July was the biggest month the company has ever had. Everyone worked six days. The deposits were enormous.
The margin was the worst of the year. This is not unusual, and it is not a mystery once you look at what a peak month is actually made of.
What a busy month does to cost
Overtime replaces regular hours, at a rate that does not appear in any estimate. Parts get bought at counter prices because there was no time to order properly. Callbacks rise because work gets done fast. Warranty claims rise with equipment volume, and the paperwork behind them is exactly what gets deferred when the schedule is full.
Every one of those costs lands on jobs that were priced from a pricebook built on a normal week.
Why the books do not show it
Because most of it arrives late. The overtime hits payroll immediately, but the callback comes in August, the supplier statement reconciles in September, and the unfiled warranty claim never arrives at all. By the time the real cost of July is visible, everyone is looking at October.
Revenue is recognised on the day. Cost is discovered over the following quarter. That gap is the whole illusion.
Measuring it without new software
Take your busiest month last year and your quietest. For each, work out revenue per completed job, average parts cost per job, and callbacks as a share of jobs completed.
If the busy month is worse on two of those three, the problem is not your pricing. It is what happens to a business at capacity, and it is fixable in the back office rather than in the field.