Catalogue campaigns covered the top of funnel; abandoned-cart retargeting brought back 22% of users.
The task
By the start, the advertiser had confirmed demand but no controllable volume. The task was to clean up the top of the funnel for an e-commerce store in Spain and bring back users who had abandoned their carts.
We worked the "E-commerce" vertical in the Spain market. Payment model — spend managed to a target ROI, main sources — catalog campaigns, retargeting.
How we did it
The setup was built sequentially, testing hypotheses on small budgets.
- Cleaned up the product feed for catalogue campaigns: removed out-of-stock items and fixed price mismatches between the feed and the live site.
- Split retargeting into two windows instead of one: users who abandoned a cart in the last 24 hours saw a simple reminder, while the 3–7 day window got a small promo code.
- Didn't merge the two groups into a single retargeting audience, as had been done before.
What went wrong
The main difficulty surfaced in the second week. Previously both groups of cart-abandoning users were merged into a single retargeting audience — this diluted the discount effect on users who didn't need it to convert.
Split retargeting into two windows: a simple reminder for the 24-hour window and a promo code for the 3–7 day window. This brought back 22% of cart-abandoning users, most of them from the 24-hour reminder alone.
The result
Here are the final numbers for the buying period. Over the 2025: roi — 260%, cart recovery — 22%, retargeting windows — 2.
Final ROI was 260%; the discounted 3–7 day segment converted worse but at a higher margin, which mattered more for overall ROI than the raw recovery rate.
What we took away
- A clean product feed (no stock or price mismatches) reduces wasted impressions on its own, before any bid changes.
- Retargeting should be split into windows by how long ago the cart was abandoned — a single audience dilutes the discount effect.
These numbers belong to one specific setup, geo, and buying period. Economics will differ on another offer or in another season — we always recalculate them before launch.