Warehouse management: the part inventory software does not cover

Warehouse management covers the physical side of holding stock: where each item sits, how it gets received, picked, packed and dispatched, and who did what. It is routinely confused with inventory management, which counts what you own and what it is worth. The distinction matters because a company can have perfectly accurate inventory records and still take twenty minutes to find an item.

For a small operation the two overlap almost completely and one spreadsheet does both. The point at which they separate is usually the point at which more than one person picks stock, because that is when where it is stops being something one person remembers.

What warehouse management is responsible for

• Receiving: checking deliveries against the order, recording discrepancies before the driver leaves, and putting goods away in a known place.
• Locations: every storage position having an identifier, so stock has an address rather than a general area.
• Picking: getting the right items for an order, in a route that does not cross the building four times.
• Packing and dispatch: the final check that what is in the box matches the order.
• Stock accuracy: cycle counting, so the record and the shelf stay in agreement.
• Returns: putting items back into sellable stock, or not, with a decision recorded.
• Space: what is stored where, with fast-moving lines near the packing bench.

Location identifiers are the single highest-return change for a small warehouse, and they cost the price of a label printer. Without them, finding stock depends on the memory of whoever put it away, which fails on their first day off and makes the operation impossible to hand over.

Picking is where the time goes

In most small warehouses, picking is the largest labour cost and the least examined. Walking is the bulk of it. Two changes cut it substantially without any software: put the fastest-moving lines closest to where orders are packed, and pick multiple orders in one pass rather than walking the building once per order.

The second change is worth the care it requires, because batch picking introduces a new error — items going into the wrong order at the packing stage. A physical separation, one tote per order, removes almost all of it.

Counting without shutting down

The annual full stocktake is a poor mechanism. It stops the business for a day or two, it produces a large adjustment nobody can explain, and it discovers in December an error that started in March. Cycle counting replaces it with a small, continuous version.

1. Split stock into groups by how often it moves or how much it is worth.
2. Count the fast or valuable group frequently, the rest a few times a year.
3. Count a small number of locations each week, on a schedule, rather than everything at once.
4. Record every discrepancy with its cause, not just its size — miscount, mispick, damage, unrecorded return.
5. Fix the cause where the pattern repeats, rather than only adjusting the number.
6. Track accuracy as a percentage over time, so improvement is visible.
7. Keep the annual count if it is required for your accounts, but expect it to be uneventful.

When a spreadsheet is genuinely enough

It is worth saying plainly: for one or two people, a few hundred lines and orders picked one at a time, a spreadsheet with locations and quantities does the job, and dedicated warehouse software would be a cost with no return. The honest signals that you have outgrown it are concurrent editing conflicts, stock accuracy below the low nineties, or picking errors reaching customers regularly.

Ettex Sheets covers the spreadsheet stage properly — locations, quantities, reorder points and a cycle-count schedule in one file that several people can open at once, with the history of who changed what. That solves the version-conflict problem, which is usually the first thing to break.

What it is not: there is no barcode scanning, no handheld terminals, no pick-path optimisation and no carrier integration. Those are the reasons to buy a warehouse management system, and past a certain volume they are worth buying. A spreadsheet that pretends to be one just moves the errors somewhere less visible.

Give every location an address, put fast lines near the bench, count a little every week, and record why each discrepancy happened rather than only fixing the number.