ABC analysis applies the Pareto principle to inventory: a small number of "A" items typically account for ~70–80% of annual consumption value, "B" items a moderate share, and many low-value "C" items the rest. Control policies, counting frequency, and safety stock are then tiered by class.
It lets planners concentrate tight control, frequent reorder-point review, and accurate forecasting on the items that drive cost, while applying simple rules to the long tail.
How to run it: for each SKU multiply annual usage (units) by unit cost to get annual consumption value, sort descending, and accumulate the percentage of total value. A common cut is A = the top ~80% of value (often 15 to 20% of items), B = the next ~15% (around 30% of items), C = the remaining ~5% of value spread over half the catalog. The cut points are a policy choice, not a law.
Policies then follow class. A items get cycle counts weekly or monthly, tight safety stock and a planner review of every reorder. C items get a quarterly count, a simple two-bin or min/max rule and a larger order quantity so nobody spends planner time on a bracket. Re-run the classification at least yearly, because a part that moved from C to A with a new program is where stockouts come from. See the inventory management guide.