As of November 9, 2026, two European fashion retailers had put the same operational number on the record within 48 hours. On 4 November Zebra Technologies said JD Sports reached 98% on-shelf availability and 20% faster replenishment after rolling RFID-equipped handhelds to 414 stores in the UK and Ireland, and on 3 November Primark told investors it will roll RFID across its range over the coming months to fix delivery accuracy. Read together, the two disclosures describe a shift in what retailers think the shelf problem is: not the scanner, but the count behind it and the allocation above it.
What the two announcements actually say
The JD Sports programme is unusually well documented because the vendor, the retailer and the trade press each published a different slice of it. Zebra's release describes a five-store pilot that compared its handhelds with hardware from two other vendors running identical software; the readers were judged faster, longer-range and easier for staff to pick up. Rollout then averaged 26 additional stores a week, and the first store went live 76 days after the project started, which is fast for a chain that operates nearly 5,000 stores in 46 countries.
Primark's disclosure is thinner but more pointed. Speaking at Associated British Foods' annual results on 3 November, chief executive Eoin Tonge framed RFID as part of a wider supply chain and data programme, and ABF chief executive George Weston supplied the diagnosis: delivery accuracy is insufficient, goods are sent to stores that do not need them while customers elsewhere cannot find a size, and summer collections are pulled nationwide in Spain at the same moment even though Mediterranean stores could sell them for another three to four weeks.
Why 98% availability is not 98% accuracy
The distinction matters because the two numbers are routinely conflated in vendor material. On-shelf availability measures whether a size that exists somewhere in the building is out where a customer can see it. Inventory accuracy measures whether the count in the system matches what is physically on the floor and in the stockroom. A store can push goods to the shelf efficiently and still hold a count that nobody trusts, which is exactly the gap Daniel McGrath, JD Sports' head of customer operations, described in May 2026 when he asked what the point of 98% accuracy is if an agent cannot trust the last 2%.
| Measure | Question it answers | JD Sports figure |
|---|---|---|
| On-shelf availability | Is the size on the floor for a shopper to pick up | 98% |
| Inventory accuracy | Does the system count match physical stock | Being rebuilt through weekly counts |
| Replenishment speed | How fast stock moves from backroom to shelf | 20% faster |
| Order completion, single-store orders | Can the store fulfil what only it holds | Up 4 to 8% |
The count that went from a day to 25 minutes
The operational change is easier to see in the audit than in the availability figure. Before RFID, JD Sports ran a physical count every 90 days that tied up ten people for a full working day. The retailer now counts an average store in 25 minutes and its 60,000 square foot Trafford Centre store in about an hour, which makes counting twice a week realistic rather than a special project. Colleagues also move 95% more inventory from the stockroom to the floor, a figure that speaks to replenishment behaviour rather than to scanner speed.
Primark's admission is about distribution, not scanners
The allocation error that no tag can catch
Allocation errors sit upstream of the shelf and survive any counting technology. If a national plan sends the same summer assortment to every store on the same date, the southern stores lose three to four weeks of selling while the northern stores start clearing early. RFID tells you what is on the floor; it cannot tell you that the floor should have received a different mix, which is why Primark paired the technology announcement with a demand-driven distribution argument.
Primark's numbers are smaller in public but larger in implication. The chain runs more than 500 stores in 19 countries and reported about 9.8 billion pounds of retail sales in fiscal 2026, so a small improvement in allocation accuracy moves more value than a perfect count in one flagship. Weston's example is a planning failure, not a scanning failure: if Spain clears summer stock nationwide on a calendar date while southern stores still have three to four weeks of selling left, the loss happens before any tag is read.
| Retailer | Stores and markets | Stated problem | Stated fix |
|---|---|---|---|
| JD Sports | 414 live, nearly 5,000 in 46 countries | Inventory visibility blocked one-hour collect and 30-minute delivery | RFID counts twice a week plus handheld rollout |
| Primark | More than 500 in 19 countries | Stock delivered to the wrong stores, sizes missing elsewhere | RFID across the range over coming months |
For brands selling through either chain, the lesson is about the data you will be asked to supply. If a retailer can count twice a week and sees availability by size, its buyers will expect size-level forecasts and pack configurations that match store-level demand, not national averages. Suppliers that still plan by country will find the conversation moving to why a specific size is short in a specific store, and they will need the same granularity to answer.
What to measure before the peak week
A practical review starts with the shelf rather than the warehouse. First, pick the fifty styles that carry the holiday plan and measure availability by size in a sample of stores; second, compare that with the system count to separate shelf gaps from count errors; third, check whether the allocation that produced the gap was national or store-level; fourth, agree a weekly count cadence for the peak weeks. Each step produces an artifact that can be repeated after the season.
None of this requires new hardware. Most suppliers already receive sell-out and stock files from retail partners; the gap is usually that the file is monthly, style-level and incomplete on sizes. Asking for size-level weekly data before the peak is a cheaper intervention than a technology programme, and it is the input any RFID-equipped retailer will now expect to receive.
FAQ
What did JD Sports and Zebra actually announce?
On 4 November 2026 Zebra said JD Sports reached 98% on-shelf availability and 20% faster replenishment using TC2 handhelds with RFD40 RFID sleds, rolled out to 414 stores in the UK and Ireland in about three months.
Is 98% on-shelf availability the same as 98% inventory accuracy?
No. On-shelf availability asks whether a size that exists in the building is on the shop floor; inventory accuracy asks whether the system count matches physical stock, and JD Sports has used the same 98% figure to describe the accuracy problem it is still solving.
Why does Primark's RFID plan matter for brands?
Because Primark linked the investment to delivery accuracy, saying stock reaches stores that do not need it while shoppers elsewhere cannot find sizes, which shifts the argument from store technology to distribution planning.
