How to measure inventory accuracy

Ask a shop owner how accurate their inventory is and you will hear 'pretty good.' Ask how they know and the room goes quiet. Most shops have never measured it, so they are running on a feeling.

The good news is that accuracy is easy to measure and the answer is just one percentage. The slightly less good news is that the first time you measure it, you will probably be disappointed. That is normal, and it is useful.

The simple method

Pick a sample of items. Count them on the shelf. Compare each count with the system quantity. An item is accurate if the two agree, or are close enough for your purposes. Then divide the number of accurate items by the number you counted.

What counts as a match

You have to decide how strict to be. For a small, cheap item, being off by one may not matter. For an expensive or regulated one, it does. A common approach is to use a tolerance that scales with value: exact for the expensive items, a small allowance for the cheap ones.

Whatever you pick, write it down and use the same rule every time. A measure that changes shape each month cannot tell you whether you are improving.

What good looks like

There is no universal target, and a number from another business is hard to compare with yours. Many guides suggest the high 90s as a goal for businesses that run formal counting programs. For a small shop starting out, a more useful target is simple: higher than last quarter.

The point is not the score. The point is the trend.

Moving the number

• Trace the misses. Look for patterns by supplier, shelf or task.
• Count the receiving door. Many errors enter there.
• Tighten what happens when something is damaged or returned.
• Count the A items more often (see Cycle counting: count a little, all year).
• Measure again in a month, with the same rule.

Most gains come from two or three habits, not from a clever system. Counting deliveries, writing off damage the same day and giving every item one home usually do more than any software setting.

What the misses tell you

The percentage is the headline. The misses are the story. After you count, sort the differences into piles. Short or over? Big or small? One supplier, one shelf, one shift?

• Mostly short: look for unrecorded picks, damage and theft.
• Mostly over: look for receipts entered twice, or returns not recorded.
• Clustered on one shelf: look for mixed bins or confusing labels.
• Clustered on one supplier: look at how deliveries are counted in.
• Random: probably counting slips. Recount and do not read too much into it.

It is the same exercise a doctor does with a blood test. One number says 'something is off.' The pattern says what to do about it.