A customer looks for their size on the rack. It's not there. They ask — or they don't — and they leave. The garment existed: it was in the back room, fifteen meters away. The system said «4 units available», so no alert fired, nobody restocked and nobody recorded a lost sale. That's the size stockout: the stock exists, just not where it sells.

Why the system doesn't see it

  • The system counts the store, not the floor. «4 units» can mean four in the back room, two in the fitting room and none on display — to the ERP it's all the same.
  • Store inventory is wrong to begin with. Between unverified receipts, misplaced returns and shrinkage, what the system says and what's actually there diverge a little more every week.
  • Counting by hand is so expensive it's done rarely. A full store count takes a whole day — so it happens a few times a year, and the rest of the time you operate blind.

What changes with unit-level RFID

Every garment carries an RFID tag with a unique serial number — the same hang tag as always, with an inlay inside. With that, the operations that are expensive today become trivial:

  • A full count takes minutes, not a day. You sweep the store with a reader and hundreds of tags respond at once. Counting daily stops being unthinkable.
  • Receiving doesn't open boxes. The box passes the reader and the system knows exactly which units arrived — against what the distribution center says it shipped.
  • Replenishment works by exception. The system distinguishes floor from back room: when size M runs out on display but there's stock in the back, it generates the restock task — before the next customer can't find it.

The effect on sales

Size stockouts aren't solved by buying more inventory — they're solved by knowing where the inventory you already bought is. With accurate inventory and replenishment by exception, the right size is on the floor when the customer looks for it. And the same tag serves the point of sale and theft protection: one label, three functions.

Where to start

With a pilot store: tag the inventory, measure the real accuracy against the system — the first count always surprises — and run a few weeks with frequent counts and replenishment by exception. The impact is measured where it matters: in sales of the sizes that used to stock out.