App Subscription Benchmarks: Trials, Churn and Pricing

Subscription apps are judged against numbers most founders have only seen in screenshots, rarely with the dataset or the year attached. This article collects the public benchmarks from RevenueCat's last three State of Subscription Apps reports and explains what each one actually measures.

Where the numbers come from

RevenueCat publishes its report each spring from apps using its SDK. The 2024 edition covered more than 30,000 apps and roughly $6.7 billion in tracked revenue ⁴; the 2025 edition grew to 75,000 apps and over $10 billion ³; the 2026 edition, published in March 2026 with data from 2025, covers more than 115,000 apps, $16 billion in revenue and over a billion transactions ^1,2. That scale is why the report is quoted everywhere, but the sample skews towards consumer mobile apps using a third-party SDK, and newer editions report medians where the 2024 edition sometimes used averages, so cross-edition comparisons need care.

Trials: who starts and who converts

Trials are front-loaded: 82% of trial starts happened on install day in the 2025 data ³, and Business apps reached nearly 90% same-day starts in 2026 ¹. Cancellations are front-loaded too: for 3-day trials, 55.4% of all cancellations occur on day 0, falling to 39.8% for 7-day, 35.7% for 14-day and 31.1% for 30-day trials; 84% of 3-day-trial cancellations happen within the first two days ^1,2.

Longer trials convert better in every edition, and 17–32-day trials were the best group in 2025 as well, at 45.7% ³. Yet 46.5% of apps used trials of four days or less in the 2026 data, up from 42.1% ². Short trials are not a mistake; they trade conversion for a faster signal and lower refund exposure.

Category and region move the median as much as length does. Travel led the 2026 trial-to-paid table at 43.5%, Health & Fitness came in at 37.7%, Gaming at 25.0% and Photo & Video at 22.2%; by region, North America converted at 34.2%, Asia-Pacific at 31.9%, Western Europe at 29.7% and India/South-East Asia at 15.2% ¹. The 2024 report, which used averages, put the overall trial conversion at 37.3%, with the App Store at 39.0% and Google Play at 28.5% ⁴.

Download-to-paid: the number that matters for ads

Trial-to-paid flatters apps with hard paywalls, because only people who already committed to a trial are counted. The fairer funnel metric is download-to-paid, measured at day 35. The 2026 global median is 2.0%, with Health & Fitness at 2.9%, Business at 2.6% and Gaming at 1.0% ¹. Hard-paywall apps reach 10.7% against 2.1% for freemium ^1,2; the 2025 figures were 12.1% and 2.2% ³, and the 2024 report had an overall 1.7% within 30 days, with a lower quartile of 0.6% and an upper quartile of 4.2% ⁴.

In dollars, the 2026 median revenue per install at day 60 is $0.34 overall, $0.55 in North America and $0.11 in India/South-East Asia, while hard-paywall apps earn $3.09 per install at day 60 against $0.38 for freemium ^1,2.

Retention at one month and twelve months

Year-one retention is where most expectations are wrong. The 2025 report put median retention after twelve months at 44.1% for yearly plans, 17.0% for monthly plans and 3.4% for weekly plans, all down from the prior period (47.1%, 18.8% and 4.2%) ³. The 2026 edition reports that roughly 72% of annual subscribers cancel within the first year, up from about 56% a year earlier, and that year-one retention of yearly subscribers is now similar for freemium (28%) and hard-paywall (27%) apps ². The 2024 edition's expert commentary gave monthly year-one retention of 11.4% ⁴. Holding on to subscribers has become harder each year.

The first month is the cliff. About 30% of annual subscriptions cancelled in month one in the 2025 data, and 35% of all annual cancellations fell in month one in the 2026 data ^2,3. First-renewal rates reinforce this: 61.7% renewed once and 53.4% twice in 2025, both lower than the year before ³. Not all of this is a choice. Billing errors caused 31% of Google Play cancellations and 14% of App Store cancellations in the 2026 data ²; Apple will retry a failed renewal for up to 60 days and lets you add a Billing Grace Period of 3, 16 or 28 days so paid-service days keep accruing ⁵.

AI apps are the outlier. They earn 41% more per paying user in year one ($30.16 versus $21.37) but churn about 30% faster, with 12-month retention on monthly plans 36% worse than traditional apps ^1,2.

Prices, plan mix and regions

Median prices barely moved: $5.99 weekly and $10 monthly in 2026, with the yearly median rising from $31.60 to $34.80 ¹. In 2024 the medians were $4.99, $9.99 and roughly $29.99 ⁴. Regional gaps are large. North American apps charge a median $39.99 a year and $9.99 a month; in India and South-East Asia the medians are about $18.32 and $3.75 ¹. Realised year-one value per payer follows the same curve: $23 globally, $32 in North America, $25 in Western Europe, $14 in India/South-East Asia ¹.

Annual plans are losing share. Monthly plans made up 42% of subscriptions and yearly 34% in 2026, and one contributor noted annual fell from 41.4% to 33.6% of plan durations year on year ¹. Productivity apps earn 91% of revenue from monthly plans; Health & Fitness earns 59% from annual; Gaming was 78% weekly in 2025 ^1,3. The annual-versus-monthly choice is a retention bet: annual plans retain better and reach Apple's 15% year-two rate sooner ⁵; monthly plans are easier to reprice. Note that on Apple, a price rise of more than 50% that exceeds about $5 per period (or $50 a year) requires active subscriber consent, as does any rise in regions that mandate it ⁶; on Google Play, subscriptions sit at a 10% service fee plus a 5% billing fee in the US, UK and EEA since 30 June 2026 ⁷.

How this shows up in the charts

Top Grossing is a retention chart in disguise. An app with 44% annual retention keeps collecting renewals from a cohort for years, so its store-billed revenue stacks even when installs flatten; the 2026 report found that apps launched before 2020 still generate 69% of subscription revenue and apps launched in 2025 or later just 3% ¹. That is why the top of the grossing list moves so slowly compared with Top Free, and why AI apps can rocket in on strong monetisation and then fade: higher revenue per payer gets them in, faster churn takes them out. A new subscription app holding a grossing position for several months is evidence of real month-two and month-three retention, not a launch spike.

Common mistakes when reading benchmarks

• Mixing trial-to-paid with download-to-paid. A 40% trial conversion on a 5% trial start rate is a 2% funnel.
• Comparing an average (2024 edition) with a median (2025 and 2026 editions) and calling the difference a trend.
• Quoting the global median when your app lives in one category and one region; the spread between India/South-East Asia and North America is more than double on most metrics ¹.
• Reporting churn from a snapshot of all subscribers instead of a cohort; cohort retention at month 12 is the only figure comparable to the report.
• Ignoring involuntary churn, which is a third of Play cancellations ².
• Treating the report as the whole market; it is a self-selected sample, and the top 10% in most categories convert at two to three times the median ³.