How to Do a Stock Take: A Step-by-Step Checklist

Last updated 5 June 2026 · ~7 min read

A stock take only earns its keep if the numbers you end up with are ones you can trust. This is a practical, repeatable checklist for how to do a stock take — counting accurately, location by location, without closing the business for a day or guessing your way through the variances.

What a stock take is (and why it matters)

A stock take — also called a stocktake or inventory count — is the process of physically counting what you actually have on your shelves and comparing it against what your records say you should have. Over time, records drift: items get miscounted on receipt, sold without a system update, damaged, or quietly walk out the door. A stock take resets your system to match reality, so the number on the screen matches the number on the shelf.

Full count vs. cycle counting

There are two broad ways to do it. A full count means counting everything in one go — usually after hours or on a closed day. It's thorough but disruptive, and most businesses dread it enough to do it rarely. Cycle counting (also called rolling counts) means counting a slice of your stock on a regular rotation — a few categories or locations each week — so the whole catalog gets verified over a cycle without ever shutting down. Many businesses use both: cycle counts through the year, and a single full count to reconcile.

The step-by-step stock take

1. Freeze movement and pick a quiet time. A count is only valid if stock isn't moving while you count it. Choose a window when you're not receiving deliveries or fulfilling orders — after hours, a slow morning, or a planned pause — and tell the team that stock is frozen until the count is signed off.
2. Tidy and organise each location first. You can't count a mess accurately. Group like items together, face products forward, clear aside empty boxes, and make sure everything has a clear home. Ten minutes of tidying per area saves far more time in recounts and disputes later.
3. Count by location, not by total. Don't try to count "how many of product X do we have everywhere." Walk one location — a shelf, a bay, a back room, a shop floor — and count what's physically in front of you. Per-location counts are easier to verify, easier to re-check, and they tell you where a discrepancy actually lives.
4. Use a scan or a count sheet — never memory. Either scan each product to look it up and record the count on the spot, or work from a printed count sheet listing what should be in that location. Scanning is faster and removes the typo risk of writing SKUs by hand; a sheet works as a fallback. Record every item, including the ones at zero.
5. Two-person spot checks on high-value items. For your most expensive or most error-prone lines, have a second person independently recount before you accept the number. The cost of a wrong count is highest exactly where the items are worth the most, so that's where a second pair of eyes pays off.
6. Record discrepancies — don't just overwrite. When the count differs from the record, note the difference rather than silently typing the new number over the old one. The size and direction of each variance is the data you'll use to find out what's going wrong. Overwrite blindly and you lose the only clue you had.
7. Investigate variances before you accept them. A gap usually has a cause: stock in a second location you didn't count, a delivery booked in but not yet put away, a return never processed, damage, or shrinkage. Chase the meaningful gaps before updating — a quick recount or a look at recent movements often explains most of them.
8. Update the system as the new truth. Once you've investigated, set your system to the counted figures. From this moment the system is the single source of truth, so every future receipt, sale, and transfer must go through it — otherwise it drifts straight back out of sync.
9. Schedule the next count. Accuracy is a habit, not a one-off. Before you close out, decide when the next count happens — the next cycle slice next week, or the next full count on the calendar — so it never again becomes a once-a-year scramble.

Common mistakes

• Counting while stock is still moving — sales or deliveries mid-count make every number suspect.
• Counting by total instead of by location — you lose the ability to see where a discrepancy came from.
• Overwriting records without noting the variance — you fix the number but learn nothing about the cause.
• Skipping the investigation — accepting gaps as "just shrinkage" hides fixable process problems.
• Treating it as a one-time event — without a schedule, the same drift is back within months.

Where stockvpro fits

stockvpro makes counts faster — scan a product with your phone camera to look it up and count it, with no manual typing, and counts are tracked per location across your warehouses and stores so you always know where a number sits. Once you've reconciled, the system becomes the single source of truth, and low-stock alerts mean fewer surprises between counts. It's free during beta.

A good stock take isn't about counting harder — it's about counting in a way you can repeat and trust. Freeze movement, count by location, investigate before you overwrite, and put a date on the next one. Do that and the number on the screen finally means what it says.