Buffer time is time deliberately left unbooked in a technician's day, before, between or after jobs, so that a job running long or a drive taking longer than planned doesn't make the next customer late.
Three kinds of buffer
- Duration buffer. Extra minutes added inside a job block. A job expected to take 60 minutes is booked as 90.
- Travel buffer. Extra minutes added to the drive between jobs, often because the drive is estimated from the shop or from a map with no time of day.
- End-of-day buffer. An empty slot left at the end of the day to catch overruns before they become overtime.
Duration and travel buffers are invisible on most boards. They sit inside blocks, so the schedule looks full while the tech has slack. That is why padded days show high booked utilization and still finish early.
How to size a buffer
A buffer should cover the spread of what it protects, not a round number. If your job durations for a job type have a standard deviation of s minutes, a buffer of about 1.3 × s after that job covers roughly 90% of overruns, assuming overruns are roughly normal. Real jobs run long more often than short, so check the result against what actually happens.
Better job duration estimates by job type shrink s, and the buffer shrinks with it. Drive estimates computed from the previous job at that hour shrink the travel buffer. The arrival window math guide shows how these spreads add up across a day.
What buffer costs
Use letters: buffer minutes per stop (B), stops per truck per day (n), trucks (T). Daily buffer across the fleet is B × n × T minutes.
Say, as an illustration, you add 20 minutes of buffer to each of 4 stops on 10 trucks. That is 800 minutes a day, or about 13 truck-hours. Some of it gets used by overruns. The rest becomes idle time: paid, unbooked and unproductive.
A common mistake
Adding the same buffer to every job. A 30-minute filter change and a 3-hour diagnosis don't have the same spread. A flat buffer is too much for one and too little for the other. Buffer by job type, and put it where the uncertainty is: after long or unpredictable jobs, not after short, predictable ones.
In CrewLink
CrewLink, in early access, computes the drive at that hour from the previous job when it offers a time, so the travel buffer doesn't have to cover a guessed drive. Time on site is recorded when each job is marked done, and future bookings use it. When a job runs long, the tech's mid-job check-in can add 15, 30 or 60 minutes and re-flow the rest of the day.
What to take away
- Buffer time is planned slack: duration, travel and end-of-day buffers.
- Size it from the spread of what it protects, about 1.3 × s for 90% cover.
- Fleet cost is B × n × T minutes a day; unused buffer becomes idle time.
- Buffer by job type, where the uncertainty is.
Related terms
- Idle time is where unused buffer ends up.
- Drive time is what a travel buffer is padding.
- Schedule density drops as buffer grows.
Common questions
How much buffer time should I leave between service calls?
Enough to cover the spread of the job before it and the drive after it, not a fixed number. Measure how much your job types usually overrun and size the buffer to that. Short, predictable jobs need little or none.
Is buffer time the same as idle time?
No. Buffer is planned slack. It becomes idle time when nothing uses it. A good buffer gets used often; a buffer that is almost never used is too big.
Should I add buffer to the drive or the job?
To whichever is uncertain. If drives are estimated from the shop or without time of day, the drive is the problem. If drive estimates are good and jobs vary, buffer after the job.