Most staffing plans I’ve seen are a headcount guess dressed up as math. Someone looks at last week, adds a few people because the forecast looks heavy, and hopes. When it works, nobody checks why. When it doesn’t, the shift ends with a backlog and a supervisor explaining overtime.
The math isn’t hard. It just has more moving parts than “volume divided by people,” and the parts people skip are the ones that make you short.
I’ve had plans that looked fine against the forecast until the shift actually started. The volume assumption was reasonable, but the schedule treated every person on the roster as if they would show up and every paid hour as if it would be productive. Once attendance, breaks, start-up time and other nonproductive time were separated from the headline headcount, the staffing gap was obvious before the shift instead of after it.
The core formula
Required headcount comes from four inputs:
Required scheduled heads = forecast ÷ rate ÷ productive hours per person ÷ attendance %
Rounded up, because you can’t schedule 0.6 of a person.
- Forecast is the workload you expect, in one unit (orders, order lines, cases, units, or pallets).
- Rate is how much one person completes in one productive hour.
- Productive hours per person is paid hours minus paid breaks minus other nonproductive time.
- Attendance % is the share of scheduled people who actually show up.
A worked example
The numbers below are made up to show the method. Use your own.
| Input | Value |
|---|---|
| Forecast | 4,600 order lines |
| Rate | 8.5 lines per productive hour |
| Paid hours | 8 |
| Paid breaks | 0.5 hour |
| Other nonproductive time | 0.25 hour |
| Attendance | 95% |
Productive hours per person: 8 − 0.5 − 0.25 = 7.25
Required scheduled heads: 4,600 ÷ 8.5 ÷ 7.25 ÷ 0.95 = 78.57, rounded up to 79.
Now check the neighbor. Schedule 78 and you expect 74.1 people to show (78 × 95%). Their regular capacity is 74.1 × 7.25 × 8.5 = about 4,566 lines. You’re roughly 34 lines short, which is about 4 labor hours of overtime spread across the crew.
So 79 covers it with no overtime, and 78 covers it with a small amount of overtime. Neither is automatically wrong. The point is that you made that call with numbers instead of a feeling.
Six mistakes that leave you short
1. You planned on scheduled heads, not present heads
If you schedule 79 and 95% show up, you have about 75 people. Skip attendance and you’ve built a hole into every shift. Use your own attendance history, by shift if it differs, and don’t use an optimistic number to make the plan look better.
2. You counted paid hours as working hours
Eight paid hours is not eight hours of picking. Paid breaks, start-up, huddles, travel, and cleanup come out first. In the example above, 8 paid hours turned into 7.25 productive hours. That’s about 9% of capacity gone before the shift starts.
3. You used one rate for every area
Inbound, outbound, and replenishment don’t run at the same speed. A single blended rate hides the area that’s actually short. Give each area its own rate, based on your own history.
4. You treated overtime as capacity
Overtime is availability, not a plan. If your schedule only works when everyone works two extra hours, you don’t have a staffing plan. Show surplus or shortfall before overtime, then show how much overtime you’d need to close it. If the answer is the maximum every day, the base schedule is wrong.
5. You planned only for the expected rate
People work slower on a bad day, with a new hire, after a system problem, or when the product mix changes. Run the same plan at a lower and a higher rate and see what happens to capacity and backlog. If a 10% drop in productivity creates a large backlog, you know where the plan is fragile before it breaks.
6. You never compared plan to actuals
A plan you don’t check is a guess you made once. After the shift, compare what you planned to what happened: completed vs. forecast, productivity vs. the rate you assumed, present heads vs. what you expected. If your rate is consistently off, fix the rate. If attendance is consistently off, fix attendance.
Where to get your rate
Don’t grab a number off the internet. Published productivity rates depend on how each source defined the work, the unit, the layout, and the product, so they usually won’t match your building. Use your own history: completed workload divided by labor hours, by area, over several weeks.
One catch is that your history is probably measured per paid hour (completed ÷ total hours worked). The planning formula uses productive hours. The two connect through this:
Paid-hour productivity = rate × productive hours ÷ paid hours
In the example, 8.5 × 7.25 ÷ 8 = about 7.7 per paid hour. If your KPI tracker says 7.7 per paid hour and your plan assumes 8.5 per productive hour, they agree. Mixing the two without converting is a common way to overstate capacity.
Two ways to plan from the same data
Plan by workload: “I have 4,600 lines tomorrow. How many people?” Use the core formula.
Plan by headcount: “I have 70 people scheduled. What can they clear?” Multiply expected present heads × productive hours × rate, then compare it to the forecast. The difference is your surplus or shortfall before overtime.
You need both. The first sets the schedule. The second tells you what to do when the schedule changes the night before.
Don’t average the percentages
Same trap as other KPIs. If you report utilization or attendance for a week, add the numerators and the denominators, then divide. Averaging daily percentages gives a small, quiet shift the same weight as a big, busy one.
A simple process that works
- Set your assumptions once: attendance, paid hours, breaks, nonproductive time, rate, overtime limit.
- Forecast by date, shift, and area, in one workload unit.
- Calculate required heads and surplus or shortfall before overtime.
- Run a lower and higher productivity scenario to see how fragile the plan is.
- Compare actuals to plan after each shift, then adjust the rate and attendance assumptions with real data.
Put it into practice
You can build this yourself in a spreadsheet. If you do, test the edge cases: attendance typed as 95 instead of 95%, a shift where nonproductive time equals paid hours, a duplicate row, a blank forecast.
If you’d rather not build, validate, and maintain it, the LogisticsDeck Warehouse Labor Productivity & Staffing Planner does the math. Set your assumptions once, add a row per date, shift, and area, and it calculates required heads, capacity, overtime, backlog, labor cost, and conservative and stretch scenarios. Then enter actuals and compare them to the plan. It’s a plain .xlsx file with no macros.
View the Labor Productivity & Staffing Planner →
Need help setting it up? See the setup guide →
The takeaway
A staffing plan is a set of assumptions you can test. Plan on the people who’ll show up, not the people on the schedule. Plan on productive hours, not paid hours. Give each area its own rate. Treat overtime as a backstop, test a bad day, and check the plan against what happened.
Then ask the only question that matters: what are we changing in tomorrow’s schedule because of what we learned today?