Back to all projects
Case study

Referral Mechanism for Kuku FM

A trust-built, fraud-resistant referral engine targeting 2,000 daily activated installs.

Role

Product Manager (case study)

Duration

4 weeks

Tools

FigmaCompetitive AnalysisFunnel Modelling
Referral Mechanism for Kuku FM
๐ŸŽง
2,000/day
verified activations target
< 5%
fraud-rate ceiling
15 min
behavioural reward gate
01Problem

Problem

Kuku FM's growth team needed a referral loop that delivered net-new listeners (not just farmed installs). Earlier referral attempts in the audio-content space suffered from heavy fraud โ€” fake numbers, SIM farms, and uninstall-after-reward behaviour.

02Users

Users

Tier-2/3 audio listeners aged 18-35 who consume devotional, romance, and self-help audio. They share heavily within close family/friends groups on WhatsApp โ€” making organic, social-graph referrals especially valuable.

03Hypothesis

Hypothesis

If we reward both referrer and referee with a meaningful, time-bound premium unlock (instead of cash), and gate rewards behind genuine listening behaviour, we can hit 2,000 verified activations/day while keeping fraud-rate < 5%.

04Wireframes

Wireframes

Designed: (1) Referrer hub with a personalised share card auto-generated per show; (2) Referee onboarding with social-proof ('your friend Riya gifted you 7 days Premium'); (3) Anti-fraud checkpoints โ€” device-fingerprint, SIM-carrier check, and 'minimum 15 minutes listened in 48 hours' before reward unlocks.

05Solution

Solution

Two-sided premium-unlock incentive with progressive milestones for the referrer ('refer 3 to get 1 month free'). Anti-fraud combines passive device/IP signals with an active 'listened for 15 min' behavioural gate. Rewards drip-released over 7 days to reduce one-shot abuse.

Learnings

What I'd take with me.

Behavioural gates outperform identity gates for fraud prevention. Asking the user to *do* something genuine (listen for 15 min) is harder to spoof at scale than asking them to *be* someone (KYC).