JD Sports Replenishment Technology: Measure Shelf Availability, Not Planning Speed

Faster replenishment planning is useful only when shoppers find the right product and size on the shelf. A system can produce recommendations in minutes while inventory remains trapped in a distribution center, delayed in transit, waiting in a store's back room, or recorded against the wrong location.
JD Sports offers a useful benchmark for closing that gap. The retailer is using RFID-enabled inventory visibility across its omnichannel operation. Supply Chain Dive reports that the program has helped JD Sports reach 98% on-shelf availability and achieve 20% faster replenishment. Those numbers belong together: speed matters, but availability is the customer outcome.
A faster plan is not the same as a faster shelfโ
Retail replenishment is a chain of decisions and physical events. Demand is forecast, inventory is allocated, the distribution center releases an order, a carrier moves it, the store receives it, and an associate places it on the sales floor. Measuring only the first step hides every delay after it.
That distinction is especially important in footwear and apparel. A store may have plenty of units for a shoe model but still lose the sale because the customer's size is missing. A category-level or style-level availability metric can therefore look healthy while the actual sellable assortment is broken.
JD Sports' rollout shows the scale at which execution data can improve. The retailer piloted RFID mobile computers in five stores, brought the first store live in 76 days, and then deployed at roughly 26 additional stores per week. The program reached 414 UK and Ireland locations in about three months, according to Supply Chain Dive. Associates can now move more than 95% of inventory from the storeroom to the shop floor, while counts can run twice weekly or even daily instead of waiting for periodic audits.
Connect the full replenishment timelineโ
A reliable control tower should create one timeline for every store-SKU replenishment movement. At minimum, capture:
- forecast creation and last revision;
- allocation recommendation and approval;
- distribution-center wave, pick, and release;
- carrier tender, pickup, and arrival events;
- store receipt and inventory-system update; and
- back-room-to-shelf confirmation.
Use a shared identifier across the allocation, shipment, receipt, and inventory records. Without that link, teams debate whether planning, warehouse operations, transportation, or store execution caused a stockout. With it, they can quantify where elapsed time accumulated and assign the exception to an owner.
This also prevents a common analytical mistake: treating forecast accuracy as the final objective. A forecast can be accurate in aggregate and still put inventory in the wrong locations. SupplyChainBrain's analysis of forecast decisions emphasizes that the value comes from turning forecasts into operating and capital-allocation decisions. Store-level availability reveals whether that translation worked.
Measure availability at size levelโ
Three metrics should sit beside replenishment cycle time.
Size-level in-stock rate measures the percentage of active store-SKU-size combinations with at least one sellable unit when demand is expected. Weighting the metric by expected demand keeps slow sizes from masking shortages in the sizes customers buy most often.
Late allocation rate measures allocation lines approved after the latest feasible distribution-center release or carrier cutoff. Segment it by cause: delayed forecast, constrained supply, manual approval, master-data error, or capacity. This separates a planning delay from an execution delay.
Lost-sales exposure estimates the units and margin at risk while a demanded size is unavailable. Use observed sales before the stockout, comparable-store demand, online searches, substitutions, and abandoned baskets. It is an estimate, but it gives teams a better priority signal than treating every missing unit equally.
Add supporting measures for back-room dwell, receipt-to-shelf time, inventory-record accuracy, fill rate, and promise failures. For omnichannel operations, compare the system's available-to-promise quantity with the physical RFID count. A mismatch can create both a disappointing click-and-collect cancellation and an unnecessary replenishment order.
Prioritize exceptions by service and demand riskโ
A long exception queue is not operational control. Rank exceptions by the commercial outcome they threaten. A practical score can combine:
- demand confidence for the store-SKU-size;
- expected lost margin during the shortage window;
- customer-promise risk, including click-and-collect orders;
- available substitutes in the same store or nearby network;
- time remaining before the next release or transport cutoff; and
- confidence in the inventory record.
For example, a high-velocity shoe size with confirmed online demand, no local substitute, and a carrier cutoff in two hours should outrank a slow-moving color variant that has alternatives on the rack. Scarce inventory can then be allocated to the location with the strongest demand signal and highest service risk, rather than split evenly across stores.
Exception rules should prescribe an action as well as raise an alert. The action might be to advance a distribution-center release, transfer stock from a nearby store, expedite a shipment, investigate an RFID discrepancy, or pause an online promise. Record the decision and its outcome so thresholds improve over time.
Build governance around the customer outcomeโ
Review availability and speed together each week. If replenishment time falls but size-level in-stock does not rise, inspect allocation quality, transport cutoffs, receipt processing, and back-room dwell. If physical counts improve but cancellations persist, investigate the synchronization latency between RFID, order management, and available-to-promise systems.
Assign each stage an owner and set service targets between stages, not only end to end. Planning owns allocation timeliness, the DC owns release performance, transportation owns movement milestones, and stores own receipt-to-shelf execution. A shared dashboard keeps these targets from becoming isolated departmental metrics.
JD Sports' reported results show the right pattern: combine faster replenishment with a measurable shelf outcome. The goal is not to make the planning engine look fast. It is to make demanded inventory available where and when a customer wants to buy it.
Ready to connect replenishment plans, transport milestones, and store-service exceptions? Request a CXTMS demo to see how one operational view can turn faster decisions into better availability.


