Benchmarks for a Commercial Kitchen Service Company

Published benchmarks for this trade fail for 1 structural reason, and it is not that the numbers themselves are wrong. 7 measures are worth tracking in their place, and the only comparison that reliably holds is against your own business as it stood 12 months ago.

The short version

Every owner wants to know whether their numbers are normal. It is a reasonable question and the available answers are worse than they look.

Industry figures for field service mix residential and commercial, mix trades with completely different parts intensities, and use definitions that vary between sources. A margin figure that includes the owner's labor and one that excludes it are not comparable, and neither source usually says which it did.

Which leaves a better approach: track a small set of measures consistently, compare yourself against your own past, and use published figures only to check that you are not wildly outside the range.

The 7 measures worth tracking

Measure	What it tells you
Gross margin, split by stream	Whether the work is priced correctly
Billable hours per technician	The denominator under every rate decision
Callbacks within 30 days	Quality, and hidden capacity
Revenue per technician	Scale, and whether growth is real
Days from completion to invoice	The cash-flow number you fully control
Contract share of revenue	Predictability, and the shape of the year
Jobs closed with a piece of equipment attached	Whether any of the above can be trusted

The last row is the precondition. Where it is low, every other figure is calculated over a partial and unrepresentative sample.

Roopairs holds equipment asset records with jobs attached to the equipment, purchase orders and parts inventory, which is what makes that seventh measure something you can move.

Why published benchmarks mislead specifically

Three problems, and the second is the one nobody accounts for.

Definitions vary. Gross margin, utilization and even revenue per technician are calculated differently across sources, and the differences are larger than the variation they claim to measure.

Work mix varies more than anything else. A service company doing mostly reactive repair on old equipment and one running a large contract book have different margin structures, different parts intensities and different utilization. Both are commercial kitchen service companies.

And survivorship. Figures assembled from businesses willing to submit data are assembled from businesses who know their numbers, which is not a random sample of this trade.

What to compare yourself against instead

Three comparisons, in order of usefulness.

Yourself, twelve months ago, measured identically. This is the only comparison that controls for work mix, geography and size, because all three are held constant.

Your own segments against each other. Refrigeration against cooking equipment, contract against reactive, one region against another. The differences are real and the definitions are identical by construction.

And your quoted figures against your delivered ones. Where quoted margin and actual margin diverge on the same job type, that is a benchmark you cannot get from anybody else and it points directly at a fixable input.

Revenue per technician is the one to watch during growth

Because it catches the failure mode that headline revenue hides.

A business adding technicians will grow revenue whatever else happens. Revenue per technician tells you whether the additional people are producing at the rate of the existing ones.

Where it falls as the team grows, one of two things is happening: there is not enough work to keep everybody busy, or the administration around each job has grown faster than the field capacity.

The second is common and fixable. It usually means the office has become the constraint, which is a records problem rather than a hiring one.

How to make the numbers comparable over time

1. Write the definitions down, one line each, and do not change them to produce a better answer
2. Take the same period each time, and prefer quarters over months in a seasonal trade
3. Note what changed in the business alongside the figures, because a number without context is a number nobody can interpret
4. Keep the old ones, so the series survives the person who built it

Point one is where most attempts fail. A definition adjusted mid-series to make a figure look reasonable destroys the only property that made the series useful.

An uncomfortable figure is a sample problem first

Check the sample before changing the business.

A poor gross margin calculated from jobs where half the parts were entered at pricebook cost is a poor calculation before it is a poor margin.

A low utilization figure built from time entries reconstructed on Fridays is measuring how people fill in forms rather than how they spend days.

So the first response to an uncomfortable number is to establish whether it is real. That takes an afternoon, and it prevents the more expensive mistake of acting on a measurement error.

The same caution applies in the other direction. A figure that looks reassuring while a quarter of the work is missing from the calculation is the more dangerous of the two, because nobody investigates good news.

The pain starts around your third or fourth technician, which is where an owner's own sense of how the business is doing stops being reliable and these measures become the only view.

Roopairs runs a product tour where you can see the jobs, equipment records and purchase orders these numbers come from.

What service companies in the trade say

“Roopairs is a top notch software that is enabling us to better serve our restaurants!!! From the Estimate to Job automation, to the Inventory and Invoicing, Roopairs is just what we needed to take our service to the next level.” — Clif Schrock

Also worth reading

• Time Tracking Technicians Do Not Resent
• Field Service Tracking: What You Can and Cannot See
• Is equipment leasing worth it

Frequently asked questions

Are industry benchmarks useful for kitchen service companies?

Only as a rough range check. They mix residential and commercial, mix trades with different parts intensities, and use definitions that vary between sources by more than the variation they claim to measure.

What should you track instead?

Seven measures: gross margin split by stream, billable hours per technician, callbacks within thirty days, revenue per technician, days from completion to invoice, contract share of revenue, and jobs closed with a piece of equipment attached.

Why is the last one a precondition?

Because where a large share of jobs have no equipment attached, every other figure is calculated over a partial and unrepresentative sample of the work.

What comparison actually works?

Yourself twelve months ago, measured identically. It is the only comparison that controls for work mix, geography and size, because all three are held constant.

Why watch revenue per technician?

Because headline revenue rises whenever you add people. Revenue per technician tells you whether the new people are producing, and a fall usually means the office has become the constraint.

What if a number looks bad?

Check whether it is real before acting. A poor margin calculated from pricebook parts costs, or a low utilization built from time reconstructed on Fridays, is a measurement problem first.