Which dispatch KPIs matter most?
The ones that follow a job from request to drop-off, worked out the same way every week.
| Measure | How to calculate it |
|---|---|
| Call answer time | Time from the phone ringing to a person answering, averaged per hour |
| Time to assign | Time from a booking being accepted to a driver accepting the job |
| Pickup lateness | Arrival time minus promised pickup time, for jobs that were late |
| Unserved bookings | Accepted bookings no car reached, as a share of all bookings |
| Cancellations | Jobs cancelled by passenger or by driver, counted separately |
| No-shows | Jobs where the car arrived and the passenger did not travel |
| Empty miles | Miles driven to pickups as a share of all miles driven |
Definitions are ours; no regulator publishes dispatch targets.
Where does the data for these measures come from?
From each trip's record. New York City bases already send the TLC, every month, the date, time and place of each pickup and drop-off and whether passenger or driver cancelled.[1]
Do NEMT providers need extra measures?
Usually no-shows by rider and by provider, because programmes track them. North Carolina's policy asks for both kinds of no-show in every trip record.[2]
How often should dispatch KPIs be reviewed?
Weekly by hour of day, so you can see which shifts need more drivers or dispatchers, and monthly for trends.
Frequently asked questions
Is there a standard pickup time a taxi firm should hit?
No regulator sets one for ordinary bookings; contracts such as NEMT or school work may set their own, so measure against what each customer was promised.
Should driver acceptance be a KPI?
Yes, as a diagnostic: a falling acceptance rate in one zone usually points to a zone, fare or timing problem.
Sources
Fees and rules change. Each source is the official page we read, with the date we read it; check it before you apply or pay.
