Every one of the 52 dots below is one store on one day. The same store, in the same week, with the same staffing structure, swings 12–39 points of labor % from day to day, and most of that swing follows sales. A store can’t be graded against a flat percentage without also grading the volume it does.
Toast, daily net sales and hourly labor cost, Fri 8/28 – Wed 9/9/2026 (13 days × 4 locations). Hourly labor only — excludes salaried Store Manager and ASM, so this is not the 6010 all-in payroll figure used in the P&L, and 22% is the hourly-labor target — not the 30% total-payroll target, which is a different measure. Bay St. Louis excluded (closed 9/7).
The curve behind each store is that store’s median daily labor spend held constant and divided by sales — no model, no fitted coefficients. It shows what the percentage does when only the denominator moves.
Same horizontal scale as above. The bar is the store’s full range of daily sales across the two weeks; the ring is the sales day at which its own median labor spend would equal 22%.
How to read it. Huntsville on Friday 8/28 did $6,608 and ran 17.4% hourly labor — better than Tchoupitoulas managed on any day in the two weeks. The following Monday it did $1,724 and ran 56.8%. Labor spend moved only $172 between those two days ($1,151 vs. $979); sales moved $4,884. Tchoupitoulas’ slowest day in the two weeks ($7,346) is larger than Huntsville’s busiest ($6,608), so the two stores never once operated at the same volume.