How to do a stock take without closing your shop

Most shop owners know the annual stock take as a bad Sunday. You lock the door, you and whoever you can bribe count every shelf, and by hour six the numbers stop meaning anything. You lose a day of trade, you get one snapshot a year, and the count is usually wrong anyway because everybody was tired.

There is a better way to do this, and it is not new. It is called cycle counting, or a rolling stock take, and it works by counting a small slice of the shop often instead of the whole shop rarely.

Why the annual count fails

A once a year count gives you accurate stock figures for exactly one day. From day two the numbers start drifting again, and you have no idea how fast. If your figures are wrong in March, you will not find out until next January.

It is also the least accurate way to count. Counting is boring, and accuracy falls off sharply after the first couple of hours. A count that takes eight hours is mostly bad data by the end.

Count a slice at a time

Split the shop into sections. A section should be small enough to count properly in fifteen or twenty minutes, so think one bay of shelving, one chiller, or one product category. A shop with forty sections can count two sections a day and cover everything in a month, without closing once.

Do it at the quietest hour you have. Early morning before the doors open is usually better than late evening, because nobody is counting while trying to leave.

Count the fast movers more often

Not everything deserves the same attention. Your top selling lines turn over constantly, so their figures drift fastest and cost you the most when they are wrong. Your slow lines barely move at all.

A simple split works well:

• Fast movers and anything expensive or easy to pocket: every month
• Everything in the middle: every quarter
• Slow lines and long tail stock: twice a year

That is not a rule, it is a starting point. Adjust it when a section keeps coming back clean or keeps coming back wrong.

Write down what you counted, not what you expected

The single biggest mistake in a stock take is counting with the expected figure in front of you. If the sheet says twelve and you can see roughly twelve, you will write twelve. Count blind, enter the number, and let the system show you the difference afterwards.

The difference is the useful part

The count itself is not the point. The gap between what you counted and what the system expected is the point, and it is telling you something specific:

• Short, on high value or small items: likely theft, internal or external
• Short, across a whole category: more likely a receiving error, so check what was booked in against what actually arrived
• Over, on one line: usually a sale rung up against the wrong product
• Short, on anything fresh or fragile: waste that nobody wrote off

Each of those has a different fix, and you cannot tell them apart from an annual count that lumps a year of drift into one number.

Make it routine, not a project

Two sections a day is fifteen minutes of somebody’s morning. It fits into opening the shop. A whole day of closing is a project, and projects get postponed until the accountant asks.

Once counting is routine, the numbers stay close enough to trust, which is what you actually wanted. Trustworthy stock figures mean you can reorder from the system instead of from a walk around the shelves, you can spot a leak in a month instead of a year, and they feed the five numbers worth checking every week.

What you need to make this work

You need stock figures that update as you sell, and a quick way to record a count against a product. If you are tracking stock in a spreadsheet that is updated weekly, cycle counting will not help much, because the expected figure is stale before you start.

StoreWave keeps stock moving with every sale, so the expected figure is always current and the difference you find is real. Counting a section takes as long as counting takes, and nothing else.