Scheduling That Matches Capacity to Demand

Most workshops book by appointment slots and hope. The result is a familiar pattern: overbooked Monday, a queue at eight in the morning, and a Thursday with idle bays.

Both ends cost the same amount, and the schedule is where it is decided.

What a schedule has to account for

Clock hours, not book hours. A day booked with eight book hours per technician is overbooked by whatever their efficiency exceeds 100%, minus whatever productivity falls short — and those do not cancel. See flat rate and clock time.

Skill, not just hours. A booked hour that only one technician can perform is not capacity if that technician is committed.

Parts lead time. A job booked before the part arrives occupies a bay and produces nothing. See parts availability.

Work you know is coming. Comebacks, warranty, internal preparation, emergencies. A schedule at 100% with no allowance will run late by the second day.

Diagnostic uncertainty. A diagnostic job has a wide distribution of durations, and scheduling it as though it were an oil change guarantees drift.

And the write-up and delivery time, at both ends, which is not in the book time.

The arrival problem

The single largest scheduling failure in most workshops, and it is a customer-behaviour problem rather than a capacity one.

Everyone arrives at eight. The workshop cannot process them, the advisors are overwhelmed, technicians wait for dispatch, and the day starts an hour behind.

Then nothing arrives after ten, so afternoon capacity is committed to work booked in the morning rush.

What helps:

Staggered appointment times, offered as genuine choices rather than as a default of "first thing".

Incentivise off-peak. A discount, a courtesy vehicle guaranteed, a faster turnaround — something that makes eleven o'clock attractive.

Early drop-off and key drops for customers who want the vehicle in early but do not need to speak to anyone.

Separate the wait-and-see jobs from the drop-offs. A customer waiting is a scheduling constraint of its own.

And level the week. Monday and the day after a public holiday are structurally busy. Moving discretionary work — scheduled maintenance, recalls, non-urgent repairs — is cheaper than adding capacity for a peak that exists twice a week.

Dispatch, which is separate from scheduling

Scheduling decides what comes in. Dispatch decides who does it, when.

A workshop can be perfectly scheduled and lose hours to dispatch — the job is ready, a technician is free, and nobody connected them. This appears in the data as short repeated gaps rather than one long one, which is why it is missed. See the three ratios.

Someone owns dispatch, actively, in real time. Where nobody does, technicians dispatch themselves, and they will reasonably pick the jobs that pay best.

Match skill deliberately. Routine work performed by the diagnostic technician is capacity spent at the wrong rate.

And keep a next job ready for each technician, so the transition is seconds rather than a walk to the counter.

Building a schedule that holds

Establish actual capacity first, from your own data rather than from headcount. See what time data tells you about capacity.

Book to a percentage of it, with the remainder held for the work you know arrives.

Categorise by duration and certainty. Short and predictable, long and predictable, uncertain. Schedule the uncertain ones where the drift can be absorbed rather than where it cascades.

Confirm the day before. No-shows and late arrivals are a large loss and a reminder recovers most of them.

And review the schedule against what happened, weekly. The gap between booked and actual is the information — if every Monday overruns by two hours, the Monday booking level is wrong and no amount of effort on the day fixes it.

Measures worth watching

Booked hours against available clock hours, by day.

Load distribution across the week.

Carry-over — vehicles not completed on the promised day.

Idle time by cause, which distinguishes a scheduling problem from a dispatch one.

No-show and late arrival rate.

And promise accuracy — how often the vehicle was ready when the customer was told. This is the measure customers actually experience and it is rarely tracked.

The trade nobody states

A workshop scheduled to 100% of capacity has no resilience, and one scheduled to 80% has idle time.

The right level is a decision, and it depends on how much unplanned work arrives and how costly a late delivery is to the business.

Making it explicit is the point. Most workshops are at whatever level emerged, and nobody chose it.

The short version

Schedule in clock hours, hold capacity for the work you know is coming, and level the week.

The arrival peak is a customer-behaviour problem — stagger appointments and make off-peak attractive.

Dispatch is separate from scheduling and loses hours in short gaps that nobody sees.

Review booked against actual weekly. A Monday that overruns every week is a booking-level problem, not an effort problem.

And choose the utilisation level deliberately, because at present it is whatever emerged.

For official guidance on organising work through a safety-management system, see OSHA safety-management guidance.