How to measure on-call response time
Measure the time from the request to the moment a person takes it, the first answer, not the time the first text went out. Give each priority its own target, track the share of requests answered inside it, and compare people, locations and times of day before you change the rotation.
Measure the first answer
Sending a text proves nothing about response. What the caller waits for is the moment someone says “I have it”. Start the clock when the request arrives and stop it when a person takes it.
Resolution is a second, separate measure: how long until the work was done. Mixing the two hides both.
Set a target by priority
One target for every request either rushes routine work or tolerates slow emergencies. Set one per priority and track the share answered inside it. DispatchMe’s reports start from these targets:
- High priority: answered within 30 minutes.
- Medium: within 1 hour.
- Low: within 2 hours.
- Goal: 90% of requests answered inside their target.
Compare before you change anything
Averages hide the slow nights. Look at:
- The median as well as the average: one slow request pulls the average up.
- Each responder’s first-answer time against the team’s.
- Each location, if some are covered by fewer people.
- Day of week and time of day, where demand and slow answers bunch up.
Read small numbers with care
A responder with five requests can look fast or slow by chance. Under about twenty requests, treat an average as a hint, not a verdict, and look at the requests themselves.
Then adjust the rotation
Slow first answers usually trace to one of three causes: too few people ranked, too long a wait before the next person, or a time block nobody really covers. Fix the cause the numbers point to, then compare the next month with this one.