ABC analysis in e-commerce: classify products by sales, profit and stock
ABC analysis shows which products truly drive store performance. In e-commerce, it should be done not only by revenue, but also by profit, margin, turnover and visibility potential.
What is ABC analysis?
ABC analysis divides assortment into three groups. A products have the strongest impact, B products are important but less critical, and C products form the long tail. The classic version follows the Pareto idea: a small share of products generates a large share of sales.
In an online store, revenue alone is not enough. A product can be A by revenue but C by profit if margin is weak, discounts are frequent or returns are high. Run ABC in several versions.
Core ABC versions for online stores
| Version | What it ranks | When to use it |
|---|---|---|
| Revenue ABC | Products by sales share. | Turnover planning, campaigns and operational priorities. |
| Profit ABC | Products by gross profit. | Visibility, promotions and marketing budget decisions. |
| Inventory ABC | Products by inventory value. | Cash tied in stock control. |
| Conversion ABC | Products by traffic-to-purchase effectiveness. | Listing improvements and choosing products to show higher. |
How to define A, B and C groups
- Choose a metric: revenue, profit, margin, inventory value or units sold.
- Sort products descending by that metric.
- Calculate each product's share of the total.
- Calculate cumulative share from the top.
- Mark products up to about 70-80% cumulative result as A.
- Mark the next products up to about 90-95% as B.
- Mark the rest as C.
Thresholds are not law. Smaller catalogs may need different boundaries. The method should be repeatable and comparable month to month.
What to do with ABC groups
| Group | Meaning | E-commerce decisions |
|---|---|---|
| A | Highest impact on performance. | Protect availability, improve images and copy, keep visibility stable, discount carefully. |
| B | Potential or steady contribution. | Test visibility, seasonal promotions and product page improvements. |
| C | Long tail with low result share. | Automate, reduce manual work, consider discontinuation or niche exposure. |
The revenue ABC trap
If you only look at revenue, you may promote products that create turnover but not profit. Compare at least revenue ABC and profit ABC. A/A products are strategic. A/C products need caution because they sell, but may not build financial results.
| Revenue / profit | Meaning | Action |
|---|---|---|
| A / A | Strategic product. | Protect availability and visibility. |
| A / C | High turnover, weak profit. | Check price, discounts, cost and returns. |
| C / A | Low turnover, high profit. | Test better exposure or campaigns. |
| C / C | Low impact. | Automate or reduce assortment. |
ABC and XYZ: sales impact versus predictability
ABC shows how much a product matters. XYZ shows how predictable sales are. AX is important and stable. AZ is important but irregular. CZ is low-impact and unpredictable.
This combination is useful for purchasing and inventory planning. AX products need high availability, while CZ products should not tie too much cash.
Limitations of ABC analysis
ABC is useful, but it is not a complete assortment strategy. Revenue-based classification can ignore margin, seasonality, new products, strategic products, complementary items and basket impact. Do not treat ABC as an automatic deletion rule.
- New products may fall into C only because they have not had time to sell.
- Seasonal products may look weak off-season and strong during demand peaks.
- Accessories may have low revenue but high margin and strong AOV impact.
- C products may be needed for assortment completeness or SEO long tail.
- A products may be risky if they depend on one supplier or unstable stock.
Using ABC for purchasing and inventory planning
ABC creates the most value when classification changes operations. A products need better forecasts, more frequent stock checks and stronger supplier relationships. C products need simpler rules and less manual attention.
| Group | Stock | Purchasing | Control |
|---|---|---|---|
| A | High availability, controlled safety stock. | Supplier priority, negotiations, backup suppliers. | Daily or several times per week. |
| B | Stock matched to trend and season. | Regular replenishment, watch for movement to A or C. | Weekly or each buying cycle. |
| C | Minimal stock or sell-through. | Cautious reorders only when strategically justified. | Monthly or during assortment review. |
ABC at variant level
One e-commerce product often has many variants. Model-level analysis can hide the issue: black size M may be A, while green size XS sits in stock. If variants have separate stock, price or availability, run mini ABC at variant SKU level.
You do not need to analyze every variant immediately. Start with categories that have high stock, frequent stockouts or large differences between variants. Detail matters most there.
ABC as a merchandising tool
A products by profit and conversion often deserve strong exposure. A by revenue but C by profit requires caution: it may dominate listing while reducing profitability. C by revenue but A by margin is a good candidate for visibility testing.
ABC then stops being only an inventory report. It becomes a decision tool for product order, campaigns, newsletters and recommendations.
How often should you run ABC analysis?
- monthly for stores with regular sales,
- weekly during season or large campaigns,
- separately for seasonal categories,
- after major price, supplier or cost changes,
- after promotions that may change the sales mix.
CSV data needed
Basic analysis needs SKU, name, category, units sold, revenue, cost of goods and stock level. A stronger version adds views, conversion, returns, first sale date and inventory value.
Make sure every row represents the same product level: either SKU or variant. Mixing products and variants distorts the result.
Turn your catalog into a priority list
Upload a CSV file to Insighteo App and analyze products by sales, profit, conversion and visibility potential.
Create an account and upload CSV