How schedules get overbooked
- Two people booking at once from a phone call and an online form.
- Emergency work added verbally and never written to the calendar.
- Job durations set shorter than the work actually takes.
- Travel between stops ignored, so the day looks feasible on paper.
- A technician marked available who is on leave, training or already committed elsewhere.
The five controls that actually work
- Single write path: every surface requests a slot; one validated engine writes it.
- Capacity check at write time on crew, skill and territory — not a nightly report.
- Travel validation between adjacent stops before the slot is confirmed.
- Hard operating-hour and time-off boundaries that a booking cannot cross silently.
- Duration derived from job type, so a 20-minute slot cannot hold a four-hour job.
When an override is legitimate
There are real cases: a high-value customer emergency, a same-day insurance inspection, a crew that has agreed to stay late. The system should allow it — and refuse to let it happen invisibly.
A safe override records who approved it, why, and against which capacity rule, and it notifies the assigned crew. An override that looks identical to a normal booking is indistinguishable from a mistake the next morning.
Should you overbook to cover cancellations?
Deliberate overbooking imports the airline model into a business where the cost of a broken appointment is a review, not a voucher. It trades a reliable failure for an unpredictable one.
Reducing cancellations is the durable answer: confirmations, day-before reminders, on-the-way notifications and clear scope expectations. Hold a small same-day reserve for the gaps that do open.