At 6:40 on a Tuesday evening, someone is standing in their garage watching water spread from the base of a water heater. They search, find your number, and call it. Your last tech left at 5:30, and the office line rolls to voicemail.
The next thirty-five minutes, by the clock
6:40 — the call rings four times, then the after-hours greeting starts.
6:41 — they hang up partway through it. Some callers leave a voicemail; in a 2025 CallRail survey of a thousand US consumers, 42 percent said they do. The rest leave nothing, which means the first you will hear of this call is an unfamiliar number in tomorrow's log.
6:44 — they are back in the search results. The next plumber's line rings out too. Third name down the list.
6:51 — the third company answers, or doesn't answer but texts back inside a minute. Either way, a conversation has started, and it is not with you.
7:15 — the appointment is booked for 8am tomorrow.
Nothing about this caller was disloyal. They had water on the concrete and a phone in their hand. In that same CallRail survey, 82 percent of consumers said an unanswered call sends them to a competitor. The window in which the job was yours lasted about as long as it takes to refill a coffee.
After-hours demand is the good kind of demand
The calls that arrive between close and open are disproportionately the valuable ones. Nobody phones about a water heater at 6:40pm to comparison shop; they phone because it failed at 6:20. An August evening in Lubbock produces air-conditioning failures at exactly the hour the dispatcher goes home, and a February norther does the same thing for heat. Garage doors jam at 9pm with the car inside. A buyer drives past a listing at dusk and wants to see it tomorrow before someone else does — a call the listing office also has nobody sitting by the phone for.
This is not only an after-hours problem, either. Invoca, a call-analytics company, analyzed more than sixty million phone calls in 2025 and found about 27 percent of calls to home-services businesses go unanswered — and that figure includes business hours, when someone was theoretically at the desk. Outside staffed hours, the answer rate is effectively zero unless something is built to catch the call.
The mechanism, plainly
The fix is not an answering service, and it is not asking anyone to work later. A programmable phone number sits in front of your published line and forwards every call exactly the way the line always worked. When a call ends unanswered, that event fires a trigger. Before anything goes out, the system checks a few things: whether the caller's number is a mobile that can actually receive a text (a landline swallows one silently), whether this number already got a message today, whether it looks like a spam dialer. Then, inside the first minute, the caller gets something like:
This is Smith Plumbing — sorry we missed you. What's going on at your place? If this is an emergency, reply CALL and this number will ring straight through to the on-call phone.
The reply lands on the phone the owner already carries, in the same inbox as the rest of the leads — not in a dashboard nobody opens.
Why an open loop beats a dead end
An unanswered call is a terminal event. It cannot be picked back up, and by 8:30 the caller has shut the water off, put towels down, and is on the couch deciding who comes in the morning. A two-hour-old missed call takes no part in that decision. A text thread does. It is sitting on their lock screen, one thumb-tap from an answer, and answering it costs them nothing at 8:40pm in a way that returning a call never quite does.
The company that texted is also, quietly, the company that responded first. The oldest serious research on lead response — a Harvard Business Review study from 2011 — found that firms responding within an hour were roughly seven times more likely to qualify a lead than firms that waited longer. That study is old and it measured web leads, not plumbing calls, so treat the multiplier loosely. The direction, though, matches what every owner already knows from losing jobs this way: the first one talking usually wins.
The emergency fork
A text is the wrong instrument for water coming through a ceiling, and pretending otherwise is how these systems fail people. The first message has to carry an immediate call path: a reply keyword or a tap-to-dial number that rings a real on-call phone, right now. The text works as triage, not as a wall — routine calls settle into the thread and book for tomorrow, while genuine emergencies punch straight through to a human. Skip this fork and the system loses exactly the highest-value calls it was built to catch.
What this does not fix
- If the on-call phone also goes unanswered, the fork fails. The system relays urgency; it cannot manufacture someone to receive it.
- Callers on landlines — older customers, commercial front desks — never see the text. Those numbers still need the old habit: someone works the missed-call list first thing in the morning.
- The most common failure is not technical. The text goes out, the customer replies, and the reply sits in a portal nobody checks. Replies have to route to a phone that is already being looked at, or the whole thing quietly dies by month three.
- US carriers now require businesses to register before sending automated texts — a process involving your legal entity and sample messages — and approval takes anywhere from days to a couple of weeks. Week one of this project is paperwork, not launch, and anyone who promises a Friday sale becomes a Monday launch is planning to send messages the carriers will silently drop.
- Replying to someone who just called you is the strongest consent footing that exists in text messaging. It is not a license to send that list promotions later; that is a different legal category requiring written consent, and this system deliberately does not do it.
How you would know it is working
Run your own numbers rather than trusting anyone's case study. Each month, count four things: after-hours calls received, texts delivered, replies, and jobs booked from those replies. Multiply the booked jobs by your average ticket and set the result against what the system costs. The arithmetic takes five minutes, and it is the only proof that should matter to you at renewal.
Where to start
Pull last month's phone bill or your carrier's call log and count the calls that arrived after 5:30pm or before 7:30am. Most owners have never done this, and most are surprised by the count. That number, times your close rate, times your average ticket, is the size of this problem at your business specifically — and it is worth having in hand before you talk to us or anyone else.