The order goes out when inventory says so.

You find the stockout the day you need the part.

Book inventory and on-hand inventory drifted apart months ago, and the reorder points live in someone's head — not always the same person. So the order goes out in a rush, at rush prices, while other items have been sitting on the shelf for a year with nobody costing them.

• Book and on-hand inventory drifted apart months ago
• Reorder points live in someone's head, and not always the same one
• The order goes out in a rush, at rush prices
• Nobody knows what's been sitting on the shelf for a year

The threshold decides, a person approves.

Each item carries its own reorder point and lead time, computed from the last twelve months of consumption rather than a figure set five years ago. When inventory drops below the limit, the supplier order is prepared with the usual quantity and supplier: all that's left is to send it.

Gaps between book inventory and the count are displayed rather than absorbed: you see where it leaks, and since when.

• Ins and outs entered where they happen, not at the end of the week
• A reorder point and lead time per item, kept current with real usage
• The supplier order prepared, ready to send
• Inventory count gaps costed, by item and by site

What hasn't moved in twelve months has a price.

Dormant inventory costs twice: the cash it ties up, and the space it takes. So each item carries its last movement date and the value it represents, which lets you decide to clear it, downgrade it or stop ordering it.

At the other end, an accurate opening count is essential: without a reliable starting point, book inventory starts out wrong and nobody trusts the screen after three weeks.

• Last movement date and tied-up value, per item
• The items whose orders may go out without approval
• An opening count required, at least on what matters
• This isn't a WMS: no bin locations, no picking waves