Supply chain glossary

Multi‑Echelon Inventory Optimization (MEIO)

Multi‑echelon inventory optimization sets stock levels across every stage of a supply network at once, instead of optimizing each location separately.

18. september 2026

3 min

Multi‑echelon inventory optimization (MEIO) is a method that sets inventory levels across all stages of a supply network at once – supplier, central warehouse, regional depot, store – instead of optimizing each location on its own. It treats the network as a single system, so safety stock can be held at the stage where it covers the most demand variability per unit of capital.

A single‑echelon calculation asks how much stock one location needs to reach its own service level and treats its supply as reliable. MEIO drops that assumption: a store is served by a warehouse that can run out too, so the safety stock at one stage depends on the service level the stage above it actually delivers. The output is a split of safety stock across stages plus the internal service level each stage has to hold for the customer‑facing level to survive. Per stage it needs:

demand variability measured where demand occurs, not where the order is placed;
lead time between stages and its deviation, taken from delivered orders;
target service level at the customer‑facing stage and the cost of holding stock at each stage;
supply constraints that bind the network: MOQ, order multiples, delivery calendar.

Multi‑echelon inventory optimization in practice

Where the network is one central warehouse plus the outlets it supplies, MEIO comes down to a single expensive decision: how much of the safety stock sits centrally and how much sits at the point of sale. Central stock covers the variability of every outlet at once, so the same service level ties up less capital – but it only reaches the shelf after one replenishment cycle. Pooling works where the internal lead time is short and quietly fails where it is not.

The bigger risk is the input, not the model. MEIO propagates every lead time it is given through the whole network, so a chain that feeds it contract lead times rather than lead times measured on real deliveries optimizes against a number that does not exist in its own data. Sales history holds the same trap: days with a stockout record what was sold, not what was wanted, and variability derived from them comes out too low at exactly the locations that need cover.

Veritico STOCK computes lead time and its deviation from historical deliveries and builds safety stock from three components – monthly forecast error, daily forecast variability and lead time uncertainty – with part of it held centrally when a central warehouse and its branches are planned together. It forecasts real demand rather than censored sales, and when supply is short it treats the network as one pool and allocates by target service level. See demand forecasting and inventory optimization and replenishment and allocation management.