How do I identify callbacks and rework in ServiceTitan data?
Look for a second job at the same location, on the same equipment or job category, within a defined window after a completed job. Do not rely on a callback flag, because flagging is voluntary and the people who create callbacks are the least likely to mark them. A derived rule applied consistently will find substantially more rework than the flag does.
Why the flag undercounts
Callback flags depend on someone deciding, at booking time, that this visit is a consequence of a previous one. Office staff frequently do not know. Technicians who caused the issue have no incentive to label it. And genuinely ambiguous cases, where a second unrelated failure occurs soon after a repair, get resolved in the direction that avoids conflict.
The flag is still worth keeping. It just should not be your measurement.
A derived rule that works
Define a callback as a job at the same location, within the same trade or equipment category, created within a window after a completed job, excluding scheduled maintenance and excluding jobs on different equipment. The window should reflect your warranty terms, and it should differ between repair work and installation work.
Run the rule and compare its output to the flag. The gap is informative on its own: a large gap means your quality data has been fiction, and a small gap means your flagging culture is unusually good.
Separate the four things people call a callback
- True rework. The original repair did not hold. This is the number that should drive training.
- Incomplete first visit. Parts were not on the truck, so a second trip was always required. This is a stocking and dispatch problem, not a skill problem.
- Adjacent failure. A different component failed soon after. Sometimes genuinely unrelated, sometimes a missed inspection.
- Customer expectation gap. The system works as designed and the customer expected something else. This is a communication issue, and it is the one most visible in call recordings.
Cost the rework, then decide
Rework is expensive in ways that never appear on an invoice: the truck roll, the labor, the capacity that could have gone to revenue work, and the customer relationship. Attaching a labor and drive cost estimate to derived callbacks gives you a number that competes for attention against the revenue projects, which is the only way it gets addressed.
The follow-up that turns the number into action is reviewing what happened on the original visit and the complaint call. Pairing job records with call analysis and structured evaluation against your own standards shows a manager the specific behavior to coach, rather than handing them a rate. The evaluation supports the coaching decision; the manager makes it.
Topics: callbacks · warranty · quality · data quality
Have a version of this question about your own business?
The useful answer usually depends on which systems you run and how they're connected. That's a conversation, not a blog post.