Trade-Aware Call Triage Explained

How asking the right technical question turns “is this an emergency?” from a guess into a decision.

The hardest job in after-hours dispatch is not answering the phone. It is deciding what to do with the call once it has been answered. Get it right and you roll trucks on real emergencies and book routine work for the morning. Get it wrong and you waste money in both directions — rolling on calls that could have waited and missing calls that could not.

The generic-script problem

A generic answering service’s intake script asks the customer “is this an emergency?” The problem with that question is that the customer is not a tradesperson. A homeowner whose breaker keeps tripping might say “no, it’s just being annoying” when in fact they have a serious wiring fault that could start a fire. The same homeowner might say “yes, emergency!” about a flickering light that is actually a loose bulb. The customer’s self-assessment is unreliable.

Trade-aware triage replaces “is this an emergency?” with “what would a tradesperson need to know to make this call?” The questions are concrete: is the breaker tripping or is the panel itself making noise? Is water flowing or is it just stained? Is the walk-in freezer the issue or just the residential fridge? The answers map to urgency reliably regardless of what the customer thinks.

The questions, by trade

The triage questions that actually separate emergency from routine, by trade:

TradeThe question that separates emergency from routine
HVAC residentialIndoor temperature now, occupant vulnerability (infants, elderly, medical), and current outdoor weather.
HVAC commercial / refrigerationEquipment type, contents at risk, current temperature, time since failure.
PlumbingActive water flow, sewage involvement, number of fixtures affected, shutoff access.
ElectricalPanel sparking or smoke vs. tripping; affected circuit count; smell of burning insulation.
RoofingActive leak in living space vs. attic only; current weather; structural damage indicators.

These are the technical questions a competent dispatcher would ask. The customer is not expected to know they are the right questions; they just need to answer them honestly. The triage logic does the classification.

Three traps in DIY triage

Most shops have an informal triage protocol they invented over time. The common traps:

  1. Asking the same questions in every season. A “no AC” call in July is a different call than in October. Weather context should adjust the urgency threshold automatically.
  2. Treating residential and commercial the same. A walk-in freezer at 41 degrees is a different problem than a warm residential fridge. The intake script should branch.
  3. Letting the customer’s tone drive the classification. Some customers under-report severity, some over-report. The questions, not the tone, should classify the call.

Why this matters more after hours

After-hours triage failures are more expensive than business-hours ones. During the day, the wrong dispatch decision costs a couple of hours of tech time. After hours, the wrong decision costs the differential between standard and premium rates, plus the opportunity cost of the next call that comes in while the tech is on a job that did not need rolling. Combined with the ~65% voicemail abandonment rate for unanswered calls, the wrong triage decisions stack with the wrong answering decisions.

How AI triage compares to human triage

A well-trained human dispatcher with years of trade experience does this well. The challenges are: it is hard to find people with that depth of trade knowledge, harder to keep them working night shifts, and impossible to scale a single person to 24/7 coverage. AI triage shines on the consistency axis — the same questions get asked the same way every call, every shift, every night. It does not match a senior human on edge-case judgment, but it beats most night-shift junior agents on consistency by a wide margin.

How Night Watch implements it

Night Watch ships triage rules for HVAC and refrigeration, plumbing, electrical, and roofing out of the box. Each trade has commercial-vs-residential branching, equipment-aware questions, and a small set of clarifiers when the initial description is ambiguous. The rules feed directly into the 3x3 dispatch loop: emergency triggers Bulldog; routine books a slot via live calendar availability. Weather data is injected at call start via the weather-aware triage layer.

The takeaway

Triage is the highest-leverage decision in after-hours dispatch and the place most shops lose money. Replace “is this an emergency?” with the small list of trade-specific questions that actually classify the call, and the answer becomes consistent regardless of who is on the phone.