ABC analysis

ABC analysis is a classification method that sorts SKUs into three buckets — A, B, and C — based on their relative importance. The principle: not every SKU deserves the same planning attention. The vital few drive the result; the long tail just adds noise.

The classic split (rooted in the Pareto principle):

• A SKUs — top ~20% of items, usually ~70–80% of revenue
• B SKUs — middle ~30% of items, ~15–20% of revenue
• C SKUs — bottom ~50% of items, ~5–10% of revenue

The exact percentages vary by business. The pattern doesn't.

How to build the analysis

Three steps:

1. Pick the metric. Revenue, gross profit, units sold, or contribution margin. Different metrics produce different classifications — gross profit weights margin in addition to volume.
2. Sort SKUs descending by the metric.
3. Cumulate the percentages and cut. A = top items contributing the first 70–80% of total. B = items adding the next 15–20%. C = the long tail.

The cuts aren't sacred. Some businesses use 80/15/5; others 70/20/10. Pick the split that produces useful operational behaviour and stick with it.

What ABC actually changes operationally

Each tier gets different planning treatment:

The point isn't that C SKUs don't matter. It's that C SKUs don't deserve the same planning spend as A SKUs. The brand has limited planning attention; ABC tells you where to put it.

Why one-dimensional ABC is too simple

Pure revenue-based ABC misses important nuance:

• A high-revenue SKU at 5% margin might be less important than a mid-revenue SKU at 60% margin.
• A SKU with steady predictable demand needs less planning attention than a high-variability SKU even if revenue is similar.
• A SKU with a 2-week lead time can recover from a forecast miss; a 16-week lead time cannot.

Hence the variants:

• ABC by gross profit. Weights margin alongside volume.
• ABC-XYZ. Two-axis: ABC for value, XYZ for demand variability (X = stable, Y = variable, Z = erratic). AX = high-value stable, AZ = high-value erratic, etc. Each cell gets its own planning policy.
• Custom multi-criteria. Some businesses build a composite score from revenue, margin, lead time, and strategic importance.

For most scaling brands, plain ABC is good enough to start. Move to ABC-XYZ once the basics are working.

Refresh cadence

ABC isn't static. SKU performance shifts:

• Annual refresh is the minimum
• Quarterly refresh is best practice for fast-moving businesses
• Trigger-based refresh (e.g. SKU's 90-day rolling revenue jumps a tier) for very fast businesses

The bigger risk is not refreshing — yesterday's A SKU declines into a B but keeps getting A-tier planning attention while the new A SKU starves.

Common mistakes

• One-time setup, never refresh. SKUs migrate between tiers; the classification has to track that.
• Treating ABC as the answer rather than the input. ABC tells you where to spend planning attention; it doesn't tell you what the planning policy itself should be.
• Using ABC across drastically different categories together. A category with 100 SKUs has a different distribution than one with 10,000. Either run ABC per category or accept that the cuts will be unbalanced.
• Confusing ABC with GMROI. ABC ranks by absolute contribution; GMROI ranks by inventory productivity. Both useful, different questions.

ABC vs Pareto

ABC analysis is the operational application of the Pareto principle ("80/20 rule"). Pareto observed that ~80% of effects come from ~20% of causes; ABC turns that into a planning method by adding the middle bucket and prescribing different treatment per tier.

When ABC is the wrong tool

Two situations where the framework doesn't fit:

• Very small SKU counts. A brand with 30 SKUs doesn't have a long tail to cut off. Just plan each one.
• Highly correlated SKU performance. If demand for the SKUs moves together (a true category brand), classifying them differently doesn't add much. Treat them as a unit.

For most consumer brands with 100+ SKUs and varied demand patterns, ABC is the right starting framework.