The leaks I count on the first day, and the arithmetic for each.
Every business I have walked into has leaks the owner has stopped seeing. Not fraud, not disaster. Small, daily, and compounding. The analysis counts them, because a counted leak gets fixed and a described one gets nodded at.
| Leak | How to count it |
|---|---|
| Rework and callbacks | Jobs revisited last quarter × average hours × loaded rate. Trades typically find 3 to 8 percent of revenue here. |
| Waiting | Hours crews stood down for parts, approvals or access, from the weekly log. Multiply by crew cost per hour. |
| Unbilled time | Hours logged against jobs minus hours invoiced. Services firms leak 5 to 15 percent of billable time this way. |
| Unapproved discounts | Quoted price minus invoiced price, summed. Then ask who decided each one. |
| Quote variance | Estimated hours versus actual, by job type. The one job type that always runs over is usually priced wrong, not run badly. |
| Receivables | Days sales outstanding × daily revenue × your borrowing rate. Cash you are lending to customers for free. |
| The owner's hours | Hours per week you spend on decisions someone else could make. At your effective rate, this is usually the largest line. |
The ranges above are what I have seen and what the practitioner literature reports. They are a C. Your numbers will be your numbers, and those are the ones that go in the report.
Most of these are conditions problems with a dollar sign on them. Waiting is a structure and decision-rights failure. Rework is a written-standards failure. Quote variance is an estimating standard nobody wrote. Unapproved discounts are decision rights again. That is why the analysis and the Gameplan are one piece of work: the leaks tell you which condition to fix first, and the arithmetic tells you what the fix is worth.
If this describes your business, the analysis is where I would start. Or score your own company in ten minutes.
Signed by Spencer Pallone Helm, 02 SEP 2026
