A/B Testing the Cart: UX Case Study from a Grocery Delivery Startup

Tech · 6 min read

The original cart attempted to upsell with recommended items, estimated savings, and perks all above the fold. Analytics showed high interaction time but a 14% cart abandonment rate during checkout. The hypothesis was that cognitive overload and perceived price complexity were the culprits.

Designers created two alternatives: a streamlined cart that deferred upsells until post-checkout, and a contextual upsell experience that showed a single, highly relevant suggestion with a clear benefit. The company split traffic evenly across the three experiences and tracked conversion, average order value (AOV), and time-to-purchase.

Results were nuanced: the streamlined cart reduced abandonment by 9% and increased conversion, but AOV dipped slightly. The contextual upsell preserved conversion while increasing AOV by 6%. The final decision combined both: a streamlined primary cart for most users and a subtle contextual upsell for segmented cohorts, implemented through server-side flags to iterate quickly.

The case reinforced a lesson for startups—measure both immediate conversion and long-term revenue impacts, and favor server-side feature flags to pivot without lengthy releases.