EN
Back to the archive

The encyclopedia · Software & IT · Operational decision · 2022–2026

ECCO's stochastic optimizer automated store replenishment for 536 shops in 27 countries.

ECCO replaced manual store ordering with a stochastic solver automating ~300,000 replenishment orders a month, saving millions a year.

ECCO

the move

Shoe maker ECCO sells through a large network of its own retail stores, and each store owner made replenishment decisions by judgement. That meant thousands of small, inconsistent choices, each prone to error, and no global picture of where inventory should go.

ECCO's Data & AI team built Intelligent Auto Replenishment (IAR), a large-scale stochastic mixed-integer program. It optimises replenishment orders for all stores at once, with a novel 'divide and conquer' method that makes the online-scale problem tractable.

The system unifies the global inventory strategy, automates nearly 300,000 replenishment orders a month, and — by reducing key operational costs by 1.09% — delivers several million euros in annual savings. A 1% cost reduction might look small, but on a thin retail margin it is the difference between a good and a bad year, and it is achieved quietly and repeatedly.

IAR made ECCO a 2026 Franz Edelman Award finalist, and the work became a template for centrally automating retail replenishment across a vertically integrated global brand.

why it works

  • One model makes consistent decisions across hundreds of stores where humans did not.
  • Stochastic programming handles demand and lead-time uncertainty explicitly.
  • A 1% cost cut is large when margins are thin and the decision repeats every month.
  • Automating ~300,000 orders a month removes a huge amount of manual work.
the payoffLet one stochastic model decide every restock.clever

what transfers

If the same decision is made hundreds of times by hand and inconsistently, consolidate it into one model that makes them all at once and better.

what came after

ECCO's Intelligent Auto Replenishment became a 2026 Franz Edelman Award finalist and illustrated how a global retail brand could move from thousands of individual judgement calls to a single, centrally optimised replenishment policy. The gain — a roughly 1% reduction in key operational costs, equivalent to several million euros a year — came not from a dramatic change but from making a recurring decision consistently well.

references

spotted an error? The archive wants to know.

same kind of clever