A simple classification that decides where to spend counting effort, working capital and management attention.
ABC analysis sorts stock items by annual consumption value and splits them into three classes. Typically around 20% of items account for roughly 80% of value — these are A items. The next band is B, and a long tail of low-value items is C. The classification is not an end in itself: it exists to let you apply different counting, stocking and approval policies to each class.
Treating every SKU with equal care is expensive and, in practice, impossible. Someone will end up deciding informally which items matter, and that decision will follow familiarity rather than value. ABC makes the decision explicit and reviewable, so counting effort and working capital follow the money rather than habit.
Annual usage quantity multiplied by unit cost, per item. Usage, not stock holding — a high-value item that never moves is a different problem.
Sort descending and compute the cumulative percentage of total value.
Commonly A to 80% of value, B to 95%, C the remainder. The thresholds are a convention, not a rule — adjust to your distribution.
Count frequency, safety stock, review cycle and approval limits all differ by class. This is where the analysis becomes useful.
Quarterly or half-yearly. Items migrate between classes as demand changes, and a stale classification misdirects effort.
A distributor ranks 3,000 SKUs by annual usage value. The top 290 items — under ten percent of the catalogue — account for just over seventy percent of value. The next 580 account for a further twenty percent. The remaining 2,130 items, seventy-one percent of the catalogue, account for about nine percent.
Two decisions follow immediately. Counting effort should concentrate on the 290, because an error there moves the balance sheet and an error in the long tail does not. And those 290 should be slotted nearest dispatch, because they are picked most often — which is a different benefit from the same classification.
The third decision is the one usually missed. The 2,130 C items are nine percent of value but they are most of the catalogue, and therefore most of the clutter, most of the locations and most of the obsolescence risk. ABC is as useful for deciding what to stop stocking as for deciding what to watch. A C item that has not moved in eighteen months is not a small asset; it is a location, a count and a write-off waiting to be recognised.
| Term | What it means |
|---|---|
| Pareto principle | The observation that a small share of items accounts for most of the value. ABC is simply that applied to stock. |
| Annual usage value | Unit cost multiplied by annual movement. The usual basis for classification, and better than either input alone. |
| XYZ analysis | Classification by demand variability rather than value. Powerful combined with ABC — an AZ item is high value and unpredictable. |
| Reclassification | Rerunning the analysis as demand shifts. Items migrate; a classification set once is a snapshot of a year that has ended. |
| Slotting | Placing stock by class so fast movers sit nearest dispatch. The most direct way ABC turns into saved time. |
A cheap item consumed in enormous volume is an A item. Consumption value is the measure, not unit cost.
A classification from implementation day is managing last year's business.
If A and C items get the same treatment, the analysis has produced a report rather than a change.
A cheap C-class component that halts the line when it runs out needs A-class attention. Overlay criticality on top of value.
| Metric | What it tells you | What distorts it |
|---|---|---|
| Share of value in class A | Whether your curve is steep or flat. A flat curve means ABC will buy you less than the textbook suggests. | Using unit cost alone rather than usage value, which misclassifies expensive slow movers. |
| Class migration rate | How many items change class at each rerun. High migration means you should reclassify more often. | Reclassifying so often that slotting never settles. |
| Pick density by class | Whether A items are genuinely placed where they are picked most. | Measuring lines rather than visits; one line can be a full pallet. |
| Obsolete value in class C | Where dead stock accumulates, and the case for range rationalisation. | Valuing at cost rather than at realisable value, which overstates what is actually there. |
We build and support the systems this page describes. Start with a free working proof of concept — no cost, no commitment.