ABC analysis is the inventory-classification practice of sorting a SKU catalog into three tiers — A, B, and C — by each SKU’s contribution to a chosen output metric, so operational attention scales with economic importance instead of being spread evenly across the catalog. Once the catalog outgrows the point where every SKU can get the same care, differentiating treatment beats spreading it thin.
The conventional split is Pareto-derived. A items are roughly the top 20% of SKUs and drive ~80% of the chosen metric; B items are the next ~30% and drive ~15%; C items are the long-tail ~50% that contribute the remaining ~5%. The cutpoints are operator conventions, not laws — 70/20/10 and 60/30/10 are common variants — and the asymmetry is the point, not the exact numbers.
The metric you classify on is itself a decision. ABC on revenue surfaces volume SKUs but obscures margin: a hero SKU at a thin gross margin can rank above a quieter SKU that prints twice the contribution dollars. ABC on gross-profit dollars — or contribution-margin dollars, where the brand has clean variable-cost data — ranks SKUs by the cash they actually generate, which is what most P&L-aware operators want.
The classification drives concrete operations cadences. A items get tight cycle counts (often weekly), safety-stock buffers, and primary vendor relationships. B items get monthly counts and standard replenishment. C items get quarterly or annual counts and become the sunset shortlist at assortment review, because they tie up inventory dollars out of proportion to the cash they return. The classification itself refreshes quarterly for DTC catalogs — annual is too slow once demand mix moves, monthly creates churn.
ABC is a classification framework, not a replenishment policy. Knowing a SKU is an A item tells you it deserves attention, not what reorder quantity to set; pair the tiers with safety-stock math and lead-time data to turn the classification into a stocking plan.