Turning Sooka's free tier into a paying audience — one experiment at a time.
Sooka launched with a dual-tier model: ad-supported free, VIP subscription. The free tier grew fast — but conversion stalled around 2%. I owned the funnel end-to-end: instrumented it, ran 20+ experiments across sign-up, onboarding, paywall and trial, and re-framed the conversation from “grow subs” to “grow the right cohort.”
Free tier grew fast but VIP conversion plateaued at 1.9%. No one knew which content hooks, which nudges, or which cohorts actually paid — the funnel was a single KPI, not a system.
Mapped the funnel from anon visit to VIP start. Built behavioral cohorts (sports, drama, organic, telco-bundled). Ran a rolling experiment backlog against every stage — paywall, trial length, onboarding, ad cap.
Conversion lifted to 2.8%. Sports cohort carried 3.1× the baseline LTV; telco-bundled users looked cheap on CAC but churned fastest (0.4×). Playbook by cohort replaced one-size campaigns.
From anon visit to VIP start.
Ten from the backlog. Six moved the needle.
Not all free users are the same.
Sports-acquired users retain at nearly 3× the rate of telco-bundled ones — even though bundle CAC was a fraction of paid. This reframed the acquisition mix for FY24 planning.
sooka OTT · Astro
Four things that compounded.
Copy + pricing surface matched to the content the user tried to watch. +14% on conversion.
Three-tap genre pick before first play cut time-to-first-play by 22 seconds.
Match-day paywalls captured intent at peak. +18.7% on EPL weekends.
14d trial for sports, 7d for drama. Shorter where D30 retention already held.