What is ABC and FSN Analysis?

ABC analysis and FSN analysis are two complementary ways of classifying inventory. ABC ranks items by their contribution to value; FSN ranks them by how fast they move — fast, slow, or non-moving. Used together they answer a question no single measure can: where should stocking effort and working capital actually go?

Why these classifications matter

A catalogue of several thousand items cannot be managed with equal attention. Reviewing every line weekly is impossible; reviewing none is negligent. Classification makes the compromise explicit — tight control where the money is, light control where it is not.

The two dimensions are genuinely different, which is why one is not a substitute for the other. A high-value item can be slow-moving, and a cheap item can be the fastest thing in the warehouse. Managing on value alone leads to careful control of expensive stock that sits; managing on movement alone leads to careful control of cheap stock while capital is tied up elsewhere.

How the classifications work

ABC applies the Pareto principle to consumption value — unit cost multiplied by quantity used, not price alone:

A — roughly the top 20% of items, typically 70–80% of value. Tight control, frequent review, accurate forecasts.
B — the next 30%, about 15–25% of value. Moderate control.
C — the remaining 50%, often under 5% of value. Simple rules, larger buffers, minimal attention.

FSN classifies by movement over a defined window:

Fast-moving — consumed regularly; availability is the priority.
Slow-moving — consumed occasionally; the risk is over-stocking.
Non-moving — no consumption in the window; the question is disposal, not replenishment.

Crossing them gives a nine-cell grid. The interesting cells are the corners: A-items that are fast-moving deserve the most attention in the business, while A-items that are non-moving are capital sitting still and are usually the single largest recoverable sum in a warehouse.

An example

A distributor holds 2,400 SKUs. ABC finds 380 items making up 76% of consumption value. FSN finds 210 items with no movement in six months. The overlap — 24 high-value, non-moving items — represents a disproportionate share of dead capital, and it is a specific, actionable list rather than a general instruction to reduce stock.

Common variations

VED analysis. Vital, essential, desirable — classification by criticality rather than value, used where a stock-out stops production or service.
XYZ analysis. Classification by demand variability, which pairs naturally with ABC because it decides where safety stock is genuinely needed.
FSN by value vs. by frequency. "Fast" can mean high volume or frequent picking; the two produce different lists.

Limitations worth stating

Both classifications look backwards. A new product has no history and will classify as non-moving until it sells, which is precisely the wrong signal during a launch. Seasonal items misclassify out of season. Classifications also drift, so a class assigned last year and never revisited slowly stops describing the item — and because the label looks like data, it tends to be trusted long after it stopped being true.

How xMatix supports ABC and FSN

ABC and FSN categories are held as attributes on the item in xMatix Inventory, and they are consumed directly by the replenishment engines in Procurement: a suggestion run can include or exclude items by ABC class and by fast, slow or non-moving status, so different classes can be replenished under different policies without custom code. Cycle-count rules can be scoped by class as well, which is the usual way to count A-items more often than C-items.

Reporting covers the analysis directly — seeded reports include an FSN and ABC movement analysis pivoted by branch, an overstock report, a stock-reaching-minimum-level report, and stock ageing with a dead-stock bucket; the inventory dashboard carries FSN and ABC breakdowns alongside ageing.

One implementation note, stated plainly: ABC and FSN categories are held as item attributes, so the classification is as current as the process that maintains it. Where classifications drive replenishment policy, that refresh cadence is worth owning deliberately — and for items whose demand is worth modelling directly, statistical and machine-learning forecasting work from observed history rather than from a class label.