How many pickers do you need? A staffing calculator that checks the queue
Pickers needed = order lines per hour ÷ (lines one picker handles per hour × the utilization you staff to), rounded up. That gives enough capacity, not short waits. This calculator also runs the queue: it shows how long lines wait for a free picker at that crew size, using your own measured time per line.
Your picking floor
Starting values are an example, not a benchmark: replace the picker time with a time study of your own floor. Target utilization is a choice; 75 to 85% is common, none is a law.
Result
| Pickers | Busy | Erlang C (random times) | Simulation, your spread |
|---|---|---|---|
| 4 | 88% | 74% wait, avg 6.2 min | press the check |
| 5 | 70% | 38% wait, avg 1.1 min | press the check |
| 6 | 58% | 18% wait, avg 18 sec | press the check |
| 7 | 50% | 8% wait, avg 5 sec | press the check |
The formula answers “is there enough capacity”. The queue columns answer “are lines released to a picker without a long wait”. Pickers walk, travel and take breaks inside your measured time per line; the model does not add hourly rosters or skills routing.
The formula
picker-hours of work per hour = lines per hour × minutes per line ÷ 60 pickers = lines per hour ÷ (lines per picker-hour × target utilization), rounded up lines per picker-hour = 60 ÷ minutes per lineThe rate per picker is yours to measure. This page deliberately quotes no standard pick rates: published figures differ widely by layout, product and method, and borrowing someone else's rate gives a confident wrong answer.
Why the formula under-staffs
The formula treats work as smooth. Real lines arrive at random and picking times vary, so there are moments when every picker is busy and a line waits. The busier the crew, the longer those waits, and the curve is steep past about 90% (see wait time vs utilization). A crew that the formula calls enough can still leave lines waiting for a long time.
Worked example: the picking template
Simflowly's picking template releases an order every 1.2 minutes (50 an hour) and a picking time that is lognormal, 4.2 minutes on average (standard deviation 1.5). That is 3.5 pickers of work arriving each hour. These are illustrative inputs, not a benchmark. At a target of 80% utilization the arithmetic says 5 pickers; the template ships with 4.
| Pickers | Busy | Erlang C wait (random times) | Simulated wait | 95th percentile |
|---|---|---|---|---|
| 4 | 88% | 6.2 min | 3.6 min | 12.6 min |
| 5 | 70% | 1.1 min | 38 sec | 3.3 min |
| 6 | 59% | 18 sec | 11 sec | 1.4 min |
Seed 7, 30 replications of 200 hours, 10-hour warm-up, lognormal picking time. Erlang C assumes exponential times, which spread more than this template's, so it shows longer waits than the simulation.
Add packing and consolidation: the next bottleneck
Staffing one stage well just moves the limit. In the same template, consolidation has 1 person at 1.03 minutes on average, so at 50 orders an hour it is 86% busy. Picking more pickers faster will not help once that station is the limit. Simflowly's insights flag the station that becomes the limit next; try the picking and pack-dispatch templates to see both.
Pickers who also pack can be modelled with a shared crew: one pool of people that several stations draw from, so at most that many jobs run across all of them at once.
Questions
How many pickers do I need per 1,000 orders?
There is no honest single number: it depends on lines per order and the time per line on your floor. The math is lines ÷ (lines per picker-hour × target utilization). For example, 1,000 orders of 2 lines each in an 8-hour shift is 2,000 lines, or 250 lines an hour. If one picker handles R lines an hour and you staff to 80%, you need 250 ÷ (R × 0.8) pickers, rounded up. R is yours to measure yours.
What utilization should I staff to?
It is a choice, and the right one depends on how variable your picking is and how long a line can wait. 75 to 85% is a common planning range, but the page shows the queue at each crew size so you can see what a given choice costs in waiting.
Does batch picking change the answer?
Yes. Batch picking changes the work per release: one release covers several orders, so lines per hour and time per line both change. Enter the released batch lines as one unit and measure its time. Simflowly has a batch size setting on stations for batched processing.
How do I get pick rates?
Time a sample of real picks with a stopwatch or from your WMS timestamps, across pickers and zones, and use the average and the spread. On the Pro plan you can import an Excel or CSV of timings and Simflowly fits a distribution to it and shows the sample size and a goodness-of-fit test.
Why does the formula under-staff?
Because it only checks capacity. Orders arrive in clumps and pick times vary, so a crew that is 80% busy on average still builds lines at the busy moments. How much depends on variability; see the wait versus utilization calculator.
Does this cover packing too?
Run it once for each stage with that stage's own time per unit. In the full simulation, Simflowly flags which station limits the flow and names the one that becomes the limit next once you fix the first.
Keep going
- Wait time vs utilization (why staffing to 100% fails)
- Bottleneck calculator
- M/M/c queue calculator
- Picking and pack-dispatch templates
- Pro plan: Excel and CSV import with distribution fitting
Updated 17 November 2026. Free to use, no account. The calculation runs in your browser; nothing you type is sent to us.