recovered revenue: $4,000–$8,000.
Day 14 is when the snapshot pays for itself. Day 90 is when it’s compounded into the maintenance-plan flywheel and the review pipeline. Day 365 is when the operator realizes they shouldn’t have run the shop without it.
What we don’t promise
We don’t promise a specific number of recovered jobs. Your service area, your pricing, your truck count, and your team’s ability to actually show up when the AI books the call all matter more than the workflow. What we promise is: if your shop is missing more than 15% of after-hours calls today, these seven automations will get you under 5% inside 30 days. Past that, it’s on the truck.
Schedule a demo and we’ll walk through your last 30 days of call logs to show you exactly where the leaks are. No charge, no pitch — just numbers.