In the first week of November, 28 of 44 ASINs were 10,760 units short.
A kit business. Selling a microgreens kit means holding hemp mats, seeds, soil, a card and a postcard, each from a different supplier on a different lead time. Forecasting the kit tells you nothing, so we forecast the components. The peak plan then gave every line 30 days of cover with no exceptions, and that is what surfaced the shortfall at the start of the month rather than after it.
A reorder is never triggered off one warehouse.
18,617 units on hand across three nodes, with 3,006 more already moving.
Two forecasts, five windows, and kits planned at the part.
The two methods are kept side by side and never averaged, because on this catalog they disagree by a median factor of 1.96.
Two independent forecast methods kept in parallel rather than averaged into one. On this catalog they disagree by a median factor of 1.96, and by more than two times on 20 of 40 ASINs.
Velocity held at 7, 14, 30, 30-day marketplace and 90 days plus year to date, across 95 ASINs.
Safety stock applied to every line in the peak plan, no exceptions granted. It is what surfaced the November shortfall.
Gated by a single hemp mat component, which is why multi-part kits are planned at the part rather than the finished good. Every part carries its own supplier and its own lead time.
If one component can stop eleven listings, plan below the finished good.
Bring the kit whose parts never arrive on the same clock. Thirty minutes with a senior lead.