What warehouse control means: the key metrics

Many warehouse owners think of control as cameras, a guard and a stock count once a year. But all of these kick in after the problem has already happened. Real control works differently: at any moment you can answer three questions with confidence.

The three questions of control

The first: where are the goods? Every item needs an exact address (rack, row, level) and that address must be recorded in the system. The second: how many are there? The system balance must match what is actually on the shelf. The third: how fast do they move? A product that ships once a month cannot be stored in the same place, and given the same attention, as one that ships ten times a day.

If you answer any of these with "roughly" or "we'll ask the accountant", control does not exist yet. The good news is that you do not need an expensive system to build it, only the habit of measuring.

The key metrics

Six metrics are enough to answer the questions with numbers:

• Stock accuracy — in how many of the sampled locations the physical count matched the system. For example, if you count 200 locations and 188 match, accuracy is 94 percent. Aim for 98 percent or higher.
• Stock turnover — how many days the stock will last. If average stock is 900 million so'm and the daily cost of goods sold is 30 million so'm, the stock covers 30 days.
• Receiving time — from the truck arriving to the goods sitting on a shelf and visible in the system. If it is long, an item you told a customer was "out of stock" is actually standing by the gate.
• Order accuracy — the share of orders picked without errors. Every error means a return, a second trip and an unhappy customer.
• Shrinkage rate — lost, damaged and written-off goods as a share of total stock.
• Dead stock share — items with no outbound movement for 90 days or more.

This does not mean you need all of them at once. A shop cares more about order accuracy; a wholesale depot cares about turnover and receiving time.

How to measure

A metric is useful only if it is calculated the same way every week. Agree on the data source once: usually the receipts and issues table in 1C or Excel. For stock accuracy you do not have to count the whole warehouse. Checking 20 to 30 locations a day (cycle counting) covers a small warehouse in a month or two, and system errors surface long before year-end.

Every metric needs an owner: receiving time belongs to the receiving clerk, order accuracy to the picking shift lead, stock accuracy to the warehouse manager. While nobody feels a number is theirs, it stays just a figure in a spreadsheet.

From number to decision

A metric on its own changes nothing; it has to lead to a decision. Imagine that in one month stock accuracy fell from 97 to 91 percent, with the gaps concentrated in the picking zone. There are three likely causes: goods were miscounted at receiving, pickers worked without a scanner, or items were put in the wrong place. To find out, recount the ten items in that zone with the biggest gaps next week and review their movement history. Usually two or three recurring patterns show up; fix them and the number recovers on its own.

Common mistakes

• Introducing more than ten metrics at once, so that none of them is followed.
• Measuring once and stopping. Control is about the trend, not a single snapshot.
• Comparing periods that do not match. Compare December sales with last December, not with January.
• Finding a number without asking why. "Accuracy is 94 percent" is information; "where did the 6 percent gap come from" is the basis for a decision.

Conclusion

Warehouse control is not an expensive program but a discipline: three questions, five or six metrics and a weekly rhythm. During a warehouse audit, SkladKontrol finds the places where exactly these numbers are getting worse and points out the loss points in receiving, storage or dispatch.