A monthly cycle count for clothing inventory means counting one zone of your stock at a time, matching physical units to your SKU records, logging each variance, and correcting the system while listings stay live. You never need to close your store or count everything in one weekend.
Here is the problem it fixes. You sell a flannel shirt on a Tuesday, go to pull it from bin C-14, and it is not there. Now you have a cancellation, a defect on your seller metrics, and a buyer who was told the item was in stock. One missing item costs you the sale, and a pattern of them costs you far more. A cycle count catches these gaps before a buyer does.
What a cycle count is, and why clothing resellers need one
A cycle count is a rolling, partial stock count. Instead of one annual or quarterly full count, you divide inventory into zones and count one zone per session. Over a month, every zone gets counted at least once.
Clothing inventory drifts for ordinary reasons:
- Mis-binned items. A jacket gets put back in the wrong bin and the SKU system says it is somewhere it is not.
- Unlogged sales. An item sold on another platform and never marked sold in your records.
- Unlogged returns. A return arrives, goes on a shelf, and never gets restocked in the system.
- Duplicates and double-listing. Two units sharing one SKU, or one unit with two SKUs.
- Personal-pile leakage. Items pulled for keeping, gifting, or donating that never left the count.
Most of this is small. The point of counting monthly is that small errors stay small. A 2% variance in a 400-item inventory is eight items, which you can chase down in an afternoon. After a year of no counting, the same drift becomes a project.
Why you do not need to shut down to count
The common fear is that counting means freezing listings. It does not, because you are only counting one zone at a time. Sales on items in other zones continue as normal. For the zone you are counting, you handle sales with a simple rule: pull sold items first, then count what remains.
Here is how the two approaches compare:
| Factor | Full shutdown count | Rolling cycle count |
|---|---|---|
| Listings paused | Often all, or at least a risk of oversells | None |
| Session length | Many hours or a full weekend | 45 to 90 minutes |
| Errors found | All at once, hard to trace | A few per zone, easy to trace |
| Frequency | Quarterly or yearly | Every zone monthly |
| Fatigue and miscounts | High | Low |
The 6-step monthly cycle count routine
This is the full procedure. Run it once per zone, and rotate through zones across the month.
- Pick one count zone. Choose a single bin, shelf, or rack section of 40 to 80 items. If you use bin locations, one bin range works well. Schedule the zone on your calendar so every zone gets hit once per month.
- Export or print the expected list. Pull the SKU list for that zone from your system: SKU, title, and bin location. This is what the system believes is there. Do not look at the shelf first.
- Process pending sales for the zone. Check for orders on items in this zone that are not yet shipped or pulled. Pull those items and set them aside in a labeled ship-ready tote. This keeps live sales from showing up as false shortages.
- Count the physical units. Go bin by bin. Read each SKU label, tick it off the expected list, and note any item on the shelf that is not on the list. Check the tag size against the listing if something looks off.
- Log every variance. Record three types: missing (on the list, not on the shelf), extra (on the shelf, not on the list), and wrong location (in the wrong bin). Note the SKU, the zone, and your best guess at the cause.
- Correct the system and fix the cause. Move wrong-location items and update the bin. Add extras to the system with a cost basis. For missing items, search other zones and your returns pile before writing anything off. Only after that, mark the item as lost and book the cost.
How to handle sales while you count
Live sales are the part that makes resellers nervous, but the rule is simple. An item sold during the count is not a variance if you know it sold.
- Pull before you count. Step 3 handles orders already in your queue.
- Mid-count sales. If an order comes in while you are counting, pull that item from the zone immediately and mark it sold on your list. Do not count it as missing.
- Tight window. Keep each session under 90 minutes. A short window means few things change while you count.
- Count high-velocity zones at a quiet time. If you get most orders in the evening, count your best-selling zone in the morning.
If you crosslist, a sale on another platform is the main source of false variances. Before you count, confirm every platform sale has been marked sold in your system. A count that finds a gone item should trace to a logged sale, not a mystery.
Throughput math: how long a cycle count takes
You can plan this from a few numbers. These are example figures, so replace them with your own timed sessions.
Say a zone holds 60 items. Reading a label and ticking it off takes about 20 seconds per item with a printed list, or 10 to 15 seconds with a scanner. Then add time for variances.
| Line | Example |
|---|---|
| Items in zone | 60 |
| Count time at 20 sec per item | 20 min |
| Pull and set up (steps 2 and 3) | 10 min |
| Variances found (3% of 60) | about 2 items |
| Chase and fix at 8 min per variance | 16 min |
| Total session | about 46 min |
That puts a counting rate at roughly 80 to 100 items per hour once setup and corrections are included, and closer to 180 per hour for the pure counting pass alone. For a 600-item inventory split into ten 60-item zones, that is about 7.5 hours a month, or about two 45-minute sessions a week. Compare that to an eight-hour weekend count you dread and skip.
If your variances run higher than 5%, expect the correction step to dominate. That is a signal to fix your intake and binning process, not to count faster.
How to read your variances
The count is only useful if you learn from it. After each zone, sort the variances by cause:
- Wrong location means a binning habit problem. Re-label bins or add a put-away check.
- Missing and later found elsewhere points to the same fix.
- Missing and never found is a real loss. Book it and note the cost basis.
- Extras usually point to an intake gap, such as items that never got a SKU.
Track variance rate per zone: variances divided by items counted. Under 2% is healthy for a solo operation. Over 5% in any zone means that zone needs a process fix before the next count.
There is a grading tie-in here too. While you hold each item, check that its recorded condition still matches what you see. A shirt logged as Excellent that now shows pilling in Fabric Condition, or a jacket with a loose lining that touches Structural Integrity, should be regraded before it sells. An NWOT item that has been handled and stored may no longer qualify. Catching this during a count is cheaper than handling a not-as-described return later. Also note any Odor & Cleanliness problems from storage, such as musty bins.
Make it a habit
The routine only works if it recurs. A few habits help:
- Put each zone on a recurring calendar slot.
- Keep a single variance log with date, zone, SKU, type, and resolution.
- Review the log monthly and pick one process fix.
- Count new zones after any large haul is binned.
Over a few months, your variance rate should fall because the count exposes the habits that cause errors. The goal is not zero variance. It is catching drift while it is still a handful of items.
Try a count on one zone
You do not need to commit to a system. Pick your smallest bin, print its SKU list, and run steps 2 through 6 once. Time it and write down the variance rate. If you want the expected list, variance log, and corrections in one place, FlipDesk can hold your SKUs and bin locations so a zone count takes a few clicks. Start with one zone and see what the numbers tell you.