The Back Office - guides
Coffee shop labor cost percentage: the real target
Ask five operators what labor should run and you get five answers, because they are running five different businesses. Here is the honest structure - from an anonymous multi-unit operator. All figures are modeled industry ranges, labeled as such.
The number everyone quotes
Total labor cost - wages, payroll taxes, workers' comp, benefits - for a healthy independent cafe usually lands at 25-30% of net sales. Over 32% for more than a month and the shop is quietly bleeding; under 22% and you are usually understaffed and paying for it in ticket times and turnover.
| Shop shape | Typical labor % | Why |
|---|---|---|
| Kiosk / window, grab-and-go | 18-24% | One person can hold the whole shift |
| Standard espresso bar, light food | 25-30% | Two-person floor most of the day |
| Full cafe with kitchen | 30-35% | Back of house doesn't scale down at 3pm |
If you run a kitchen, stop comparing yourself to espresso-bar numbers. That comparison has ended more cafes than bad coffee.
The metric that actually matters: sales per labor hour
Labor % is a lagging number you learn at the end of the month. The number you can steer by daily is sales per labor hour (SPLH) - net sales divided by total hours worked that day.
| SPLH | Reading |
|---|---|
| Under $45 | Overstaffed or underpriced - something has to move |
| $45-65 | Healthy band for most shops |
| Over $75 | Great, until the line forms and regulars stop waiting |
Pull it by daypart. A shop doing $90 SPLH at 8am and $22 at 3pm doesn't have a labor problem - it has a 3pm problem. Cut the shift, not the week.
The cuts that don't hurt service
1. Cut by daypart, not by day
Chart transactions by hour for four weeks. Most shops find one to two dead hours where the second person rings $40 total. That is the trim - a shift that starts at 7:30 instead of 6:30 is worth roughly $6,000 a year at typical barista wages, and no customer ever notices.
2. Fix the Friday-Monday gap
Staffing to the Saturday rush all week is the classic leak. Your weekend floor and your Tuesday floor should not look alike. Stagger the second person to arrive at the rush, not at open.
3. Cross-train before you cut
The shop that can run two people at peak because both can pull shots and run food beats the shop running three specialists. Training hours are the cheapest labor you buy.
What not to cut
Manager hours disguised as barista hours are the silent inflator - an owner working 50 unpaid hours a week makes the P&L show 24% labor and a lie. Put yourself in the schedule at a market wage when you model, even if you don't take the draw. Otherwise every decision downstream is built on a number that isn't real.
Run your own numbers
The full staffing model - daypart templates, the SPLH tracker, the training ladder - is in the playbook. Get on the letter below and it goes to you first.