Multi Criteria Inventory Classification
Multi-Criteria Inventory Classification (MCIC) is an inventory stratification method that ranks SKUs using more than one criterion—commonly demand, unit price, lead time, criticality, variability, obsolescence, and substitutability—and assigns classes such as A/B/C for differentiated control policies. Unlike classic ABC, it combines financial and operational risk factors into a weighted score to guide inventory management.
On a manufacturing shop floor, MCIC links each SKU's class to a specific control policy. Fast-moving or critical parts receive tighter review cycles, higher safety stock, and expedited replenishment, while stable low-priority items use simpler order rules and lower service targets. In warehousing, classification drives slotting, cycle counting, receiving attention, and replenishment expediting, so scarce labor focuses on items with the highest combined operational and financial impact. The process often begins with demand characterization using mean and standard deviation, then combines unit price and lead time into a weighted score, ranks SKUs, and applies ABC thresholds. A five-step framework is common: data collection, statistical characterization, item classification, control-system selection, and performance assessment with sensitivity analysis. This ensures that a critical low-dollar component is not starved just because its annual spend is small, and that high-variability long-lead items get the attention they need.
Why not use classic ABC alone?
Classic ABC ranks mostly by consumption value, but MCIC incorporates demand variability, lead time, criticality, and obsolescence, which better reflect operational risk in manufacturing and spare-parts environments.
What criteria are most common in manufacturing MCIC?
Demand level, demand stability or variability, unit cost, lead time, criticality, stockout risk, obsolescence, durability, substitutability, and repairability are repeatedly cited in the literature.
How are weights chosen in MCIC?
Published methods use decision-maker input, AHP/FAHP, fuzzy Delphi filtering, optimization-based weights, or utility functions to reduce subjectivity and convert multiple criteria into a single ranking score.