RevenueCat

Mojo used new-cohort 7-day ARPU as its single monetization metric. Mojo picked a 7-day window to cover its 3-day trial and confirmed paywall and pricing lifts in RevenueCat Charts, which offered 7-day ARPU as a view.

RevenueCat's blog first published Mojo's yearly-only default paywall test.

App Masters

CEO involvement signals priority, and a clear line prevents product drift. Jacob Eiting gets involved in priorities, which signals importance and brings his knowledge. He also recalled mascot characters creeping onto every dashboard page until the product looked like a McDonald's.

Subversive

Overlay revenue spikes on posting dates to see which content converts. Sideshift lays RevenueCat and Stripe revenue, plus App Store installs, over creator video performance to see which organic videos drive revenue, connecting Superwall too.

Attribution stays imperfect since you only feel the million-view hits; checking creator audience geography helps detect view botting. "It's always the attribution piece that is so foggy."

Sub Club▶ 55:50

Longer lookbacks and combined views help teams judge matured cohorts. RevenueCat widened its experiment lookback window to over a year, which BoldVoice finds easier than homemade analytics. Ryan Thorpe pulled RevenueCat, App Store Connect, Google Play and ad data into spreadsheets before building Flo to combine them.

Coconote's Zack Hargett watched the RevenueCat chart from day one, treating paying customers as a stronger signal than usage events.

Subversive

Overlay revenue spikes on posting dates to see which content converts. Sideshift lays RevenueCat and Stripe revenue, plus App Store installs, over creator video performance to see which organic videos drive revenue, connecting Superwall too.

Attribution stays imperfect since you only feel the million-view hits; checking creator audience geography helps detect view botting. "It's always the attribution piece that is so foggy."

abtest.design

CEO involvement signals priority, and a clear line prevents product drift. Jacob Eiting gets involved in priorities, which signals importance and brings his knowledge. He also recalled mascot characters creeping onto every dashboard page until the product looked like a McDonald's.

Sub Club

Check whether an AI feature is affordable for every user every day. RevenueCat builds on a frontier Claude model and estimated one LLM call per app per day would cost tens of thousands of dollars a month. Opus pricing had not come down in six months.

Cheaper models often fail on trivial tasks once you have used the best one, so routing is not a free fix. Eiting's view is that venture money fits products with viral growth and high inference costs, betting costs fall in time. "Can we afford to do this for every user every day?"

App Masters▶ 15:48

Payments integrations open new revenue, like HSA billing. Natal added HSA billing through the RevenueCat and Flex integration and saw tens of thousands of dollars come through in two weeks with no announcement.

Pre-tax money acts like a 30 to 40% discount, and RevenueCat's Daily Payouts is a factoring product that could extend to loans against run rate.

App Masters

Payments integrations open new revenue, like HSA billing. Natal added HSA billing through the RevenueCat and Flex integration and saw tens of thousands of dollars come through in two weeks with no announcement.

Pre-tax money acts like a 30 to 40% discount, and RevenueCat's Daily Payouts is a factoring product that could extend to loans against run rate.

Superwall▶ 19:40

Hard paywalls convert about five times better than freemium without hurting retention. Download-to-paid by day 35 is 10.7% for hard paywalls against 2.1% for freemium (cited as 2.7% in another talk), while median year-one retention is 26.8% versus 27.7%.

Phil Carter saw one client's shift to freemium cut subscriber conversion by more than 50%, so it was pulled after a couple of weeks. Freemium suits apps chasing a huge top of funnel like Spotify, Duolingo and Strava, and only works if free users bring value through upsells, ads or organic growth. The hard paywall figure ignores users who never reach a paywall.

Sub Club

Longer trials often convert better, yet the industry keeps shortening them. RevenueCat's data shows 14 to 30 day trials often convert better, while most companies trend toward shorter trials. Carter suspects AI apps drive the trend because free usage burns compute.

The hosts admit longer trials may look better only because of selection bias.

Sub Club▶ 56:00

Compute costs push AI apps to short trials, credits and paid tiers. Serving each user costs real compute, so AI apps use very short trials or none, metered credits and extra paid tiers. A 30-day free trial lets heavy users rack up costs and churn without paying.

Tolen started with a 7-day trial and adjusted trial length and freemium over time. "You have whales racking up massive compute costs without paying you a dime and then they churn."

Sub Club▶ 56:00

It uses top models for its data chatbot, expecting costs to fall. RevenueCat is building Rico, an agent that answers questions about revenue data, plus agentic paywall creation for vibe-coded apps.

Carter argues model choice should match the customer, planning 6 to 12 months out as prices fall; David Barnard pushed back that premature cost optimization can cost you the best experience.

Sub Club▶ 50:50

RevenueCat turned its customer data into the benchmark operators lacked. Teams stare at their own dashboards and can't tell if their numbers are good, so RevenueCat publishes an annual State of Subscription Apps report from what Rik Haandrikman calls the biggest cross-platform monetization data set. It powers purchases for over 100,000 apps across Apple, Google, web and Amazon.

The 2026 edition is the fourth, runs about 300 pages and took four months to write; it isn't capped, so it grows each year. Data is anonymized and aggregated, which is why it shows distributions rather than scatter plots, and only apps with active subscription revenue above install and revenue thresholds are included.

App Masters

Growth concentrates at the top while new app supply explodes. The top 10% of apps grew 306% year over year while the median grew 5.3%, and the top 25% grew 80%. New subscription apps per month rose from 2,000 in January 2022 to 14,700 in January 2026, about 7x, most of it in the last six months.

iOS now takes 77% of new launches, up from 67% in 2023. RevenueCat sees daily app submissions roughly triple, with about 7x more new users a month but roughly the same support ticket volume, because new customers talk to an AI agent that sets up their account. Jacob Eiting: "These markets are just sorting machines."

App Masters

Old apps still earn most of the money because winners compound. 69% of store revenue comes from apps released before 2020; apps from 2020 to 2021 earn 9% and apps from 2022 to 2023 earn 14%.

Strava, ClassDojo, Duolingo and the big media apps are named as compounding winners. Eiting says this chart is where disruption would show first if newer cohorts start taking share.

App Masters

With polish now table stakes, niche focus and community decide winners. In the Shipyard hackathon, 15% as many participants delivered twice as many finished products at a much higher quality bar, so a finished-looking app no longer stands out. Cal AI won users who wanted speed over perfect macros and was acquired by My Fitness Pal; Shotsy started as a tool for its founder's GLP-1 Reddit group.

Rik says niche apps can now reach $10 to 20 million ARR and that building in public only works inside the community you serve, such as Studyfetch on TikTok where students look for study help. "You could probably make something that looks and feels like Duolingo for the first 30 minutes, but then it'll be done."

App Masters

Read report benchmarks as correlations and test them on your app. The figures mix selection effects, such as who chooses long trials, and the median 5.3% growth includes apps owners have abandoned. Run the test on your own business before adopting a benchmark.

The data covers only apps using RevenueCat's SDK, and which apps count as AI is a subjective call. Eiting treats a 33% drop as the sign to rethink strategy. "You should never take one of these single data points and treat it as gospel."

App Masters

Hard paywalls convert about five times better than freemium without hurting retention. Download-to-paid by day 35 is 10.7% for hard paywalls against 2.1% for freemium (cited as 2.7% in another talk), while median year-one retention is 26.8% versus 27.7%.

Phil Carter saw one client's shift to freemium cut subscriber conversion by more than 50%, so it was pulled after a couple of weeks. Freemium suits apps chasing a huge top of funnel like Spotify, Duolingo and Strava, and only works if free users bring value through upsells, ads or organic growth. The hard paywall figure ignores users who never reach a paywall.

App Masters

Most users decide on day zero, so onboarding and paywall carry the business. 80 to 90% of trial starts and half of conversions happen on day 0, and 55% of 3-day trial cancellations happen on day zero. About 80% of payers across the industry convert on day one.

34% of annual subscribers who churn cancel in the first month, and only 11% cancel right at renewal. Mojo gets about 50% of payers on day one and reads that lower share as a sign of good retention and a base of existing users still to monetize.

App Masters

AI apps earn more per payer but lose users faster. AI apps make 41% more revenue per payer but churn 30% faster, with year-one retention of 21% against 31% for other apps.

Users hop between AI tools like streaming services and return when a better model launches, so reactivation is strong. Rik argues only products built on a bond beyond being a tool hold users.

App Masters

Compute costs push AI apps to short trials, credits and paid tiers. Serving each user costs real compute, so AI apps use very short trials or none, metered credits and extra paid tiers. A 30-day free trial lets heavy users rack up costs and churn without paying.

Tolen started with a 7-day trial and adjusted trial length and freemium over time. "You have whales racking up massive compute costs without paying you a dime and then they churn."

App Masters

A judgment-free Slack channel for dumb ideas produced major initiatives. The default assumption in the channel is that an idea is bad, so people share without fear. The AppGrowth conference, planned in three months, and App Store rejection socks both came from it.

App Masters▶ 7:04

Staff spend one to two hours a day tinkering with AI tools. Rik tells his team to pause regular work daily to try tools like Perplexity, computer use and OpenClaw, and the company moved everyone onto Manus for complex projects, making the team much faster.

Rik runs an OpenClaw setup on a Mac Mini that builds niche apps around the clock. "Whatever you invest now is almost certainly going to pay dividends."

App Masters▶ 9:26

Longer lookbacks and combined views help teams judge matured cohorts. RevenueCat widened its experiment lookback window to over a year, which BoldVoice finds easier than homemade analytics. Ryan Thorpe pulled RevenueCat, App Store Connect, Google Play and ad data into spreadsheets before building Flo to combine them.

Coconote's Zack Hargett watched the RevenueCat chart from day one, treating paying customers as a stronger signal than usage events.

App Masters

RevenueCat turned its customer data into the benchmark operators lacked. Teams stare at their own dashboards and can't tell if their numbers are good, so RevenueCat publishes an annual State of Subscription Apps report from what Rik Haandrikman calls the biggest cross-platform monetization data set. It powers purchases for over 100,000 apps across Apple, Google, web and Amazon.

The 2026 edition is the fourth, runs about 300 pages and took four months to write; it isn't capped, so it grows each year. Data is anonymized and aggregated, which is why it shows distributions rather than scatter plots, and only apps with active subscription revenue above install and revenue thresholds are included.

Superwall

Growth concentrates at the top while new app supply explodes. The top 10% of apps grew 306% year over year while the median grew 5.3%, and the top 25% grew 80%. New subscription apps per month rose from 2,000 in January 2022 to 14,700 in January 2026, about 7x, most of it in the last six months.

iOS now takes 77% of new launches, up from 67% in 2023. RevenueCat sees daily app submissions roughly triple, with about 7x more new users a month but roughly the same support ticket volume, because new customers talk to an AI agent that sets up their account. Jacob Eiting: "These markets are just sorting machines."

Superwall

Old apps still earn most of the money because winners compound. 69% of store revenue comes from apps released before 2020; apps from 2020 to 2021 earn 9% and apps from 2022 to 2023 earn 14%.

Strava, ClassDojo, Duolingo and the big media apps are named as compounding winners. Eiting says this chart is where disruption would show first if newer cohorts start taking share.

Superwall

Subscription retention makes paid acquisition a sound bet. With about 30% annual retention, a $1M ad spend in March 2026 yields about $300K of free cash flow in March 2027, assuming the spend pays back within the first year.

Churned customers also stay on a list that can be marketed to for years. Eiting still sees venture capital as a rare fit for apps.

Superwall

Read report benchmarks as correlations and test them on your app. The figures mix selection effects, such as who chooses long trials, and the median 5.3% growth includes apps owners have abandoned. Run the test on your own business before adopting a benchmark.

The data covers only apps using RevenueCat's SDK, and which apps count as AI is a subjective call. Eiting treats a 33% drop as the sign to rethink strategy. "You should never take one of these single data points and treat it as gospel."

Superwall

Hard paywalls convert about five times better than freemium without hurting retention. Download-to-paid by day 35 is 10.7% for hard paywalls against 2.1% for freemium (cited as 2.7% in another talk), while median year-one retention is 26.8% versus 27.7%.

Phil Carter saw one client's shift to freemium cut subscriber conversion by more than 50%, so it was pulled after a couple of weeks. Freemium suits apps chasing a huge top of funnel like Spotify, Duolingo and Strava, and only works if free users bring value through upsells, ads or organic growth. The hard paywall figure ignores users who never reach a paywall.

Superwall

Most users decide on day zero, so onboarding and paywall carry the business. 80 to 90% of trial starts and half of conversions happen on day 0, and 55% of 3-day trial cancellations happen on day zero. About 80% of payers across the industry convert on day one.

34% of annual subscribers who churn cancel in the first month, and only 11% cancel right at renewal. Mojo gets about 50% of payers on day one and reads that lower share as a sign of good retention and a base of existing users still to monetize.

Superwall

Longer trials often convert better, yet the industry keeps shortening them. RevenueCat's data shows 14 to 30 day trials often convert better, while most companies trend toward shorter trials. Carter suspects AI apps drive the trend because free usage burns compute.

The hosts admit longer trials may look better only because of selection bias.

Superwall

Turn on grace periods and billing retry on Android. Nearly a third of all subscription cancellations on Google Play are involuntary billing failures, from causes like prepaid cards running out and no auto-recovery path like iOS.

Enable grace periods and retry features and tell users about payment problems, especially once Android is more than experimental revenue.

Superwall▶ 10:35

AI apps earn more per payer but lose users faster. AI apps make 41% more revenue per payer but churn 30% faster, with year-one retention of 21% against 31% for other apps.

Users hop between AI tools like streaming services and return when a better model launches, so reactivation is strong. Rik argues only products built on a bond beyond being a tool hold users.

Superwall

It uses top models for its data chatbot, expecting costs to fall. RevenueCat is building Rico, an agent that answers questions about revenue data, plus agentic paywall creation for vibe-coded apps.

Carter argues model choice should match the customer, planning 6 to 12 months out as prices fall; David Barnard pushed back that premature cost optimization can cost you the best experience.

Superwall

Most users decide on day zero, so onboarding and paywall carry the business. 80 to 90% of trial starts and half of conversions happen on day 0, and 55% of 3-day trial cancellations happen on day zero. About 80% of payers across the industry convert on day one.

34% of annual subscribers who churn cancel in the first month, and only 11% cancel right at renewal. Mojo gets about 50% of payers on day one and reads that lower share as a sign of good retention and a base of existing users still to monetize.

Sub Club▶ 47:29

Mojo used new-cohort 7-day ARPU as its single monetization metric. Mojo picked a 7-day window to cover its 3-day trial and confirmed paywall and pricing lifts in RevenueCat Charts, which offered 7-day ARPU as a view.

RevenueCat's blog first published Mojo's yearly-only default paywall test.

Sub Club▶ 14:47

Longer lookbacks and combined views help teams judge matured cohorts. RevenueCat widened its experiment lookback window to over a year, which BoldVoice finds easier than homemade analytics. Ryan Thorpe pulled RevenueCat, App Store Connect, Google Play and ad data into spreadsheets before building Flo to combine them.

Coconote's Zack Hargett watched the RevenueCat chart from day one, treating paying customers as a stronger signal than usage events.

Sub Club▶ 57:47