Seven-Eleven Japan let stores reorder single items daily instead of forecasting from HQ

Japanese convenience stores operated in a few hundred square feet with almost no backroom, serving customers who shopped daily rather than stocking a weekly pantry, which meant shelf misses were immediate and visible — a rice ball not on the shelf at lunchtime was a lost sale that could not be recovered later that day.

Centralizing purchasing decisions at headquarters, the standard chain-retail answer, would only average demand across thousands of stores and still miss the hyper-local swings — a single rainy afternoon or a nearby school event could flip what a specific store needed within hours. Suzuki needed ordering decisions made at the shelf, informed by real sales as they happened, without asking undertrained part-time clerks to become forecasters.

Pushed real-time item sales data to clerks, letting them order per store several times daily, with no returns, so shelf-level decisions beat central forecasts.

Real-time item data lets clerks see demand as it happens, not averaged across stores.

Frequent deliveries let clerks act on hourly signals and correct quickly.

Abolishing returns forced clerks to make accountable predictions, not hedges.

When central forecasting misses hyper-local swings, move the decision to the person closest to the demand signal and give them real-time item-level data.

By 1981 Seven-Eleven Japan had 1,000+ stores dominating Tokyo; by 2003 it had 10,000+ stores and $28 billion in sales. Tanpin kanri became a standard case taught at Harvard Business School and Stanford GSB and the template for convenience-store and quick-commerce inventory systems worldwide, credited with keeping Seven-Eleven Japan's per-store sales far above chain averages for decades.