What 115,000 Apps Tell Us About Subscription Monetization

An analysis of 115,000+ apps generating $16 billion in subscription revenue produces a picture that contradicts several common assumptions about how subscription apps grow. The market is not graduating from freemium to paid. It is polarizing: apps that figured out monetization are pulling away, and those that haven't are shrinking. The middle ground is disappearing.

The polarization problem

The top quartile of subscription apps grew MRR by more than 80% year-over-year. The bottom quartile shrank by 33%. The distance between these two groups is not explained by category, platform, or team size — it is explained almost entirely by monetization strategy. Apps in the top quartile made deliberate choices about paywall timing, trial structure, and pricing. Apps in the bottom quartile largely inherited defaults.

Hard paywalls outperform freemium by a wide margin

At day 35, hard paywalls convert at 10.7%. Freemium converts at 2.1%. By day 60, hard paywalls generate 8× more revenue per install than freemium.

This surprises teams that build freemium products on the assumption that a larger free audience eventually produces more paid subscribers. The data does not support that assumption. Freemium grows install counts; it does not reliably grow revenue. The users who enter a hard paywall app have already decided to engage with it — the install itself represents stronger intent.

• Hard paywall: user sees the paywall before meaningful use — conversion is front-loaded
• Soft paywall: user gets limited access before hitting a paywall — lower conversion, but more time to demonstrate value
• Freemium: user can use a meaningful portion of the app for free indefinitely — lowest conversion rate

The right model depends on the app's value proposition. Apps where value is immediately obvious — a specific tool that does one thing well — convert well with hard paywalls. Apps that require time to demonstrate value (habit trackers, language learning) may need soft paywalls or trials.

55% of trial cancellations happen on day zero

More than half of trial cancellations occur on the first day — before the user has had a chance to experience the full product. This is almost never a product quality problem. It is an onboarding problem: the user did not immediately see why the app is worth keeping.

The practical implication is that the first session is not a product demo — it is the conversion event. Features that users discover on day three do not affect trial-to-paid conversion because the majority of cancellations have already happened. Teams that optimize onboarding ahead of feature development consistently see better conversion than teams that ship more features first.

The Android billing failure gap

31% of Google Play subscription cancellations are billing failures — the card declined, the payment method expired, or the payment processor returned an error. On the App Store, the equivalent figure is roughly half that.

Billing failures are recoverable revenue. A user whose card declined is not a churned user — they are a subscriber who encountered a friction point. Google Play provides grace periods and account-hold states specifically to allow recovery. Apps that handle these states — surfacing in-app messaging when a payment fails, sending push notifications with a direct link to update payment details — recover a meaningful share of this otherwise-lost revenue.

• Implement Google Play's grace period — continue providing access for a few days while the payment is retried
• Use account hold state — clearly communicate to the user that their subscription is paused, not cancelled
• Send a direct deep link to the Play Store subscription management screen, not a general prompt
• Time the recovery message — the first 24 hours after a billing failure have the highest recovery rate

What the data recommends

The patterns in the data translate into a short list of decisions that account for most of the gap between high-performing and low-performing subscription apps:

• Test hard paywall timing before assuming freemium is the right model — the conversion difference is large enough to justify A/B testing early
• Treat day-one onboarding as the highest-leverage product surface — it determines whether 55% of trials stay or cancel
• Build billing failure recovery for Android — 31% of cancellations are recoverable, not true churn
• Price by market — the same subscription priced identically across all regions will underperform in markets where purchasing power differs significantly
• Measure retention by cohort, not overall — overall retention numbers mask which acquisition channels produce paying users vs. free users