Case Study: Moonpig boosts revenue per user by optimizing product recommendations with SiteSpect

Moonpig Increases Conversions By Optimizing Product Recommendations

Moonpig, a UK-based retailer of personalized cards and gifts, needed to surface the right products to the right customers to boost add-on sales. Their client-side A/B testing tool caused flicker, bucketing and accuracy issues, so the team sought a solution that would integrate with their ML-driven recommendations and let them reliably test and optimize product displays — specifically to improve the “attach” rate of recommended gifts shown after adding a card to cart.

Using SiteSpect’s server-side Engine API, Moonpig updated its recommendations algorithm, added metadata to enable meaningful categorization, and redesigned the recommendations into a horizontal carousel with items grouped by category. The change increased revenue per user by raising attachment rates without reducing purchase conversions, and the variation was rolled out to 100% of traffic.

Moonpig

James Huppler

Head of Product