Case · retail chain, 650+ stores
A forecast that beat the chain's own auto-ordering
For a 650-store retail chain, our machine-learning model forecasts every product in every store for every day and turns it into order quantities. On the chain's real sales history, covering about 90% of its automatic orders, preliminary results before final sign-off:
- on-shelf availability, vs. 94.5% before
- 99.4%
- lost sales, at the same inventory turnover
- −37%
- order error (WAPE)
- −8 pts
This is a retailer's ordering, not a bakery. The same forecast by product, store and day drives a bake plan.


