What the best subscription apps do differently, by the numbers
Where the paywall goes, how long onboarding is, what to charge and which plan to lead with — the four levers with the largest documented effect on iOS subscription conversion, with the published figures and their sources.
Most of what is written about app funnels is folklore
Ask ten founders how many onboarding screens an app should have and you will get ten confident answers and no citations. The measured record is thinner than the advice suggests, but it exists, and where it exists it is surprisingly consistent. Broctic's funnel playbook collects it: every figure with the report it came from, the metric as that report defines it, the year, the sample and — where the publisher sells something the number flatters — a caveat.
The paywall goes at the end of onboarding
The largest documented effect is placement. Apps with a hard paywall convert a median 12.1% of installs to payers by day 35; freemium apps convert 2.2% (RevenueCat, State of Subscription Apps 2025). The cost is refunds — 5.8% against 3.4% — because a hard gate converts some people who had not yet decided. Timing explains most of it: 82% of trials start on the day the app is installed. An offer met on day three, in Settings, is an offer to someone who has already concluded the free version is enough.
Onboarding is three to five screens, then value
Design onboarding around the first useful outcome, defer nonessential setup and measure your own cohorts. Earlier numerical onboarding claims on this page were retired because their primary study could not be verified. See the reusable app system research for current sources and experiment guidance.
Pricing low to convert is the expensive mistake
High-priced apps start more trials (a median 9.8% of downloads against 4.3% for low-priced apps), convert more installs to payers, and earn roughly seven times the first-year value per new subscriber — $55.21 against $8.08 (RevenueCat 2025). Price signals seriousness, and the people it selects are the ones who wanted the outcome enough to pay for it; they also retain better. And which plan you lead with is a retention decision in disguise: weekly plans renew at 3.4% after a year, monthly at 17%, annual plans keep about half.
What is not measured
There is no independently published figure for the effect of a specific onboarding question, a tutorial style, or a paywall layout — only vendor case studies, which the playbook quotes with their caveats, and practitioner consensus, which it labels as such and gives no number. The checklist version of the playbook is served at /llms.txt so a coding agent building an app can read the evidence in one pass and measure its own funnel against it.