Throughput

How throughput is worked out

Count what passed through, over a period you can act on. The counting is simple; the two choices are the unit and the direction.

For the unit, pick what matches the work. Outbound is usually counted in order lines, sometimes in units, parcels or pallets. Inbound is usually counted in pallets or receipts. A warehouse that handles both full pallets and single items needs two figures rather than one blended number, because one pallet and one lipstick are not one unit of anything.

For the direction, keep inbound and outbound apart. A building can be shipping comfortably while goods-in sits two days behind, and a combined figure is exactly the wrong shape to show that.

The important thing to be clear about is what throughput is not. It is a volume, not a score. It tells you what moved, and nothing at all about what it cost to move it. That question belongs to labour productivity, and the two should be read on the same page.

There is no good number

Throughput is the one figure in this glossary where comparison between operations is not just unhelpful but meaningless: a bigger warehouse has more throughput, and that is all the comparison says.

Where it earns its keep is capacity. Throughput tells you what the building did; the question worth asking is what it could do, and that is set by the narrowest point in the flow. A single packing bench, one dock door, one induction point on a sorter. Adding pickers upstream of a saturated bench produces a longer queue and the same throughput.

In practice

A warehouse ships 11,000 lines on an average day over a ten-hour operating window, so 1,100 lines an hour.

Plot it by hour and the average turns out to describe almost no hour of the day. The morning runs at 700, the hour after the second wave release runs at 1,900, and the last ninety minutes before the carrier cut-off runs at 1,700. The packing benches can clear 1,500 lines an hour between them.

So the building averages 1,100 against a bench capacity of 1,500 and looks comfortable, and yet two hours a day queue, which is where the slow tail of the order cycle time comes from. Flattening the wave releases would move more lines through the same benches than buying a third bench would. The figures here are an illustration.

What moves the figure

• The bottleneck, and only the bottleneck. Everything else is preparation for moving the bottleneck.
• Wave release and cut-off design. The shape of the day is usually a planning choice rather than a demand pattern, and flattening it is the cheapest capacity available.
• Shift pattern against the arrival pattern. Staff on at eight when the orders land at eleven costs you both ends.
• Dock and yard capacity. Two dock doors set a hard ceiling on inbound pallets however fast the put-away team is.
• Order and product mix. A week of large pallet orders and a week of single-item orders produce very different line counts from the same tonnage.
• Automation at the right point. A sorter fed by a pick face that cannot keep up raises the equipment bill and not the throughput.