To do a stock count, tidy the storage area, stop or log all stock movements, count every location with a clear method, compare the result with your records, recount and investigate the differences, and post adjustments only after they are explained. Done in that order, the count tells you both how much stock you have and why your records were wrong.
A count that only resets the numbers wastes the effort. The value is in finding the causes: unrecorded transfers, picking mistakes, supplier short deliveries, damage and theft. This guide covers how to plan a stock count, run it, reconcile it and use the result in your accounts.
Why stock counts matter
Your stock figure feeds your accounts, your purchasing and your promises to customers. If the system says 40 and the shelf holds 31, you will oversell, reorder wrongly and misstate profit. A regular count keeps the figure honest.
For most businesses, a full count is also needed at the year end or before preparing accounts, because closing stock affects cost of sales and profit. Ask your accountant what evidence they expect, such as signed count sheets and a reconciliation.
Full count or cycle count
| Approach | How it works | Best for | Drawback |
|---|---|---|---|
| Full (wall-to-wall) count | Count everything on one date, often with operations paused | Year end, small stockrooms, system go-live | Disruptive, tiring, error-prone for large sites |
| Cycle count | Count a rotating slice every day or week | Ongoing accuracy in busy operations | Needs discipline and a count schedule |
| Spot count | Count selected items after an issue or alert | Investigating a discrepancy | Does not prove overall accuracy |
| Blind count | Counters do not see the system quantity | Reducing bias in any count | Slightly slower to reconcile |
Many companies do cycle counts through the year and a full count once for the accounts. Our guide to inventory management methods explains how ABC analysis decides which items to count most often.
Step-by-step stock count process
- Choose the date and scope: which sites, zones and stock types are included, including stock held at third parties and goods in transit.
- Prepare the area: remove rubbish, group like items, label locations and separate damaged, returned and quarantined goods.
- Prepare the data: print or load the count lists, and clear any pending receipts, picks and transfers.
- Freeze movements or log them. Either stop receiving and dispatching for the count, or record every movement during it with a time.
- Brief the teams on the method, units of measure and how to record partial boxes.
- Count in pairs: one counts, one records, and each location is signed off.
- Recount any location that differs from the system by more than your tolerance.
- Investigate the differences before changing anything.
- Post adjustments with a reason code and an approver.
- Release the freeze and review what you learned.
Preparing the site and the people
Most counting errors come from the preparation, not the counting. Mixed pallets, unlabelled bins, loose items and unclear units lead to double counts and omissions.
- Give every location a code and a label that is easy to read.
- Decide the unit for each product: each, box, pack or pallet. State what counts as a full carton and how to treat a part-opened one.
- Assign a team leader for each zone and a person who reconciles.
- Provide pens, tags or scanners, and a simple way to mark counted locations.
- Keep counters separate from the people who normally pick in that zone where possible, to reduce bias.
Counting methods that reduce errors
Blind counting hides the expected quantity, so people count what they see rather than confirming the system. Two-person teams catch slips, and tagging each counted pallet or bin prevents repeat counts.
Barcode scanning speeds up identification and reduces keying mistakes; the principles are in our guide to a barcode inventory system. If scanning is not available, use pre-printed sheets in location order and have counters write the quantity and initial each line.
Reconciling the differences
Compare counted quantities with system quantities line by line, and list the differences by value and by percentage. Start with the biggest value differences. Recount them, then look for the cause.
- Timing: goods received or shipped but not yet recorded, or recorded on a different date.
- Unit mistakes: boxes recorded as singles, or the reverse.
- Location errors: stock found in a different bin, counted twice or not at all.
- Picking and receiving mistakes: the wrong item or quantity moved.
- Damage, expiry and loss that was never written off.
Only when a difference is explained, or a recount confirms it, should you post an adjustment. Keep a reason code for each one so you can see patterns over time.
Stock valuation and your accounts
Once counted quantities are agreed, value them. In the UK, stock is normally valued at the lower of cost and net realisable value, so slow, damaged or obsolete items may need to be written down. Your accountant will advise on the method for your business and the evidence needed for the year end.
Worked example, with assumptions: a count finds 120 units of an item that the system lists as 140, at a cost of £8 each. The difference of 20 units is £160 of stock lost or unrecorded. If the cause is a delivery logged twice, you correct the record; if it is damage, you write it off.
Using software to count
If your counts also feed the books, our custom software team can connect them to the systems you already use, and you can reach us from our homepage.
A system that supports count sessions lets you generate lists, scan items, compare automatically and post approved adjustments to both stock and accounts. In Dika Ops we built stock movements and adjustments on the same record as finance, so inventory and finance work from one record; the product is in closed beta with a waitlist. Whatever tool you use, keep the same discipline: freeze, count blind, recount, explain, then adjust. For a broader look at tools, read our guide to inventory management software.
Common mistakes
- Counting while goods still move, with no record of the movements.
- Letting the people responsible for the stock count it alone.
- Adjusting quantities before investigating causes.
- Ignoring stock in vans, at customers, at suppliers or in transit.
- Forgetting damaged and returned goods in separate areas.
- Not recording the count date and time against the figures.
Frequently asked questions
How often should you do a stock count?
At least once a year for the accounts, and more often for accuracy. Many businesses count high-value or fast-moving items monthly and the rest less often through cycle counting.
How long does a stock count take?
It depends on stock volume, layout and team size. A small stockroom may take a morning, while a large warehouse can take several days. Good preparation and scanning shorten it considerably.
What is the difference between a stock take and a stock count?
In everyday use they mean the same thing: physically counting items and comparing them with records. Some businesses use stock take for the full annual exercise and count for smaller checks.
What do I do if the count does not match the system?
Recount the difference, check for timing and unit errors, and look for unrecorded movements. Post an adjustment with a reason only after you understand the cause or have confirmed the result.
Do I need to stop trading during a stock count?
Not always. A pause makes a full count simpler, but cycle counts and well-logged movements allow trading to continue. Choose the approach that keeps the numbers reliable.
Who should count the stock?
Ideally people who do not normally handle that stock, working in pairs, with a supervisor who reconciles. Independence reduces both mistakes and the risk of concealment.
Next step
Pick one zone and run a trial count this week using the steps above. It will show you what your real accuracy is and which preparation steps your site needs before a full count.
