Updated 15 November 2026 · 8-minute read
A stock count (or stock-take) checks that what's on your shelves matches what your records say. You don't have to close for a day to do it. This guide covers full counts versus cycle counts, why blind counting works, and what to do about the differences you find.
Full counts and cycle counts
A full count counts everything at once. It gives a clean snapshot, but it usually means closing or working late, and tired people miscount. A cycle count counts a small part of the stock — one shelf, one aisle, one van — on a rolling schedule, so everything gets counted over a few weeks without stopping work.
Most small businesses do best with cycle counts plus an occasional full count (for example at year end). A simple schedule:
| Items | Which ones | Count every |
|---|---|---|
| A | The fifth of items that make up most of your sales, usage or value | Month |
| B | The next third or so | Quarter |
| C | Slow, cheap items | Six months |
Counting by location is easiest in practice: “Shelf A every Monday” is a habit people keep. Put your A items on shelves that get counted most often.
Count blind
In a blind count the counter doesn't see the expected quantity. It sounds harsh, but it matters: if the sheet says 24, a quick glance at roughly two dozen boxes becomes “24”. Without the number, people actually count. Show the expected quantity only at review time, to whoever approves the differences.
How to count without closing
- Pick a small area you can finish in one go — a shelf, a bay, a van.
- Agree what “in the area” means. Stock in the delivery bay waiting to be shelved isn't on the shelf yet; receive it first or leave it out.
- Keep recording sales and use as normal. The trick is to compare each count with the expected quantity at the moment that item was counted, not with a snapshot from the start of the day.
- Count each item once, carefully. Left to right, top to bottom. Count in the unit you stock (boxes, rolls, lengths).
- Recount the big differences before anyone changes the records — ideally a second person.
- Apply the adjustments with the reason “count”, so the history shows where each change came from.
StockKit does step 3 for you: when a line is counted, it stores what the system expected at that moment. Sales and deliveries after that are kept when you apply the count, so you never undo a sale by applying an old snapshot.
Counting with a team
Split the area so two people never count the same shelf. In StockKit, everyone opens the same count on their own phone, scans or searches each item and enters the quantity; each line saves immediately with who counted it, and the progress bar shows how much is left. If someone recounts a line, the newer number wins and the history shows who changed it.
Dealing with differences
A difference (or variance) is counted minus expected. Some are inevitable; what matters is the size and the pattern.
- Sort by value, not by count. Ten missing cable ties matter less than one missing extension lead. StockKit's review lists differences by value first.
- Set a recount threshold. For example: recount anything over 2 units or over $20.
- Look for the cause before adjusting. The usual suspects are stock used but not scanned out, deliveries received short, items on the wrong shelf, damage, and unit mix-ups (boxes counted as single items).
- Watch for repeats. The same item short every month points to a process problem — maybe it's used on jobs and never scanned out.
Uncounted items
If an item is expected on the shelf but nobody counted it, decide whether it's really gone (count it as zero) or just missed (leave it for the next count). StockKit asks you this when you apply a count: uncounted lines are left alone unless you choose to treat them as zero.
After the count
- Keep the count report — StockKit lets you download it as CSV or PDF.
- Fix what caused the biggest differences: a missing scan step, a confusing shelf, a unit that needs a clearer label.
- Schedule the next area.
Prefer paper? Our printable stock count sheet has the same columns.