ABC-XYZ Classification

Classifies SKUs based on their contribution to total inventory cost or revenue, following the Pareto principle.

• A-Items: High-value products representing ~10-20% of SKUs but 70-80% of total inventory value. Require tight control, accurate forecasting, and frequent cycle counting.
• B-Items: Moderate-value products representing ~30% of SKUs and 15-20% of value. Managed with automated systems and periodic review.
• C-Items: Low-value products representing ~50% of SKUs but only 5-10% of value. Suitable for simplified, rule-based replenishment with higher safety stock.

Categorizes items by the stability and forecastability of their consumption patterns using the coefficient of variation (CV).

• X-Items: Highly stable demand with low variance (CV < 0.5). Suitable for zero-stock or just-in-time strategies with minimal buffer.
• Y-Items: Fluctuating demand with moderate variance (0.5 ≤ CV < 1.0). Often driven by trends or seasonality, requiring statistical forecasting and moderate safety stock.
• Z-Items: Highly erratic, lumpy, or intermittent demand (CV ≥ 1.0). Forecasts are unreliable; managed via make-to-order strategies or high safety stock.

The intersection of ABC and XYZ dimensions creates nine distinct segments, each requiring a tailored planning approach.

• AX / BX: High-value, predictable items. Ideal for Vendor-Managed Inventory (VMI) and tight JIT synchronization.
• AY / BY: High-to-moderate value with trending demand. Use statistical forecasting with dynamic safety stock calculation.
• AZ / BZ: High-to-moderate value with chaotic demand. Requires manual review, Available-to-Promise (ATP) checks, and make-to-order where possible.
• CX / CY: Low-value, predictable-to-trending items. Automate with simple reorder-point logic and high Economic Order Quantity (EOQ) batches.
• CZ: Low-value, unpredictable items. Minimize management overhead with generous safety stock or eliminate from the catalog if non-strategic.

The statistical foundation of the XYZ classification is the coefficient of variation (CV), a normalized measure of dispersion.

Formula: CV = (Standard Deviation of Demand / Mean Demand)

• Calculated over a representative historical period, typically 12-24 months.
• A CV of 0.2 indicates highly stable demand; a CV of 1.5 indicates extreme volatility.
• For intermittent demand (many zero-demand periods), use specialized methods like Croston's method or ADIDA (Aggregate-Disaggregate Intermittent Demand Approach) before calculating CV to avoid misleading classifications.

The core value of the matrix is applying the right planning intensity to the right items, avoiding over-engineering low-value SKUs.

• Forecasting: X-items use simple exponential smoothing; Z-items may require demand sensing or qualitative input.
• Replenishment: AX items use continuous review with low safety stock; CZ items use periodic review with high order-up-to levels.
• Inventory Audit: A-items undergo frequent cycle counting; C-items are audited annually.
• Supplier Strategy: AX items benefit from long-term contracts and tight collaboration; CZ items can use spot buying.

ABC-XYZ classification serves as a critical input parameter for Multi-Echelon Inventory Optimization (MEIO) engines.

• The classification determines the service level targets and holding cost penalties assigned to each SKU in the optimization model.
• AX items at a central warehouse might be assigned a 99.5% cycle service level, while CZ items at a retail node might be set to 90%.
• This segmentation prevents the MEIO solver from allocating excessive safety stock to low-value, unpredictable items at the expense of high-value, stable ones.