Picking and Packing in a Small Warehouse: Lists, Locations and Error Control

A picking error costs three times: in the item sent by mistake, in the return trip, and in customer trust. In a small warehouse it can be reduced to nearly zero without an expensive system, using only three things: a fixed location per item, a pick list ordered by route, and one check before the box is sealed.

The three foundations

What surprises businesses is that the first contributes most. In a warehouse where items are "roughly there", every pick is a small search, and that adds up to more time than any other improvement you could make. It is also the easiest to implement: one day of rearranging and labelling shelves, with no tool at all.

Organising locations without a system

Not by category, but by frequency: fast-moving items near the packing bench, slow ones further away. It sounds obvious and most small warehouses are arranged the opposite way - by product type, because that is how people think.

The simple method: shelf-number codes like A3 or B7, written on the shelf and in the catalogue, and visible from where a picker stands rather than only on a label at knee height. It can be maintained in a spreadsheet where there is no system. What matters is one location per item - an item sitting in two places is an item that gets counted twice and lost during picking.

What belongs on the pick list

• The order number, not only the customer name.
• Items in route order through the warehouse, not in order-entry order.
• The location next to each item.
• Quantity, made prominent - the field most often missed.
• A tick box for each item picked.
• A note for items needing special handling: fragile, refrigerated, pre-packed.

The fourth is a particularly common error source: the picker reads the name and not the quantity, and sends one instead of three. A simple fix is to highlight quantities greater than one visually.

The check before sealing

Thirty seconds, containing three checks: that every item on the list is ticked, that quantities match, and that the order number on the list matches the shipping label. The third prevents the worst error of all - a correct parcel carrying another customer's label.

With more than one person, it is worth having the checker not be the picker. That is not distrust but a cognitive fact: someone who picked sees what they expected to see. Where that is impossible, a short break between picking and checking achieves part of the effect.

What do you do when an item is missing?

This is the moment orders stall silently. The procedure should be defined in advance rather than re-decided each time:

1. Mark the item as missing on the list, rather than leaving the order open to one side.
2. Check immediately whether it is a genuine stock gap or an item shelved in the wrong place.
3. Tell the customer the same day with two options: wait, or ship without it and be credited.
4. Correct stock in the system immediately, so the item is not sold again.
5. Record the case, because an item missing twice in a month points at a counting problem.

Point 4 is the one that gets forgotten, which is why the same error repeats: the item stays "available" in the system, sells again, and the situation worsens. The full logic is in inventory in a system versus a spreadsheet.

What to measure

Two numbers are enough: percentage of orders with an error (wrong item, wrong quantity, missing item) and average time from order to packed parcel. The second matters less than people assume - speed bought at the cost of accuracy costs more than it saves.

When analysing errors, check whether they cluster on particular items. The answer is usually yes, and usually they are items that resemble each other - two colours, two sizes, two versions of the same product. The fix there is physical: separate them on the shelf rather than adding another control. An extra control slows every pick to solve a problem that exists on three items; physical separation solves exactly those three.

What a good packing bench looks like

The bench itself affects the error rate more than people expect. What should be within reach: packaging in sizes that match what you actually ship, a label printer, and clear space to lay an order out before sealing it. A crowded bench is why an item gets left outside the box.

Two small details that repay themselves: a fixed place for parcels awaiting collection, entirely separate from the packing area, and one box for "problems" - orders that stopped and are waiting on a decision. Without the second, a problem order stays on the bench, mixes with the next ones, and is forgotten.

What changes as you grow

Up to a few dozen orders a day this manual process works well and is not worth replacing. The signs it is time to add a layer: more than one person picking simultaneously and colliding; an error rate that will not fall despite the controls; and picking becoming the bottleneck that threatens the carrier's collection window.

The next layer is not necessarily a system: often it is a division of labour - one person picks, another packs - or simply rearranging shelves by frequency. Both are cheaper than any software, and worth exhausting before adding a tool.

Sources

• Rivhit - inventory and warehouse management
• HFD - fulfilment and logistics services