How to measure inventory record accuracy

The owner asks, "Is our inventory accurate?" and would like one number in reply. But where that number comes from is not obvious. Some count products, some bins, some money, and some treat an error of a few pieces as "accurate". For the same count, one definition gives 82 percent and another 98.5 percent. Record accuracy is easy to calculate once you have decided what to measure and how.

Three ways to calculate it

1. By SKU: add up each SKU's quantity across all bins. The number of SKUs whose book total equals the counted total, divided by the number of SKUs. It answers "overall, is this product right?"
2. By bin: take one SKU in one bin as a line. The number of lines where book quantity equals counted quantity, divided by the number of lines. It is stricter, because misplaced stock and bleed between bins show up here.
3. By value: one minus (the sum of the absolute gap values divided by the total book value). It reflects the financial deviation, and high-value goods naturally weigh more.

On top of any of these you can add a tolerance: a difference within plus or minus 2 pieces, or 2 percent, still counts as accurate. It suits high-volume goods where a few pieces do not matter; for high-value goods, keep the tolerance at zero.

One count, three numbers

Suppose a count covers 300 SKUs, spread over 500 SKU-and-bin lines, with total book value of 200,000 dollars.

• By SKU: 276 SKUs match exactly, so accuracy is 276 divided by 300, which is 92 percent.
• By bin: 410 lines match exactly, so accuracy is 410 divided by 500, which is 82 percent. With a tolerance of plus or minus 2 pieces, 455 lines qualify and accuracy is 91 percent.
• By value: the absolute gap values total 3,000 dollars, so one minus 3,000 divided by 200,000 gives 98.5 percent.

By SKU is 10 points higher than by bin because misplaced stock cancels out in the SKU totals: one bin is over, another is short, and the total is right. Only the bin-level figure shows it. And 82 percent by bin against 98.5 percent by value suggests the errors sit mostly on low-value items. If accuracy by value were clearly lower, look first at the few high-value products.

None of the numbers is wrong; they speak about different things. Reporting 82 percent one month and 98.5 the next looks like progress when only the method changed, so once the definition is fixed, do not change it lightly.

Two easy mistakes

First, do not use the net gap. Some people measure book total minus counted total. Surpluses and shortages cancel, the totals agree and accuracy reads 100 percent even when every line is out. Measure by lines, or by absolute gap values.

Second, agree the denominator in advance:

• Do lines where book and count are both zero count? If they do, accuracy is flattered, because those lines are always "accurate".
• Do discontinued items that still hold stock count?

A sound default is to include only lines where the book or the count is not zero, with the same rule every time.

Setting a target

There is no standard figure that fits every business, so do not borrow someone else's. A more practical way:

1. Treat the result of your first count as the baseline.
2. Set targets by category: tighter for high-value goods, looser for cheap bulk goods.
3. Keep the same definition afterwards and compare monthly or quarterly.
4. Use cycle counts to supplement the data instead of waiting for the annual count.

First count and recount

The accuracy of the first count and the accuracy after recounting are two useful numbers. Continuing the example: the first count is 82 percent by bin. Recounting the 90 mismatched lines shows that 30 were counting errors. Once corrected, 440 lines match and accuracy rises to 88 percent.

So 6 points came from errors in counting itself, and the remaining 12 percent is the genuine problem in the book records. Seen separately, they tell you whether to improve the counting or the daily bookkeeping.

Pick one definition, fix the rules and stay with them, and the accuracy figure becomes useful. InventoryHub works out the gap quantity and value for every SKU and delivers a gap table with per-warehouse summaries, so you can calculate and track accuracy on whichever definition you choose.