Marketing A/B Test: Resolving Mobile Checkout Friction

This project evaluates a critical A/B test aimed at solving a significant drop-off in conversion rates during the checkout phase, specifically for mobile users. Using a sample size of 40,000 users and rigorous statistical analysis (Two-sample Z-test & Logistic Regression), the goal was to determine if a new UI design could increase conversions without hurting existing user habits.

Key Findings & Business Impact:

Massive Conversion Uplift: The new UI design was a huge success, driving a +65.4% relative uplift in overall conversions. The conversion rate jumped from 11.36% to 18.79%, strongly supporting a 100% product rollout.

The Mobile Friction is Fixed: The primary objective was successfully met. The mobile segment was the true hero of this test, with conversions nearly doubling (+85.5% relative uplift, jumping from 9.98% to 18.52%).

Universal Improvement: The new design didn't just fix mobile; it improved the experience everywhere. Desktop users also saw a strong ~41% relative uplift, proving the new UI is universally better across all devices.

Zero Risk & No "Novelty Effect": The data proved that returning users adapted positively to the change. There was no novelty effect detected, ensuring that the conversion boost is sustainable and not just a temporary spike.