Skip to content
GetProfit
Log in

← All terms

ABC analysis

ABC analysis is a way to split a catalogue into three classes by revenue share: A for the few products that bring most of it, C for the many that bring little.

How it works

You sort products by revenue, largest first, and add up their shares of total revenue down the list. The products that make up the first large slice become A, the next slice B, the rest C. The method comes from inventory management, where it ranks items by the value of stock used in a year. The cut-offs are a convention with no single standard, and the idea is close to the Pareto principle.

For ads, you run it on the revenue each product brought from advertising over a set period. To use the classes in campaigns, you write them into a custom label in the feed and split campaigns or listing groups by it. ABC analysis for an online store shows how to run it on a product export.

Formula

Cumulative share = revenue of the product and every product above it in the list ÷ total revenue

Example

Example store, not client data.

The tableware shop has 3,000 products and 180,000 of revenue from ads in a month; 220 products sold at least once. With cut-offs at 80% and 95%, the first 100 products by revenue bring 144,000 and become A. The next 70 bring 27,000 more, up to 171,000: class B. The other 2,830 share 9,000 and become C: 50 sold something and 2,780 sold nothing, 1,800 of them never shown.

Not to be confused with

  • Labels — the portal’s eight groups, set by how a product behaved in ads over 12 months: sales together with spend and ROAS. ABC looks at revenue alone.
  • XYZ analysis — splits products by how steady their demand is, not by how much they bring. The two are often combined as ABC-XYZ.

Right and wrong readings

  • Wrong: “Class C is what we switch off in ads.” Right: C lumps together products that spent budget without a sale and products that were never shown, since both have zero revenue. In our study of 1.4 million products, 37.1% of products spent budget without sales and 55.0% got no budget at all.
  • Wrong: “Our A products are safe for the whole year.” Right: a class describes one period. In the same study, 64.7% of winners were one-offs: one conversion in one month.

Sources

  • ABC analysis — Wikipedia: inventory origin, annual consumption value, cumulative shares, varying cut-offs. Checked 2 October 2026.
  • Listing groups — Google Ads API: dividing products by custom labels. Checked 2 October 2026.