Metrics & analytics · Easy
How would you measure the success of a new 'save for later' feature in an e-commerce app?
Short answer
Start from the feature's goal: helping undecided shoppers come back and buy instead of abandoning. Track adoption (share of sessions where items are saved), engagement (return visits to the saved list) and outcome (purchases of saved items within a set window). Add guardrails so the feature does not cannibalise immediate purchases, and compare against a control group to prove causality.
Detailed answer
Framework: goal → adoption → engagement → outcome → guardrails
Work backwards from why the feature exists. Metrics without a stated goal are just a list.
Worked example
Goal hypothesis: many shoppers are not ready to buy (waiting for payday, comparing options), so they abandon the cart or forget the product. Saving items should bring some of them back.
Adoption
- Share of product-page sessions where an item is saved.
- Share of weekly active users who have saved at least one item.
Engagement
- Share of savers who revisit their saved list within a set period.
- Items moved from saved list to cart.
Outcome (primary)
- Conversion of saved items to purchases within a defined window, compared with a control group without the feature.
- Net change in purchases per user, because saved-item purchases might simply replace purchases that would have happened anyway.
Guardrails
- Immediate conversion on product pages: if users save instead of buying now, and then forget, the feature could hurt.
- Cart abandonment trends.
- App performance on product pages.
Use an A/B test at launch. Look at the result by segment: new versus returning users, and price bands, because the feature probably matters more for expensive items.
Common pitfalls
- Celebrating adoption (lots of saves) without checking purchases.
- Ignoring cannibalisation.
- Measuring for too short a window when the purchase decision takes days.