Product segmentation
Product segmentation in ecommerce advertising groups products by shared performance or business characteristics so merchants can give each group a suitable priority.
By Berend Vrakking, founder of Product Metrics. Updated 28 October 2026.
Product Segmentation example
| Product group | Advertising clicks | Return |
|---|---|---|
| Stars | High | High |
| Question Marks | High | Medium |
| Dogs | High | Low |
These groups describe advertising volume and return against your chosen target, not sales volume. Illustrative data.
Put it into practice
- Explore it: Product SegmentationSee the six segments, rules and label workflow.
- Read the guide: Optimise Google Shopping adsApply product-level performance context to Shopping.
Product Segmentation: one product and the whole account
Product Metrics ML puts products into six segments using advertising click volume and return measured as ROAS or POAS. An account average hides this distribution; product groups show where to investigate performance.
The six segments are Stars, Question Marks, Cash Cows, Dogs, Drainers and Ghosts. Product Metrics writes approved segment labels to Merchant Center. Configured campaigns then use those labels; the merchant controls campaign structure and priorities.
Product Segmentation: common mistake
Confusing a segment with an instruction to change a price or budget automatically. A segment describes performance; merchant rules and campaign configuration determine how it is used.
Product Segmentation: questions
Do I need margin data to segment products?
Are segments the same as custom labels?
Keep reading
- POAS calculatorCompare contribution return before choosing priorities.
- Custom labelsThe fields that carry a product group into campaigns.
- ROASOne return measure used for segmentation.