Stock Smart — Multi-Echelon Inventory at Noordveld Components

A four-round, intermediate inventory and operations simulation set inside Noordveld Components B.V., a €148M Tilburg-based distributor of industrial drive components serving the Benelux and western Germany. Noordveld runs a two-echelon network — one central distribution centre (CDC) holding 6,200 SKUs that replenishes four regional stocking points (RSPs) in Rotterdam, Eindhoven, Antwerp and Düsseldorf — and guarantees a 97% line-item fill rate within 24 hours to its top accounts. On 9 March 2026 the promise is breaking: measured fill rate has slipped to 91.4%, yet inventory has ballooned to €31M and turns have fallen from 4.1 to 3.6. The board gives the planning team one S&OP cycle to fix service WITHOUT adding net working capital — or it outsources planning to a 3PL, cutting nine roles. Low service and high stock at once is the classic signature of inventory at the wrong echelon. Playing the supply-chain planning team, you work the multi-echelon logic of Ton de Kok's research. Round 1 (Diagnose): segment the 40-SKU basket by volume AND variability (coefficient of variation), locate the €5.8M of dead RSP buffer, and find the under-buffered fast movers actually driving the fill-rate miss. Round 2 (Plan): set a differentiated cycle-service-level policy and size safety stock with SS = z × σ, building reorder points on demand-during-lead-time — discovering that service is non-linear, with slow lumpy lines costing ~3× the buffer per service point of the fast movers. Round 3 (Decide): place safety stock across the two echelons, exploiting risk pooling (the square-root law) by consolidating lumpy demand at the CDC while forward-deploying the fast stable lines, funding the redesign cash-neutrally from the dead stock and optimizing the network jointly rather than echelon-by-echelon. Round 4 (Recover): stress-test the policy against a supplier lead time that doubles from 3 to 6 weeks and a +40% Düsseldorf demand spike, recomputing exposed reorder points and lateral-shipping from the central pool, then defend the plan to the board on the service-cost frontier. The math rewards variability-aware segmentation, differentiated service targets, demand-during-lead-time reorder points, central risk pooling, cash-neutral redeployment and joint network optimization — and punishes the five classic errors: uniform 98% targets, forward-deploying everything, average-demand reorder points, buying out of the problem, and single-stage silo thinking. Final KPIs track fill rate (%), inventory on hand (€M), inventory turns (×) and quarterly service P&L.