How to read a variance report

The count is finished, and on screen or in Excel sits a table of hundreds of rows. Many people try to read it from top to bottom and tire out. In fact a variance report is simple: you compare two numbers and look at the gap between them. The real skill is reading it in the right order: what to look at first, what to leave for later, and which figure not to trust.

The main columns

Almost every variance report has the same columns:

• Article and name. Which item it is about.
• Booked. The balance in the software or in Excel, meaning how many there should be.
• Counted. The quantity actually counted in the warehouse.
• Difference. Counted minus booked.
• Difference value. The difference multiplied by the cost price, shown in so'm.
• Comment. Some reports have this column too: the counter or the person in charge notes the likely cause of the difference here.

The last column matters most, because "5 units" means something quite different for two items: one costs 2,000 so'm and the other 2 million.

What minus and plus mean

If the difference is negative, there are fewer goods than the books say: a shortage. If it is positive, there are more: a surplus. For example, the books show 120 sacks of flour, the count finds 112, and the difference is minus 8. At a cost of 9,000 so'm per sack, the shortage is 72,000 so'm.

A surplus is as much of a problem as a shortage. It signals that goods were received but never entered in the documents, or that they were mixed up with another item.

Which row to start from

Sort the report not by quantity but by value, so that the largest sums of difference rise to the top. Usually a few big positions account for most of the total loss, while the remaining hundreds of rows hold small differences.

1. First look at the 10 to 20 biggest differences and have them recounted.
2. Then hunt for unit-of-measure errors: pieces versus boxes, kilograms versus grams.
3. Review the small differences in groups: they often come from rounding or scale error.

The net difference can mislead

At the bottom of the report there is usually a total, say minus 80,000 so'm. It looks small, but you have to check what is inside. Suppose the shortages total minus 500,000 so'm and the surpluses plus 420,000 so'm. The net difference is small, yet the turnover of errors on both sides is 920,000 so'm. That points to goods being mixed up: item A sits on item B's shelf and both look wrong. Even with a small net difference, the process may be disordered. So look at three figures separately: total shortage, total surplus and the net result.

At the top of a report there is often an accuracy percentage: the number of positions with no difference divided by all positions. For example, 1,840 of 2,000 positions show no difference, so accuracy is 92 percent. This figure shows the general state of the process but says nothing about money lost: the most expensive item may hide among the remaining 8 percent.

Looking for causes

A difference does not point to a culprit by itself; it only tells you where to look. These questions usually give a quick answer:

• Was there receiving or dispatch during the count, and were the documents entered?
• Is the item sitting somewhere else, in a different bin?
• Is the pack size right: how many units are in one box?
• Has the item been set aside as returned or damaged?
• Could it be a counter's mistake: does a recount give the same number?

Until the cause is found, do not post an adjustment to the books. If you "even out" the number, the problem hides and comes back next time.

If the same difference recurs in two counts in a row, it is no coincidence. Look for a systemic cause: a wrong unit of measure, a document entered twice, or an item that is always muddled in the same spot.

Conclusion

A variance report is not for reading every row but for looking in the right place. Sort by value, check the signs and units, look at shortages and surpluses separately from the net, and do not adjust until the cause is found. If you have no time to analyse the result, the StockCheck team hands over the discrepancy list with causes in Excel and PDF.