Reading a difference list: patterns that point to the cause

The count results are in: 87 items differ. The easiest response is to adjust the system to the counted figures and declare that books and stock agree. The price is that the same problems come back next year. A difference that is smoothed over often carries useful information: where the process leaks and which kinds of items go wrong. In fact the difference list itself holds clues, and reading them saves a lot of wasted walking.

Five common causes in one line each

1. Recording errors: a wrong quantity typed, the wrong item chosen, one document entered twice, or a document not entered at all.
2. Mixed goods: the same product under two codes, similar items in neighbouring locations, or goods put in the wrong place.
3. Timing: document dates that differ from the real movement, or a cut-off that was never drawn clearly.
4. Physical loss: damage, spoilage, cutting waste, weight change and also theft.
5. Unrecorded movements: samples, gifts, returns, transfers, items sent for repair or used in production without a document.

The clues in a difference list

Each cause leaves a different trace in the data:

• Recording errors: the difference is a neat multiple, or looks like swapped digits, such as 120 typed as 210, and often sits on one document.
• Mixed goods: they come in pairs, one item short and one over, with similar codes or names and roughly equal quantities.
• Timing: the difference disappears when you move the cut-off by a day.
• Physical loss: the difference is steady, no document explains it, and it is typical of small, valuable or perishable items.
• Unrecorded movements: quantities are round numbers, such as 5 or 10, concentrated in a few items that sales or projects often take.

Another clue is repetition. If the same item, location, shift or supplier causes trouble at every count, the fault lies in the process and not in one slip.

Four quick tests in Excel

1. Sorting test: sort by the absolute value of the difference and look at the top 10 lines first, because they set the tone of the whole result.
2. Pairing test: place items with similar codes or names on neighbouring rows and see whether shortages and surpluses come in pairs of similar size.
3. Timing test: compare each difference with the last documents for that item around the cut-off, and see whether it disappears when the cut-off moves by a day.
4. Multiple test: check whether the difference is an exact multiple of the carton size or the pallet quantity.

After these four tests most differences already have a first classification, and only the rest need a walk to the shelf or a conversation. The order of checking is paper first, shelf second, people last. When you do ask, ask "what happened to this document afterwards?" rather than "who took it?".

An illustration

Suppose 87 items differ, with gross differences of 64 million UZS, and the ten largest account for 41 million. Going through them: 4 are mixed goods (offsetting pairs, 18 million together), 3 are misdated documents (12 million), 2 are samples issued without a record (6 million) and 1 cannot yet be explained (5 million). 18 + 12 + 6 + 5 = 41. (The figures are invented.) Of the biggest differences, only 5 million needs further investigation. The rest can be fixed by raising documents or correcting codes.

When you report to management, a simple table works best: cause group, number of lines, amount, action and owner. Managers decide faster from that table than from an eighty-line difference list.

Fix the cause, not the symptom

• Recording errors: add a second check, or a second signature on key documents.
• Mixed goods: unify item codes and add labels that tell similar items apart.
• Timing: require documents to be entered on the day of the movement, and draw a clear cut-off at month-end.
• Physical loss: improve storage and protection, and agree a reasonable loss allowance with management.
• Unrecorded movements: keep a register for samples, gifts and items sent for repair.

Fixes are aimed at processes, not at blaming one person.

In short

Only a difference that is understood can be fixed. UzWarehouse's warehouse audit and difference tracing follows each count difference to its source, whether entry error, mixed goods, loss or process gap, and finishes with practical recommendations.