Skip to content
AvidraAvidra
  • Home
  • Pricing
  • How it works
  • Blog
  • About
  • Contact
Sign inStart free
AVIDRA

Every call answered.

AI-assisted missed-call recovery for local service businesses.

Product

  • Pricing
  • How it works
  • Features
  • Compare
  • Blog
  • System status

Industries

  • Dental
  • Electrical
  • Garage door
  • HVAC
  • Locksmith
  • Med spa
  • Plumbing
  • Roofing

Company

  • About
  • Contact
  • FAQ
  • Book a demo
  • Sign in

Legal

  • Privacy
  • Terms
  • Security
  • Refunds
  • Cookies
  • SMS policy
  • DPA
  • Acceptable use
  • Accessibility
  • Do not sell my info

© 2026 Avidra™. Avidra™ and Hypnics™ are trademarks of Hypnics. All rights reserved.

Avidra is a product of Hypnics. Pickering, ON, L1X 0G2, Canada.

UPTIME 99.9% · 4.2s avg · WORKS WITH YOUR NUMBER

Industry

Plumbing emergencies after midnight: how to capture without burning out

2026-05-08 · 8 min read · By Asad Mohammad

12:47 in the morning. The phone on the nightstand rings. The on-call tech (who, this week, is you) picks it up. A homeowner in Oakville says there's water coming through her kitchen ceiling and she can hear it dripping. Her husband shut off the main but she can already see the drywall sagging. Tomorrow is her daughter's birthday party. Can someone come now.

This call is why plumbing is plumbing. It's also why most plumbing shop owners are tired all the time. The math on after-hours emergencies is real money. The math on your own sleep is also real. The trick is building a system that captures the money without burning out the team that has to answer the phone at 1am.

What an after-hours plumbing emergency actually is

Three call shapes show up after midnight in residential. The first is the one above: real emergency, water somewhere it shouldn't be, panicked homeowner, has to be addressed now. This is the most valuable call shape because the alternative for the homeowner is real damage. Average ticket on these is high. Conversion is high.

The second is the "I think I have an emergency" call. Maybe it's a slow drip, maybe a hot water tank that's hissing weirdly in a way the homeowner has never heard before. The homeowner doesn't know if it can wait. Half the time it can. Half the time it can't. Either way they want to talk to someone right now to find out.

The third is the surprise. A landlord calling because his tenant just texted at 1am about a sewer backup. Or a real estate agent calling because a showing tomorrow needs the basement to not smell like a swamp.

Each shape has a different cost structure and a different right answer. Treating them the same is the mistake most after-hours systems make.

The triage problem

The hard part about after-hours plumbing is that most after-hours calls are legitimate but only some are true emergencies. If your on-call tech rolls every call, you burn him out fast and most of the rolls are routine work that could've waited until 7am.

The cheap version of triage is voicemail with a "press 1 for emergency" greeting. It works some of the time. Often the homeowner panic-mashes the keypad or just hangs up because they think nobody's home. Voicemail can't pick the urgent calls out of the routine ones because it can't ask follow-up questions.

The next version is a 24/7 answering service. Real human picks up, asks the script you gave them, decides whether to dispatch the tech. This works better but costs $400-$1,200/month for the kind of trades-shop volume we're talking about. And the script the human follows is exactly the script you wrote. If you wrote it wrong, the dispatch decisions go wrong too.

The version we built into Avidra is different. AI picks up, asks the homeowner via text or voice what's going on, and routes based on the answers. If the homeowner says "water is coming through my ceiling," the AI flags as emergency. If the homeowner says "my toilet runs sometimes but it's been like that for a week," the AI captures the lead but doesn't wake the tech up at 1am. Both calls get captured. The on-call tech only gets paged for the first one.

Related reading
  • What a missed service call actually costs your shop
  • HVAC busy season is two months. Don't lose calls in them.
  • After-hours call answering: in-house, service, or AI?
The triage tax

Every minute your on-call tech spends listening to a non-emergency is a minute he's not going to be sharp tomorrow at 9am. After-hours triage isn't free even if you're "just answering the phone." Sleep debt costs your shop in slower job completion and in mistakes on the install the next day.

A playbook that works

The version that holds up against both the lost-lead math and the sleep-cost math has four parts.

Set a real urgency threshold

Write down what counts as an emergency. Not in your head, on a doc that the answering layer can use. Something like:

  • Active water leak, more than a trickle, going somewhere it shouldn't.
  • Sewer backup into the house.
  • No water at all in a house with people in it.
  • No hot water in winter, with people in the house, and an existing customer.
  • Gas smell (which you then transfer to the gas company, not your truck).

Everything else gets logged as a non-emergency callback for first thing in the morning. The homeowner gets a confirmation by text that you'll be calling at 7am, plus the option to text back if it escalates.

This threshold should be your call. Not the answering service's, not the AI's. The human running the shop knows what's a real emergency for their customer base and what isn't.

Capture every call, but only page on emergencies

The trap is treating "missed it" and "didn't page" as the same thing. They're not. Every after-hours call should be captured: name, address, what's going on, urgency level. That's the lead. Whether you wake up your on-call tech is a separate question. Capture everything. Page on the threshold above.

If you're using voicemail as your capture system, you're going to lose half of these. Most homeowners hang up. SMS-based capture has a much higher response rate because the homeowner can answer in their own time, with their own thumb.

Rotate the on-call tech, with real backup

One person every night is a recipe for burnout, even with good triage. The rotation that tends to hold up is two-week shifts with a clear primary and a clear backup. The primary takes pages first. The backup takes calls only if the primary is on another job. Same threshold for both. Write the after-hours pay structure into your operating agreement before the rotation starts, not after the first dispute.

Measure the false-positive rate

Once a month, sit with the on-call tech and look at the pages from the past four weeks. Mark each one: was this actually an emergency, or did the triage layer get it wrong? If your false-positive rate is over 25%, your urgency rules need updating. If your false-negative rate (real emergencies that didn't get paged) is over 5%, your rules are too tight and you're losing customers to the plumber who DID send someone.

What this looks like on a Tuesday morning

The morning after a good after-hours system is running, here's what the owner walks into. The dispatch dashboard shows 4 captured calls from overnight. One was paged: the kitchen-ceiling leak in Oakville. The other three weren't urgent. A slow toilet drip plus a hot-water-no-heat call from someone who specified she could wait until morning, with a stray question about opening hours mixed in.

The on-call tech got paged once, drove out at 1:15am, was back home by 3:30am with a $700 emergency call closed out. He slept until 9. The other three calls go into the day's schedule. Two of them get booked for that afternoon. One was a tire-kicker who never replied to the morning follow-up.

That morning is the system working. Most shops without a triage layer have either of two alternatives. The on-call tech rolls four times that night and is wrecked Tuesday. Or three of the four calls go to voicemail and never come back. Both versions cost money. The first costs you in tech wellness, the second in lost jobs.

The honest part

This kind of system isn't trivial to set up. Writing real triage rules takes time. Training whatever answering layer you're using (human or AI) on your specific operation takes a couple of iterations. Adjusting the false-positive rate after the first month is real work.

What's not trivial is the alternative. Either you keep losing after-hours calls, or your on-call tech burns out from getting paged on every voicemail. Spending $500-$1,000/month on a generic answering service that doesn't actually know your business is the third option, and it's the most expensive of the three.

The right answer probably involves AI doing the triage and capture against a clear threshold you wrote, with a sane rotation underneath. If you want to see how Avidra does the triage part, pricing is on the site. The hard part you'll still own: writing down what an emergency is for your shop. Nobody else can do that for you.