Technician Utilization and What the Denominator Assumes

Why this matters

Utilization is the one operations number that gets acted on hardest and audited least. A shop reads it low and spends a quarter chasing sales it does not need, or trimming a crew that is already stretched. A shop reads it healthy and books the calendar to the edge, then pays for it the following month in premium hours and second visits. Both mistakes come from the same place: the top of the fraction is measured and the bottom of it is assumed, and almost nobody writes the assumption down.

The assumption sitting on the face of the number

The usual construction is logged hours in the window over assumed available hours, where available is a nominal full-time week multiplied by the count of active technicians multiplied by the number of weeks. The numerator is a measurement: somebody logged time against a job. The denominator is a model with three inputs, and all three are guesses that nobody revisits.

That matters because the denominator is the larger number and small percentage errors in it move the answer more than large errors in the numerator do. It also matters because headcount is a snapshot and hours are a flow, so a roster read on one day is being divided into thirteen weeks of work. That mismatch has its own article; see the reference on days sales outstanding for the general shape.

Three ordinary situations break the model, and every one of them is present in most small shops at any given time.

Break one: the part-timer counted at a full week

A technician on a nominal three-day week contributes 24 hours of capacity, not 40. Counted at a full week across a 13-week quarter, that one name adds 16 hours a week of capacity that never existed, or 208 hours over the quarter.

Break two: the mid-window joiner or leaver counted for the whole window

Somebody who left at the end of week 5 was available for 5 weeks and is being counted for 13. That is 8 weeks of phantom capacity, or 320 hours. A joiner works the same way in reverse: a technician who started at the beginning of week 10 was available for 4 weeks, and counting them for 13 adds 360 hours that nobody could have worked.

The uncomfortable version of this break is the one that runs the other way. If the roster is a snapshot at period end, the leaver is simply gone from it, so their five weeks of logged hours sit in the numerator with no matching capacity in the denominator. Pick one rule, weeks present rather than heads at a point in time, and apply it to both sides.

Break three: a name on the active list that does not turn a wrench

A service coordinator, a working owner who now mostly quotes, an apprentice on classroom weeks. Each is a real employee and none of them is field capacity. Counted as a technician for a 13-week quarter, one such name adds 520 hours of assumed capacity against which zero field hours will ever be logged.

The three breaks, worked on one small crew

A 6-name active list, a 13-week quarter, a nominal 40-hour week. Logged field hours for the quarter: 1,616.

Reported denominator: 6 x 40 x 13 = 3,120 hours. Reported utilization: 1,616 / 3,120 = 51.8 percent.

The real crew behind those six names:

True utilization: 1,616 / 2,072 = 78.0 percent.

The phantom capacity is 3,120 minus 2,072, or 1,048 hours, which is 33.6 percent of the reported denominator. A third of the capacity in the report did not exist. The reported figure and the true one are the same numerator over two different denominators, and they are 26.2 percentage points apart.

Why the three corrections do not add up

Fix each break on its own, holding the numerator at 1,616 hours:

Those individual gains sum to 20.0 points and the combined correction is worth 26.2. They are not additive because each correction shrinks the base the next one acts on. The practical consequence is that fixing one break and declaring the number clean understates how wrong the rest of it still is, and a shop that fixes the easiest break first is the one most likely to stop there.

Crew size decides how much one name is worth

The reason this bites small shops and not large ones is arithmetic, not discipline. On the 6-name list above, one full-time name is 16.7 percent of the assumed denominator. On a 30-name list the same name is 3.3 percent. The identical misclassification, one coordinator carried as a technician, moves the small shop's figure roughly five times as far.

That sets the review effort where it belongs. Under about ten field names, every single row of the capacity list has to be right, because any one of them can move the answer by more than a period's worth of genuine change. Above that, the list matters in aggregate and the numerator is usually where your next hour of checking pays.

What counts as a logged hour, and the failure underneath it

Everything above is denominator. The numerator carries its own definition and it needs writing down once: does travel between jobs count, does a warranty return count, does shop time on a Monday morning count, does a half day of training count. There is no single right answer, but there is a wrong one, which is having no answer and letting each technician decide. Two crews logging the same work under different habits will differ by more than the band is wide.

The common failure here is not a definition dispute. It is hours that were worked and never logged at all. Drive time back to the shop, the last half hour of paperwork, the parts run somebody treated as an errand rather than as job time.

In the crew above, two people leaving roughly 3 hours a week between them unlogged across 13 weeks is 39 hours. Against the true denominator of 2,072 hours that is 1.9 points, which is real but not decisive. Scale it up and it becomes the whole story: four technicians each leaving 6 hours a week unlogged across the same quarter is 312 hours, or 15.1 points on that denominator. A shop in that state reads as having a capacity problem when it has a record-keeping problem, and the tell is that the gap between paid attendance hours and logged field hours is much wider than travel, PTO and shop time can account for.

Check that gap before you touch the denominator. It is one subtraction and it tells you whether you are looking at the number's top or its bottom.

The band, and the basis it belongs to

State your basis in the same breath as your target, because the band travels with the basis and the two are meaningless apart.

On the basis used above, gross nominal available hours with PTO and holidays still in the denominator, a residential service crew that sustains 60 to 70 percent is doing well. Use 65 percent as the starting point and tune it against your own trailing four quarters, matching season to season if your work is seasonal.

A shop that nets PTO and holidays out of the denominator is on a different basis. At roughly three weeks combined per technician per year, that removes about 6 percent of the denominator and lifts the same work by roughly 4 to 5 points, so the band for that basis sits nearer 64 to 75. A shop that builds the denominator from scheduled field hours only will read higher again, sometimes above 85, and that figure answers a different question entirely: not how much of what you pay for reaches customer work, but how tightly the days you already committed were filled.

The crew above, at 78.0 percent on the gross basis, is not under-loaded. It is above the band, which is the overbooking end.

The two ends, and telling slack apart

Above the band is overbooking, and it does not present as a problem in the period you measure it. It presents one period later as premium hours, as jobs finished at the edge of the day, and as return visits on work that was rushed. Utilization is a leading indicator at that end, which is exactly why a figure that reads flatteringly low is expensive: it invites a shop to sell into capacity it does not have.

Below the band is slack, and slack has two causes that need opposite responses. Either not enough work is coming in, which is a sales problem, or enough work is coming in and it is not reaching the crew, which is a scheduling problem. The cut that separates them is cheap: count technician-days that carried no assignment at all, over technician-days available. If a meaningful share of days carried no work, the calendar had nothing to give and the constraint is demand. If almost every day carried work and the hours are still thin, the work is arriving and dispatch is losing it to travel, waiting or scope that runs short.

Running that same cut is also what catches the denominator error that runs the opposite way to the three above, where capacity is built from days that already had work assigned; see the troubleshooting case on a utilization figure that read ninety percent over an idle crew.

Two cases where the number stops carrying information

The one-person shop. The denominator is a single head, and every hour that person spends quoting, buying parts and invoicing is capacity that is real and unbillable. The figure tops out well below any band and tells them nothing they did not already know from their own week. Utilization is a number about allocation between people. With one person there is nothing to allocate, and the useful question becomes what share of the week goes to work somebody pays for.

The shop that logs every paid hour to a job. Shop time to a shop code, drive-back to the last job, a meeting to an internal code that still counts. The numerator approaches the denominator and the figure approaches 100 percent regardless of how busy anyone actually is. It has stopped discriminating, because what it now measures is how completely the crew logs rather than how much of what you pay for goes into customer work.

Those two cases bracket the rule. The figure only carries information when the numerator is narrower than the denominator by a definition somebody wrote down, and when both sides are drawn from the same list of people over the same stretch of time. Write both definitions on the same page as the number, and re-read them whenever somebody joins, leaves, or changes their standing hours.

References

• See related: Utilization Read Ninety Percent While the Crew Was Idle - the denominator error that runs the other way
• See related: Jobs per Day per Technician: Why More Is Not Better
• See related: Overtime Percent as a Capacity Signal, Not a Payroll Line
• See related: Days Sales Outstanding and the Mismatch Inside It - for a snapshot figure divided into a windowed flow