Workforce and Labor
Technician Capacity Planning When Demand Is Lumpy
Lumpy demand makes the average useless. The real question is not how many technicians you need but which buffer is cheapest, and overtime wins more often than managers expect.
When service demand is lumpy, the capacity question stops being "how many people" and becomes "which buffer". You can buffer with headcount, with overtime, with subcontractors, with backlog, or with response-time commitments that admit variability. Each has a different cost and a different failure mode, and staffing to peak with permanent headcount is usually the most expensive of the five while feeling like the safest.
The overtime crossover, worked
Suppose peak demand exceeds base capacity by 300 hours a year. At the current manufacturing wage of $30.35/hour (Jul 2026, BLS) with a mid-range 35% burden, the loaded rate is about $41 and an overtime hour at time-and-a-half runs near $61. Covering the surplus with overtime costs roughly $18,438. A permanent additional technician costs about $85,223 a year fully loaded before recruitment and ramp. Overtime is the cheaper buffer by a wide margin at this volume, and it stays cheaper until the surplus approaches a substantial fraction of a full-time position.
- Overtime hour, loaded: $61
- Covering 300 surplus hours: $18,438
- One additional technician, annual: $85,223
What the arithmetic leaves out
Sustained overtime has costs the hourly comparison misses: fatigue-driven error rates, callbacks, and eventually attrition, and losing a trained technician costs the recruitment plus the ramp. That is why the honest rule is not "overtime is cheaper" but "overtime is cheaper up to a threshold of hours per person per period, beyond which it stops being a buffer and becomes understaffing". Set that threshold explicitly, monitor overtime per individual rather than in total, and the model stays defensible.
Overtime is a buffer. Permanent overtime is a staffing shortfall that has been renamed, and it bills at time and a half.
The buffers people forget
Two of these cost nothing to hold
- Backlog on deferrable work. Preventive maintenance can usually slip a week; emergencies cannot. Scheduling deferrable work as the shock absorber costs nothing and is the most underused lever available.
- Subcontract capacity, arranged in advance rather than in a crisis, which converts a fixed cost into a variable one at a premium you negotiated calmly.
- Response-time tiering, where not every ticket carries the same commitment, so peak load is smoothed by the contract rather than by the roster.
Measure the peak before you size for it
Most operations know their average demand and guess at their peak. Pull the last two years of ticket data, aggregate to the period your roster actually responds to, usually a week, and look at the distribution rather than the mean. The ratio of the busiest weeks to the median is the number that should drive the buffering decision, and it is frequently smaller than the organizational memory of the worst week suggests.
Use the labor cost calculator to price overtime against additional headcount at your own rates. Compare the options
Published 2026-08-08.