Retail procurement carries a scale problem most other industries do not face in quite the same way: thousands of SKUs, seasonal assortment turnover, and pricing that can legitimately vary by region or banner.
Where category-level forecasting breaks down
A forecast built at the category level looks reasonable in a summary report and misses exactly the SKU-level swings that actually drive markdowns. A category can look stable in aggregate while individual SKUs inside it swing hard enough in either direction to cause a stockout on one and a clearance markdown on another, in the same week.
Regional pricing, without a shared benchmark
The same SKU frequently carries different negotiated prices across regions or banners, for reasons that made sense at the time each deal was struck. Without a shared benchmark connecting those negotiations, the gap between what one region pays and what another pays for the identical item goes unnoticed until someone happens to compare invoices directly.
Reorder timing
Reorder signals that reach sourcing too late are one of the most common ways retail loses a negotiated price it already earned, particularly on imported or long-lead SKUs where the reorder window closes fast.
The fix across all three is the same: forecast and benchmark at the SKU, not the category, with reorder signals tied directly to the registry instead of a separate planning spreadsheet.