After the count: analysing differences

The count is finished and the figures are in the table: a shortage of several million so'm and a smaller surplus. Many people stop here and ask, "why so much?" In fact there is no single answer, because the total is made up of different causes. The work begins by grouping the differences by cause. Then instead of "lost money" you get a clear list of tasks: some are fixed with a document, some by changing a process, and only the last part needs further checking. Start the analysis only with differences confirmed by a second count: looking for the cause behind an unverified row is wasted effort.

Look at shortages and surpluses together

Do not read the list of differences as separate "red" and "green" rows. A shortage on one item is often tied to a surplus on a similar one: another size, another colour, another pack, or two cards for the same product. Example: "washing liquid 1 l" shows a shortage of 8 and "washing liquid 1000 ml" a surplus of 8. That is not a loss but muddle in the books. Suppose the total shortage is 9.4 million so'm and the surplus 2.1 million so'm. Once such pairs are closed, the net difference falls to 7.3 million so'm.

Sort by value

Sort the rows not by quantity but by value: multiply the difference by cost price and start with the largest. Usually it is about ten positions, not dozens or hundreds, that make up a large part of the sum. Do not spend time on the small differences yet; you will close them later against an agreed threshold.

Four groups of causes

Before you write an item off as "lost", check where it could be:

• another zone or the stockroom, the returns box;
• orders set aside for customers or packed for delivery;
• the area by the till, the display, the sample shelf;
• goods out for repair or prepared for return to the supplier.

For each large row ask three questions: what changed in the documents (latest receipts, sales, returns)? Could the goods be somewhere else (returns box, display, another zone)? Did this position show up in the previous count too? The answers place the difference in one of four groups:

• document error: a receipt miscounted, a unit mixed up, goods kept on two cards, a return not entered;
• unrecorded outflow: goods opened for a trial, spoiled, given as a sample, used inside the shop;
• process error: not counted at receiving, wrong price entered, goods moved from the stockroom to the floor without a document;
• unexplained remainder.

For example, the net 7.3 million so'm might split like this: document and process errors 3.2 million, unrecorded outflows 1.6 million, unexplained 2.5 million so'm. So 4.8 million of the difference, about 66 per cent, is tied to a definite cause.

Be careful with the unexplained remainder. Show it in the report as a separate line marked "cause not established" rather than a guess. Recheck such positions with a short recount within the next month and add a control point to receiving and issuing. If the difference repeats, that becomes grounds for a closer review, but until then there are no grounds for blaming anyone.

Look for repeats

One difference is an event; the same difference repeating is a system. Regroup the result by zone, product group, supplier and day of the week. For instance, if small spare parts come up two to four pieces short at every count, they may be received by the box, sold by the piece and written off by the box. Bringing such a position to one unit on its card closes the problem. Put results next to the previous recount too: if a cause has appeared two or three times, make it a separate task.

Turn the result into a task table

Four columns per group are enough: cause, what we do, who is responsible, by when. Correct balances in the accounting system only through a document with the reason written down, and keep the old and new values. Watch the "explained share" indicator: what part of the net difference is tied to a cause. If that share has grown by the next recount, the analysis is working.

Conclusion

A list of differences is not a verdict but raw material for analysis: pair surplus with shortage, sort by value, split by cause and look for repeats. The Pereuchet team recounts any position with a difference, confirms the result and hands it over as a ready file for Excel or 1C with a signed certificate, and the analysis is built on that.