Here is a scenario that plays out in franchise operations every single week. Two locations in the same system both do $25,000 in weekly sales. Corporate allocates 400 labor hours to each. On paper, they are identical.
But one location does 60 percent of its sales through delivery. The other does 26 percent. The high-delivery location has seven to ten drivers on staff at any given time. The low-delivery location has three. Both draw from the same 400-hour pool.
The result is predictable and painful. The high-delivery store is chronically understaffed in the kitchen. The low-delivery store has more sandwich makers than it needs. Both are hitting their labor targets. Neither is actually staffed correctly.
Why Sales Per Labor Hour Is Not Enough
The standard metric for labor efficiency in restaurants is sales per labor hour. It is a useful number. It is also dangerously incomplete.
Sales per labor hour treats all labor as interchangeable. An hour spent driving is counted the same as an hour spent on the make line. But those hours serve fundamentally different functions and have fundamentally different productivity profiles. A driver generates revenue by completing deliveries. A kitchen worker generates revenue by producing orders. They are not substitutes for each other, and lumping them into the same denominator obscures what is actually happening in the building.
When you dig into the numbers, a pattern emerges. Across a group of locations, the overall sales per labor hour might average $60. But the sales per driver hour could be $115 to $140, while the sales per kitchen hour is something entirely different. These are not just different numbers. They represent different operational realities that demand different staffing approaches.
The Hidden Understaffing Problem
A location with 400 hours and eight drivers on shift during lunch burns 24 hours just on drivers during a three-hour window. That is 24 hours that the standard labor model assumes are being spent making food. They are not. They are being spent in cars.
So the GM looks at the labor budget and sees that they are at target. Corporate looks at the numbers and sees compliance. But the kitchen is running with two fewer people than it needs during the busiest period of the day. Speed of service drops. Out-the-door times go red. Customer satisfaction declines. And nobody can figure out why, because the labor number looks fine.
This is not a management failure. It is a measurement failure. The metric being used to evaluate performance does not capture the complexity of the operation it is measuring.
The Opposite Problem Is Just as Real
At the low-delivery location, the situation is reversed. Three drivers on shift means only 9 hours consumed during the same lunch window. That leaves dramatically more labor available for kitchen production. The GM has more people on the line than they need. They hit their labor percentage easily. The food is fast. The customers are happy.
But the labor budget says both stores should be staffed the same way because they have the same sales volume. This means the low-delivery store is carrying excess kitchen labor that it does not need, and the high-delivery store is short on kitchen labor that it desperately needs. The system is working exactly as designed, and it is producing the wrong answer in both directions.
Splitting the Labor Model
The fix is conceptually simple but operationally nuanced: separate delivery labor from production labor in the staffing model. Instead of allocating one pool of hours based on total sales, allocate two pools based on channel-specific demand.
Production hours should be driven by in-store and pickup order volume. Delivery hours should be driven by delivery order volume and geographic factors like average delivery radius and route density. A location with high delivery demand needs more driver hours, and those hours should not come at the expense of kitchen staffing.
This requires the labor model to understand not just how much a location sells, but how it sells. The same $25,000 in revenue tells a completely different staffing story depending on the channel mix. And that channel mix can shift week to week based on weather, marketing promotions, and local competitive dynamics.
Average Wage Is a Lever, Not a Ceiling
There is a related insight that most operators discover only after years of managing labor: hours are not the only variable. Average wage matters just as much.
If a location can reduce its average wage by $0.50 per hour, the math creates room for 15 to 20 additional hours per week at the same total labor cost. That is not cutting staff. That is restructuring the mix. It might mean scheduling a part-timer for the opening shift instead of a full-timer. It might mean hiring two part-time closers instead of one full-time closer. The total hours go up, coverage improves, and the labor dollar stays flat.
But this only works if the operator can see the relationship between average wage, total hours, and coverage in real time. If they are working off a static budget that says “400 hours this week,” they will default to cutting hours when they should be restructuring the schedule.
Making Labor Visible at the Right Level
The operators who consistently hit their labor targets while keeping their teams properly staffed share a common practice: they measure labor at a granular level, by role, by shift, by channel, and they make those measurements visible to the people making scheduling decisions.
When a district manager can see that Store A is running eight drivers and two kitchen staff during the lunch rush, the problem is obvious. When they can see that Store B has the opposite ratio, the solution becomes a conversation rather than a mandate. The data does not replace the manager’s judgment. It gives them something concrete to work with.
The rack-and-stack scoreboard approach, where locations are ranked weekly on labor metrics, food cost, speed of service, and customer satisfaction, can be remarkably effective at driving behavior without top-down pressure. People move to the top of the rankings not because they are told to, but because visibility creates accountability.
The Standard Model Is Not Wrong. It Is Incomplete.
Sales per labor hour is a fine starting point. It is a terrible ending point. For any concept with a meaningful delivery component, the labor model has to account for the fact that not all hours are created equal. Until it does, the locations that need the most help will keep looking like they are on track, and the ones that need the least will keep getting rewarded for a problem they do not have.
Need a labor model that accounts for how your restaurants actually operate? Let’s Talk
