ABC analysis sorts inventory into three tiers by how much value each item ties up, so you can spend tight control on the few items that matter and loose control on the many that don't. The A items, roughly the top 20% of SKUs, usually account for about 80% of the value; the C items, more than half the SKUs, account for a small slice.
Every storeroom has the same lie hiding in it: that all the parts deserve equal attention. They do not. A handful of high-value items drive most of your inventory dollars and most of your stockout risk, while a long tail of cheap parts sits there quietly. ABC analysis is the tool that separates the two so your counting, your reviews, and your ordering effort land where the money is. This post walks through classifying SKUs, setting a policy per tier, and where the method misleads you.
What is ABC analysis?
ABC analysis is an inventory classification method that applies the Pareto principle, the 80/20 rule, to your item catalog. The idea, named for the economist who noticed that a small share of causes drives most effects, is that inventory value is never spread evenly. A small fraction of items commands most of the dollars. By ranking items and cutting the list into three groups, you turn one undifferentiated pile of thousands of SKUs into three groups you can manage on purpose.
The typical split looks like this: A items are the vital few, often around 10 to 20% of SKUs but 70 to 80% of annual value. B items are the middle, perhaps 30% of SKUs and 15 to 20% of value. C items are the trivial many, often half or more of the SKUs but only 5 to 10% of value. The exact percentages vary by operation; the shape, a steep front and a long flat tail, is remarkably consistent. Some shops add a fourth D tier for dead or obsolete stock, and others split A into A1 and A2, but three tiers is the workhorse. The number of groups matters less than the discipline of treating them differently once you have them.
How do you classify SKUs into A, B, and C?
You classify by annual dollar usage: multiply each item's unit cost by its annual demand, rank the list from highest to lowest, then draw the tier lines where the cumulative value hits your chosen thresholds. Annual dollar usage, not unit price alone, is the standard criterion, because a cheap part used constantly can tie up more money and more risk than an expensive part used once a year.
Run it as a repeatable procedure, not a one-time spreadsheet exercise.
- Pick the value metric. Usually annual dollar usage: unit cost times annual usage quantity. Use a consistent time window, typically the trailing 12 months.
- Calculate it for every SKU. Compute annual dollar usage for each item so every part has a comparable number.
- Rank and accumulate. Sort descending, then compute each item's running share of total value.
- Draw the tier lines. Set thresholds, commonly the top ~80% of cumulative value as A, the next ~15% as B, the last ~5% as C, and label each item.
- Sanity-check against reality. Override the math where judgment demands it: a low-dollar part that shuts down a line if it stockouts belongs in A regardless of its dollar rank.
- Reclassify on a schedule. Demand and prices move, so rerun the analysis periodically, quarterly or at least yearly, so the tiers stay honest.
That last point matters more than it sounds. An ABC classification set once and never revisited slowly drifts out of date as new products launch and old ones fade, and you end up lavishing A-level attention on a part that became a C two years ago.
What control policy applies to each tier?
Each tier gets a different level of control, review frequency, and buffer, because the whole point is to stop treating a fastener like a flagship component. A items earn tight, hands-on management; C items get simple, low-touch rules that keep them available without eating your time.
| Policy lever | A items (vital few) | B items (middle) | C items (trivial many) |
|---|---|---|---|
| Review | Frequent, often continuous | Periodic | Infrequent, simple rules |
| Ordering | Tight lot sizes, close to demand | Moderate | Bulk orders, order-up-to levels |
| Safety stock | Lean but closely watched | Moderate buffer | Generous buffer, low risk to hold |
| Forecasting | Detailed, item-level | Standard | Simple or none |
| Cycle count | Most frequent | Medium frequency | Least frequent |
The logic behind the C column surprises people: for cheap items, carrying a bigger buffer is often smarter than spending management time trying to trim it. The cost of holding an extra box of washers is trivial next to the cost of a stockout or the labor to manage it tightly. Spend your scarce planning attention on the A items, where a percentage point of buffer is real money and a stockout is real disruption. The mistake most storerooms make is the reverse: they micromanage the cheap parts because they are numerous and visible, and coast on the expensive ones because there are only a few. ABC exists to break exactly that instinct.
How often should you count each tier?
Count A items most often, C items least, and let the tier drive the schedule. This is where ABC analysis marries cycle counting: instead of one disruptive wall-to-wall count a year, you count a slice of inventory every day, weighted so the high-value A items get counted several times a year while C items might get counted once. Typical cadences run something like A items quarterly or monthly, B items twice a year, and C items annually, though the right numbers depend on your accuracy and your risk. The effect is that your counting effort, like your control effort, concentrates on the items where an error costs the most.
What do the numbers say?
Context and definitions from standards bodies and primary data:
- ABC classification is a defined inventory-management practice in the body of knowledge maintained by the Association for Supply Chain Management (ASCM/APICS) which codifies the tiered-control approach in its APICS Dictionary.
- Inventory is a large, live number on the economy's books: the U.S. Census Bureau's Manufacturing and Trade Inventories and Sales series tracks business inventories in the trillions of dollars, with the inventories-to-sales ratio hovering in a roughly 1.3 to 1.4 range in recent years.
- The underlying 80/20 pattern is the Pareto principle, the observation that a small share of inputs drives most of the output, which ABC analysis operationalizes for a parts catalog.
The practical takeaway is that a small number of SKUs sits on top of most of that money, and ABC is the cheapest way to find them.
What are the limits of ABC analysis?
The biggest limit is that a single value criterion ignores everything else that matters. Annual dollar usage says nothing about lead time, criticality, or how erratic demand is. A cheap gasket that grounds a line and takes eight weeks to source is a C item by dollars and an A item by risk, and pure ABC will misfile it. Good practice overrides the math for critical and long-lead items, and many operations add a second dimension, classifying demand variability alongside value, so an item is scored on both how much it costs and how hard it is to predict. ABC is a starting lens, not the whole eye. Treat it as the first cut that tells you where to look, then layer in criticality and lead time before you set the final policy, the same way a good advanced planning and scheduling approach weighs more than one constraint at a time.
Where ABC analysis breaks in practice
The method is simple; keeping it current is the hard part. Classifications go stale, count frequencies drift back toward "whenever we get to it," and the analysis lives in a spreadsheet disconnected from the actual stock movements on the floor. When usage data is scattered across an ERP, a warehouse system, and a maintenance storeroom, nobody reruns the ranking, and the tiers quietly stop matching reality. Harmony is an AI-native layer that connects machines, software, and paperwork into one operational layer, with no rip-and-replace, so item usage, counts, and stock movements become one live record instead of three disconnected ones. AI search returns cited answers across those records, so a planner can ask which items drive the most value or which A items are overdue for a count and get a real answer, and Harmony's digital workflows route each count and reorder to the right person. It is not an inventory-optimization product; it keeps the classification honest by keeping the data in one place, the same paper-to-digital move Harmony makes on the floor (see the CLS case study), where it also sharpens inventory turnover and cuts the working capital trapped in the long tail. That long tail is exactly the kind of waste a lean operation works to squeeze out.