Why Inventory Classification Matters
Not all inventory items deserve the same planning attention. Treating a $50,000/year component the same as a $200/year consumable wastes your planners' time and produces worse outcomes for both. Inventory classification is how you formalize that distinction — and apply consistent, appropriate planning policies to each tier.
ABC classification segments items by annual dollar value. XYZ classification segments by demand variability. Together, they give you a nine-cell matrix that tells you exactly how to plan each item in your portfolio.
ABC Classification: Segmenting by Dollar Value
ABC classification divides your inventory into three tiers based on annual dollar usage (quantity consumed × unit cost):
- A-items — the top 80% of annual dollar spend, typically representing 10-20% of your SKU count. These deserve your most planning attention.
- B-items — the next 15% of annual dollar spend. Moderate value, moderate attention.
- C-items — the bottom 5% of annual dollar spend, typically representing 50-70% of your SKU count. Low value, minimal planning effort required.
How to calculate ABC classification
- Pull 12 months of demand history for all active items
- Multiply quantity consumed by unit cost to get annual dollar usage per item
- Sort items from highest to lowest annual dollar usage
- Calculate cumulative percentage of total spend down the list
- Items in the top 80% cumulative = A. Next 15% = B. Bottom 5% = C.
XYZ Classification: Segmenting by Demand Variability
XYZ classification segments items by how predictable their demand is — measured by the Coefficient of Variation (CoV):
- X-items — CoV ≤ 0.5. Stable, predictable demand. Statistical safety stock and MRP work well.
- Y-items — CoV 0.5-1.0. Variable or seasonal demand. Adjusted safety stock with seasonal awareness.
- Z-items — CoV > 1.0. Erratic, unpredictable demand. Do NOT use statistical safety stock. Use min/max buffer logic and manual planner judgment.
The critical insight: you cannot use the same planning logic for a stable X-item and an erratic Z-item. The statistical safety stock formula assumes normally distributed demand. Z-items violate that assumption — applying the formula gives you a number that sounds precise but is actually meaningless.
The 9-Cell Planning Matrix
Combined ABC/XYZ classification produces nine planning classes. The most important ones to know:
- AX — high value, stable. Tightest controls, weekly review, statistical safety stock at 97-99% service level.
- AZ — high value, erratic. Your most dangerous combination. Weekly manual review, buffer logic only — never statistical safety stock.
- CX — low value, stable. Simple min/max, quarterly review, minimal planner attention.
- CZ — low value, erratic. Fixed buffer quantity, exception-only attention.
The Most Common Classification Problems
- Classifications set at ERP go-live and never updated. Demand changed, costs shifted, products were added and discontinued — but the A/B/C codes stayed where someone put them years ago.
- XYZ classification never implemented. Many ERP systems support XYZ fields but they're rarely populated. Without XYZ, every item gets planned the same way regardless of demand volatility.
- Z-items running on statistical safety stock. The formula produces a number, the number goes in the system — even though the formula doesn't apply to erratic demand.
In a typical SMB ERP, 30-50% of active items are misclassified when checked against current demand data. That means 30-50% of your inventory is running on the wrong planning policy.
How Often Should You Reclassify?
At minimum, annually — run the analysis against the last 12 months of data and update classifications that have shifted. Also reclassify after significant business changes: new customers, lost customers, product line additions or discontinuations, major cost changes. Document the review process in an SOP with a named owner and a scheduled review date.