Cleo Referral Home Overhaul
Redesigned Cleo's referral home screen around two psychological levers, social proof and a gamified progress tracker - to break a multi-quarter plateau in share rate. Led end-to-end: discovery, test design, and full rollout, while Product managing the project.
Overview
Company
Cleo
Date
1 month (Discovery → launch)
Role
Senior Product Designer
Team
1 iOS Engineers, Backend Engineer, AX Lead
Impact
• +7% share rate across 71,634 users (T1)
• +2.43% clicked share, winning variant vs control
• Rolled out 100%, 24 Feb 2026
Discovery & insights
I ran workshops to map where users hesitated before referring a friend, and analysed customer support transcripts to find the real questions people were asking. The pattern was clear and specific: the dominant friction wasn't reward size or general trust in Cleo.
There was uncertainty about whether a referral would actually work for the friend being invited. That insight reframed the brief from "improve the referral screen" to "remove the doubt that's stopping the share."
The CS analysis also surfaced a recurring set of questions that became the basis for a new FAQ section, built directly from real user language rather than guesswork.
The design
Two levers, designed to be tested both in isolation and combined:
Social proof — contextualised proof points placed directly at the moment of decision, showing real, successful referrals rather than generic reassurance copy.
A gamified boost tracker — reframing the reward from a flat, one-off incentive into a visible, trackable progression. The CTA was rebuilt around tracking an invite rather than a generic share prompt, applying the same structural mechanic that makes game progress bars compelling to a referral reward.
I structured the test as a clean three-arm split — control, social proof alone, and both levers combined — to isolate which lever was doing the work.
