Contents
Every year it is the same. Orders start climbing in October, the crew starts staying late, and by the third week of November people are working ten-hour days and Saturdays. Someone calls the temp agency, and the temps arrive in the busiest week, when nobody has time to train them. Errors go up, a good picker quits in December, and in January the warehouse is half idle.
Peak season does not have to go like that. It is the most predictable problem a warehouse has. This part uses the example distributor's year to compare four ways of staffing it, with the work each order brings taken from the earlier parts of the series.
Step 1: turn the forecast into hours
Labour planning starts with two numbers: how many orders are coming, and how many hours each order takes. The first comes from last year's sales by week, adjusted with what the sales team knows about this year. The second is a labour standard for each zone, measured on the floor. For the example, with the fixes from Parts 1 to 5 in place:
| Zone | Work | Time an order |
|---|---|---|
| Receiving | Unload and scan 0.61 pallets an order (Part 3) | 2.0 min |
| Put-away | Put away three pallets in four, cross-dock the rest (Part 3) | 1.8 min |
| Picking | Batch picking in the re-slotted racks (Part 2) | 1.8 min |
| Packing | Scan every case and pack (Part 5) | 2.5 min |
| Shipping | Stage, load and paperwork | 1.0 min |
| Total | 9.2 min |
\[\text{People needed} = \dfrac{\text{orders a day} \times \text{hours an order}}{\text{productive hours a person a day}} = \dfrac{300 \times 0.153}{7} \approx 6.5\]
A day of 8 paid hours gives about 7 productive hours after breaks, meetings and set-up. So an average day of 300 orders needs about 6.5 people, a crew of 7. At the holiday peak of about 489 orders a day it needs about 10.7, a crew of 11.
Step 2: look inside the day
A daily total hides a second problem. Each zone has its own rhythm: receiving is busiest first thing, when the trucks arrive; packing builds through the late morning; and shipping is concentrated in the last two hours before the carriers come. A crew that is big enough for the day can still be short in shipping at 13:00 while receiving stands idle. This is why cross-training matters as much as headcount, and it is what Figure 2 shows hour by hour.
Four ways to staff the year
Plan 1: lean crew plus overtime. Keep a crew of 7, sized for the average day, and cover peaks with overtime. In BC, most employees earn time and a half after 8 hours in a day or 40 in a week, and double time after 12 hours in a day. The example allows up to two hours of overtime a person a day. When even that is not enough, orders ship a day late.
Plan 2: crew sized for the peak. Hire 11 people and keep them all year. Nothing ships late, but for most of the year a large share of paid time is idle.
Plan 3: core crew plus planned temps. Keep the crew of 7, and book agency temps from the forecast, starting each one 2 weeks before the week they are needed so they are trained before the rush. Temps are assumed to work at 60% of full pace in their first week and 85% in their second, and each one costs 8 paid hours of training and 2 hours of a supervisor's time.
Plan 4: temps plus a cross-trained crew. As Plan 3, but 60% of the core crew is trained in more than one zone and moves each hour to where the work is. The crew covers the zones' different rhythms itself, so fewer temps are needed and fewer hours go to overtime.
Figure 2 is a labour heat map of the example's warehouse. Each zone's floor is coloured by how loaded its people are in that hour: blue is idle, green busy, amber stretched and red more work than people. The figures are the people in each zone, blue for the core crew and orange for temps. The day runs from 06:00 to 14:30 on a loop, and the figure tours four typical weeks of the year on its own. Choose a plan, or Compare all to see the four side by side, and pick any week from the year chart.
This browser cannot show the 3D model. The controls below still update the numbers.
People needed each week (bars) and the plan's people (line). Click a week.
- Idle (under 60%)
- Busy
- Stretched
- More work than people
- Core crew
- Temp
The result
| Plan | Core crew | Temps hired | Overtime hours | Late orders | Heavy overtime weeks | Idle time | Cost a year | Cost an order |
|---|---|---|---|---|---|---|---|---|
| Lean crew + overtime | 7 | 0 | 2,107 | 2,836 | 17 | 23% | $751,000 | $9.63 |
| Crew sized for the peak | 11 | 0 | 473 | 0 | 0 | 42% | $944,000 | $12.10 |
| Core crew + planned temps | 7 | 4 | 959 | 0 | 0 | 30% | $772,000 | $9.90 |
| Temps + cross-trained crew | 7 | 4 | 517 | 0 | 0 | 15% | $652,000 | $8.36 |
Overtime alone looks cheaper than it is. At $751,000 a year, Plan 1 costs less on paper than planned temps. But 2,836 orders ship late, and for 17 weeks of the year the crew averages more than eight hours of overtime a person a week, up to 10.0 hours. Research on long hours, best known from Pencavel's study of munitions workers, found output per hour falls once people work beyond about 50 hours a week, and that very long weeks add little. Long peaks also bring errors (Part 5) and resignations, which this table does not count.
Staffing for the peak is safe and expensive. Plan 2 never ships late, but it costs $192,000 a year more than Plan 1, and 42% of paid productive time has no work.
Planned temps buy reliability. Plan 3 ships everything on time with no heavy overtime weeks, for $772,000 a year. Without cross-training, though, temps are booked zone by zone and still leave gaps at each zone's busiest hour, which overtime fills.
Cross-training is the cheapest capacity there is. Plan 4 needs fewer temp weeks and less overtime, and its idle time falls to 15%. At $652,000 a year, or $8.36 an order, it is the lowest-cost plan and the most reliable.
A peak-readiness calendar
| When | Action |
|---|---|
| 12 weeks out | Weekly forecast agreed with sales. Labour standards checked on the floor. Temp needs by week and zone |
| 10 weeks out | Agency booked with dates and numbers. Cross-training plan for the core crew. Holiday and leave requests settled |
| 6 weeks out | Re-slot fast movers for peak items (Part 1). Clear dead stock. Check scanners and spare batteries |
| 2 to 4 weeks out | Temps start, are trained and work alongside the core crew. Short written work instructions at each station |
| Peak weeks | Daily stand-up on yesterday's numbers. Overtime capped and shared. Carrier pickups confirmed |
| After the peak | Temps released in order of need. Review what the forecast missed. Thank the crew |
A simple skills matrix
Cross-training works when it is planned and visible. A skills matrix on the wall shows who can work where, and who is learning what:
| Person | Receiving | Put-away | Picking | Packing | Shipping |
|---|---|---|---|---|---|
| Crew member A | Trainer | Trained | Trained | Learning | |
| Crew member B | Trainer | Trained | Trained | Trained | |
| Crew member C | Trained | Trainer | Trained | Learning |
Aim for every zone to have at least three trained people, and every person to be trained in at least two zones. Forklift work also needs certified operators, so plan that training early.
Measures to track
| Measure | What it shows |
|---|---|
| Labour hours an order, by zone | The labour standard. It drives every staffing number |
| Forecast against actual orders, by week | Whether the plan can be trusted, and how much buffer it needs |
| Overtime hours a person a week | Fatigue and cost. Set a ceiling and keep to it |
| Orders shipped on time in peak weeks | Whether the plan worked |
| Temp productivity by week of service | Whether the training and the start dates were right |
Common mistakes
- Booking temps in the peak week. They arrive when nobody can train them, and work at half pace when you need them most.
- Planning by headcount, not hours. Ten people at seven productive hours is not the same as ten people at eight.
- Ignoring the hours inside the day. The shortage is usually in one zone at one time, not in the day's total.
- Running the crew on overtime for weeks. Output per hour falls, errors rise and people leave.
- Skipping the post-peak review. Next year's plan starts with what this year's forecast missed.
Key points
- Peak season is predictable. Turn the weekly forecast into hours with a labour standard for each zone.
- Look inside the day. Each zone peaks at a different hour, so a crew that is big enough can still be short in one place.
- Overtime alone is cheap on paper and expensive in late orders, errors and burnout. In the example, 2,836 orders ship late.
- Book temps from the forecast and start them two weeks early, so they are trained before the rush.
- Cross-train the core crew. In the example it is the lowest-cost and most reliable plan, at $652,000 a year.
Earlier in this series: Part 1 space, Part 2 picking, Part 3 the dock, Part 4 stock records and Part 5 errors. Next: Part 7 asks when an SME needs a warehouse management system.
Sources
- Government of BC: Overtime wages, Employment Standards Act Part 4, Section 40
- Pencavel (2014), The Productivity of Working Hours, IZA Discussion Paper 8129
The distributor, its demand pattern, labour standards and rates are an illustrative example built from stated assumptions. Overtime rules are British Columbia's standard rules; averaging agreements and other provinces differ. This is not legal advice.
