The two numbers every PSAP reports to someone are call answer time and call processing time. How fast did you pick up the phone, and how fast did you get a unit dispatched. NENA has standards for both. Accrediting bodies measure them. Directors put them in presentations.
Neither number tells you whether the call was handled well.
Speed is not quality
A dispatcher who answers in two seconds and dispatches the wrong unit type in 45 seconds has excellent metrics. A dispatcher who takes an extra 15 seconds to ask the follow-up question that changes the call from a single engine to a full first alarm has worse metrics and a better outcome.
The call answer time standard exists for a good reason - people calling 911 should not wait. The processing time standard exists for a good reason - delayed dispatch costs response time. But these metrics measure the system's speed, not its accuracy. They tell you how fast the machine ran. They don't tell you whether it ran in the right direction.
What would a meaningful metric measure?
The metric that actually predicts outcome quality is something like: "Did the initial dispatch match the actual incident?" In other words, when the first unit arrived, did they find what the dispatcher told them they'd find? Were the right number of units dispatched? Was the call type accurate? Was the information broadcast on the radio sufficient for the responding units to make good approach decisions?
That metric is harder to track. It requires matching the dispatch record against the incident record after the fact. It requires someone to review whether the initial resource assignment was appropriate given what the dispatcher knew at the time. It requires judgment, not just a timer.
No accrediting body measures it. No national standard defines it. Most centers don't track it. And yet it's the single number that tells you whether your dispatch operation is actually working.
The perverse incentive
When you measure speed and don't measure accuracy, you get fast inaccuracy. Dispatchers who are evaluated on processing time will optimize for processing time. They'll dispatch faster, which sometimes means dispatching with less information, which sometimes means dispatching the wrong thing.
The dispatcher who holds the line for an extra 20 seconds to get a better description, a more accurate location, or a clearer picture of the threat level is producing a better outcome and a worse metric. In a system that rewards the metric, that dispatcher gets coached to be faster. The incentive structure is backwards.
Where to start
You probably can't change your accreditation metrics or your director's reporting requirements. But you can add an internal quality metric to your QA process: did the initial dispatch match the incident? Track it for six months. See where the mismatches cluster - by call type, by time of day, by dispatcher, by position. Those clusters are your actual performance story, and they'll tell you more about your operation than every call answer time report combined.