A/B’ing Trust Signals: Crypto Wallet Nova’s Three-Week Redesign Experiment

Design · 5 min read

Nova faced a classic crypto UX problem: users saw the wallet as risky and balked at granting permissions. The product team hypothesized that adding trust signals would increase adoption, but there was disagreement about which signals to display — partner logos, security audit badges, or granular explanations of permission scope.

They designed three variants and ran a three-week A/B test with new users: one showed partner logos (exchanges and integrations), another displayed security audit badges plus a short explanation of what was audited, and a third simplified the screen to one clear permission statement with a small “Learn more” link. Metrics tracked included permission grant rate, dropoff, and subsequent retention at day 7.

Surprisingly, the simplified permission-first design outperformed both credibility-heavy variants: conversion rose by 12% and day-7 retention by 8%. Heatmaps and session replays showed users were overwhelmed by dense badges and logos, often scrolling past important permission details. Follow-up interviews revealed users wanted one clear statement about what the wallet would and wouldn’t access.

Nova’s designers concluded that trust signals must be contextual and minimally intrusive; badges help later in the funnel (settings or help center), while the onboarding surface needs first-contact clarity. The outcome guided a new pattern library rule: present one primary trust cue and hide secondary signals behind progressive disclosure.