Sub Club by RevenueCat Dynamic Paywalls That Drove Millions in New Revenue – Shawn Gong, Tinder
9 snips
Mar 4, 2026 Shawn Gong, a Tinder product leader in growth and monetization, explains ML-driven paywalls that boosted millions in revenue. He discusses decision overload, using machine learning to surface the single best offer, smart a la carte pricing to avoid cannibalizing subscriptions, and why Tinder Select failed to scale. Short, practical takes on pricing, experimentation, and designing for emotional choices.
AI Snips
Chapters
Transcript
Episode notes
Decision Overload Kills Paywall Conversion
- Decision overload reduces conversion on paywalls and leads users to buy suboptimal tiers or buy nothing.
- Tinder observed many users bought Platinum out of price-based heuristics while only using Gold features, revealing overwhelmed choices harm outcomes.
Surface One Best Offer With ML
- Use ML to predict and surface the single best product for each user instead of showing every tier to reduce choice friction.
- Tinder moved from static pricing to dynamic pricing that asks an ML model which SKU to recommend for each user.
A-B Test ML Paywalls Incrementally
- Validate ML-driven paywalls with A/B tests against the control paywall and measure conversion plus revenue uplift.
- Tinder started small on a subset of features, routed paywall rendering to the model, and compared conversion and total revenue.