What should I do?
Loading…
Under the hood
Reorder point – = lead-time demand + safety stock. Safety stock – = z(–) × σdemand – × √lead. σ comes from the forecast's own P90−P50 spread (÷1.2816) — three global LightGBM quantile models trained across 30,490 Walmart series.
How good are our predictions for this product?
How to read it: the white line is real daily sales. Left of the marker is the past. Right of it, the dotted amber line is what we predicted before seeing those days, and the shaded band is the range we promised sales would stay inside about 8 weeks in 10. The band hugging the white line = the forecast you can trust.
Your whole store, one glance
All 3,049 products in – ranked by how urgently they need a restock — so Monday morning starts with a list, not a guess. (Demo assumes 20 units on the shelf for every product — change “units on your shelf” above.)
| Product | Status | Order | Sells / day | Cushion |
|---|
Click any row to inspect that product above.
Proof it works
We replayed the final 28 days of real sales across all 30,490 products. Two managers, same store: one reorders on gut feel (average sales × delivery days), one follows RestockIQ.