RefreshFC forecasts demand for every fresh SKU in every store — daily, 21 days out. Weather, promos, holidays, and shelf life in one number your buyers can act on before the truck leaves.
A single warm weekend can swing berry demand 40%. Most buyers still order off last year's numbers, then absorb the difference as shrink — or watch shoppers walk past an empty case.
Read-only feeds from your POS, DSD receipts, and promo calendar. Two hours of IT time, no migration, no new hardware in store.
Models trained per-category on weather, local events, price elasticity, and harvest windows — with shelf life carried through so the number is orderable, not academic.
Case-rounded quantities land in your ordering system with the reason attached. Buyers keep the final call and the model learns from every override.
Each category gets its own model family — bananas don't behave like bagged salad, and neither behaves like a Tuesday in February.
“We cut berry shrink by a fifth in one quarter. The part my buyers actually love is that it tells them why the number moved.”
A backtest on your own stores in five business days — the shrink you'd have avoided, SKU by SKU, before you commit to anything.