A job duration estimate is the expected on-site time for a job, from the technician starting work to finishing it. It is the length of the block a job takes on the schedule, excluding travel.
How to set a job duration estimate
Base it on your own history, by job type:
Duration estimate = Median actual on-site time for that job type
Then decide how much to plan for the slow cases:
Planned duration = Median + Buffer (for example, the gap between the median and the 80th-percentile time)
Say your last 50 furnace tune-ups took a median of 55 minutes, and 80% finished within 70 minutes. Booking 70 minutes covers most visits without padding every one to the worst case. Booking 120 "to be safe" wastes nearly an hour on most visits.
Use the median, not the average. A few very long jobs pull the average up and make the typical job look longer than it is.
Why it matters for scheduling
Every job's start time depends on the previous job's end time. If estimates are short, the day runs late and each later job inherits the delay. If estimates are padded, trucks finish early and sit idle, and the board looks full while capacity goes unused.
Duration uncertainty is also the first reason arrival windows run long. The more spread in how long a job takes, the wider the window has to be for the job after it. That mechanism is laid out in Why arrival windows are so long.
A common mistake
Putting travel inside the duration. If a 60-minute job is booked as 90 minutes "to cover the drive," the estimate is wrong for every job except one at exactly that distance. A job next door gets the same padding as one across town. Keep duration as on-site time only and put drive time next to it. CrewLink's technician scheduling works this way: the drive sits next to the job instead of hidden inside it, so the board shows what the day actually contains. Why this distinction matters for utilization is in 100% booked is not 100% utilized.
Related terms
- Drive time belongs next to the duration, not inside it.
- Arrival window width follows from duration uncertainty.
- Idle time grows when durations are padded.
- Wrench time is the measured version of what this estimate predicts.
Common questions
How often should job duration estimates be updated?
Review them every quarter, or after any change in equipment, process or crew experience. Durations drift.
Should different technicians get different estimates?
If the difference is large and consistent, yes. An apprentice and a senior technician often take different times on the same job.
What if I have no job history?
Start with your technicians' best estimate per job type, then record actual start and finish times for a few weeks and replace the guesses with medians.