24/7 Emergency Calls: Triage, Thresholds, Wake-Ups

There are two kinds of "24/7," and a company's health depends on knowing which one it sells. The first is a marketing position: the company promises an answer at any hour, and keeps it with a system — an automated first tier answering everything, and thresholds deciding what wakes a human. The second is a human load: someone answers in person at any hour, whatever the call. The first holds for years; the second collapses within months, unless properly staffed and paid for — which is rare.

Some trades require genuine night capacity — the locksmith, the emergency repair service, the security company — and those trades have built it in: night surcharges, rotation, thresholds. For the rest, the honest question is: what real need does the night serve? The answer lives in the call log: a few overnight calls a week, a majority of which aren't urgent in the strict sense. A well-designed system absorbs that reality — answer everything, wake rarely — rather than faking it with a human picking up at 3 AM to read out the hours.

A night triage system holds in three colors, defined per trade and written in black and white. Red: act now — people's safety at stake, active damage growing, contractual commitment to immediate response. Orange: call back within the hour — serious but stable problem, customer in measurable difficulty, contained damage risk. Green: handle in the morning — everything else, including what sounds urgent to the caller but doesn't degrade before 8 AM. Each trade fills its three columns with its real situations; a plumber and a lawyer wouldn't write the same lines.

The taxonomy gets validated by experience, not in a room. The first month, every classified night call gets reviewed in the morning: was that 2 AM wake-up truly red? Was that midnight message really green? Reclassifications refine the list, and after a quarter, the taxonomy reflects what the actual night contains — not what was imagined. A taxonomy never reviewed is a taxonomy never believed, and the on-call person pays the price.

The wake-up threshold document is of total apparent boredom and extreme silent profitability. It lists the situations — typically three to five — justifying disturbing a human at night, each with its reason: safety, cost of delay, commitment. Everything else, whatever the caller's tone or insistence, gets handled green or orange. The document's discipline spares a thousand nocturnal micro-decisions, and every avoided micro-decision is preserved sleep — literally workplace safety for a trade handling tools the next morning.

Writing the document is a one-hour exercise with a method. Review the last six months of after-hours calls, list everything that disturbed someone, and ask for each case: "what would have happened if we'd waited until 8 AM?" The cases where the answer costs dearly rise to the wake-up list; the rest sink into green. That yields a list grounded in the trade's facts, not in phone anxiety — and the night on-call inherits a rule more reliable than their own 3 AM judgment.

The script that answers at night follows an unchanging order: safety first, information second, frame last. Safety: any danger signal — medical, fire, gas, threat — redirects to emergency services before any qualification, without exception or question hierarchy. Information: name, address or location, callback number, nature of the problem, occupants or stakes involved — the collection the receptionist executes in two minutes. Frame: what the customer should expect — "the on-call will call you back before [time]", never a vague promise — and confirmation that the message goes out immediately.

Two details separate a good night script from a mediocre one. First: announce the callback delay before hanging up — the customer who knows they'll be called at 7:30 AM waits calmly; the one who knows nothing dials three other companies. Second: transmit everything to the on-call in writing — a structured text summary reads in twenty seconds on waking, when a voicemail replays poorly at 3 AM. At night, writing beats voice everywhere voice isn't indispensable.

The real test of a night service is what happens at 8 AM. Without a structured handover, overnight messages vanish between coffee and the first job; with one, the day starts with ten minutes that sort everything. The morning report: reds handled overnight (and their status), oranges to call back within the hour, greens to handle during the day, and the special cases — the furious customer, the ambiguous file, the wake-up that shouldn't have been. One page, read standing, that sets the battle order.

The handover also closes the quality loop: every night call is data about the system. Were callback promises kept? Did the thresholds classify correctly? Did the recovery text open the conversations it should have? Ten minutes of reading every morning, a monthly one-hour review, and the night service improves like everything else in the company — by iteration, not miracle.

The fatal risk of 24/7 isn't technical, it's human: fatigue. A tradesperson who sleeps badly drives badly, welds badly, judges badly — and ends up paying in safety what they thought they were gaining in availability. The warning signs are known: multiple wake-ups a week, early-morning irritability, silly errors at the start of the day, a phone dreaded in the evening. Spotting them early and acting — tightening thresholds, recruiting a second on-call, outsourcing the night — is a management decision, not an admission of weakness.

The sustainability rule is countable: beyond a few wake-ups a week on average, a lone human doesn't hold — the system must evolve. Evolving means one of three things: tighten the red definition, pay a second on-call to alternate, or hand the overnight first tier to an external service and keep only the true reds. Companies that sustain 24/7 over the long haul have all made one of these three choices; the ones that shut their night service made none.