Franchise Network Labor Cost Analysis: Finding the Locations Driving the Drift

Franchise network labor cost analysis means decomposing the network number rather than reading it. A network labor line that drifts two points is almost never drifting uniformly: part of the move is common to every location, and the rest sits in a handful of them, usually with a personnel change underneath.

The report on your desk gives you one number for the system, one for last year, and a difference. Every decision made from it will be a network decision, and most network decisions are wrong here.

The network line is an average of forty different weeks

A network labor percentage is a weighted average, and an average conceals two things: how far apart its components are, and whether they moved together. Three quite different events produce the same headline drift.

• A common shift. A state minimum-wage step, a market-wide rate correction, or a softening denominator. Every location moves, roughly the same distance, for reasons no operator caused.
• A concentrated shift. Most of the network holds and a few locations move a long way. Different cause, different fix, same line on the report.
• A mix shift. Nothing moved, but the locations carrying the volume changed. New units at half their mature sales sit in the average at a much worse labor ratio, so a network opening steadily drifts upward while every location in it improves.

Take a network of 40 locations, weighted equally to keep the arithmetic visible, whose labor line went from 29% to 31%. Read as one number, that is a two-point problem. Decomposed, it can be almost anything.

Suppose every location moved about 1.2 points — a wage floor or a menu-price effect, and most of the drift. The remaining eight tenths sits in six locations that each added roughly 5.3 points on top. Thirty-four locations are fine. Six are not, and one is probably in serious trouble.

Those readings imply opposite responses: a budgeting conversation, or six phone calls.

Franchise network labor cost analysis in four cuts

The decomposition is not sophisticated: four sorts of the same data, and most brands have never run the second one.

1. Rank locations by contribution, not by ratio. Multiply each location's point move by its share of network sales. Your worst-ratio site is often small and stable; your biggest contributor is often a mid-table site that moved three points on a large base.
2. Split common from concentrated. Take the median location's move as your estimate of the common shift, then look at what sits above it. If the median moved as far as the network did, you have a market problem and no location list worth building.
3. Split rate from hours from sales. A location where the hourly rate rose has a market problem; one where hours rose has a schedule problem; one where sales fell has a labor line that is a symptom of something else, and cutting hours there will make it worse.
4. Split by daypart before you split by location. A location whose whole week drifted is different from one whose Friday evening drifted, and the second is far more common — which is why a location list ends up in the unit-level work of shaping a schedule block by block, and why a network memo about labor targets never gets there.

Run those four and the ranked list is short. Six locations, sometimes four, and rarely the ones the field team has been worried about.

Under a concentrated move there is usually a person

One pattern is worth knowing before you make the calls, because it changes the conversation.

When a location adds several points of labor over a couple of months, the cause is rarely a decision anybody made. It is that whoever used to build the schedule stopped building it. A general manager left. An assistant was promoted into a job nobody trained them for. A shift lead who quietly fixed the rota every Tuesday moved to days.

Scheduling is the most consequential unwritten skill in a restaurant. Whoever holds it learned it by getting it wrong for a year, and when they go the location has lost a calibration, whatever the org chart says it lost.

So before the call, check three dates: when the drift started, when the last manager change was recorded, and when a new scheduler was last trained. Eight to ten weeks between the second and the first is the common gap, and long enough that nobody at the location connects them.

It is the same signal shape as the purchasing drift behind franchise approved vendor compliance: a small unexplained move in a controllable line, weeks before anything appears in the report built to catch it.

The data will fight you before the locations do

Two locations can post different labor percentages while running identical weeks, and you will spend a month chasing it.

Salaried managers included at one site and excluded at another. Payroll taxes and workers' compensation in the line for some, below it for others. One location pulling hours from the point-of-sale clock while its neighbor reports from payroll, which differ by every unapproved minute. Sales net of discounts here, gross there.

None of that is dishonesty; it is what happens when a definition is published once and never checked — the network reporting version of compliance data drift: the number arrives on time, in the right format, meaning something slightly different at every site.

So before you rank anything, rebuild the labor line of your top and bottom five by hand from source. If two of the ten fall off the list, the list was measuring bookkeeping. The caution applies upward too. PAR's QSR Operational Index quick-service labor at 26.69% of sales for 2024, after 28.35% in 2023, and an external target is worth nothing unless your own line is assembled the way the index assembles its own.

Six conversations, not a network memo

Decomposition matters because the response to a concentrated problem and the response to a common one share no steps.

If the move is common, the available answers are pricing, schedule shape and a revised target. Telling forty operators to watch their labor when thirty-four did nothing wrong spends credibility you will want later — the same mistake as funding a national campaign to fix a problem in three trade areas, which is the argument running through franchise national vs local marketing budget.

If it is concentrated, the response is a coach on a plane, in the order the contribution ranking gives you — not a target letter. Somebody sitting with whoever now builds the schedule, rebuilding one week with them.

Then publish the decomposition to the network, including the part that is nobody's fault. An operator told the network moved two points hears an accusation. An operator shown that the market took 1.2 of it and six named sites took the rest hears an organization that knows what is happening inside itself — and the next number you ask them for arrives faster.

Six names on a ranked list, and the actual work starts when they become a schedule: franchise labor scheduling by daypart.