Hard Paywalls Convert 5x More Trials Than Freemium. A Year Later, Retention Is a Coin Flip Apart.

Every mobile subscription team has had some version of the same argument. Someone wants to gate the app behind a hard paywall to push conversion. Someone else pushes back: won't that just convert a bunch of price-shocked, half-committed users who cancel the second the trial ends? The assumption underneath that objection is that a hard paywall buys you volume at the cost of quality — more subscribers, but worse ones. RevenueCat's 2026 State of Subscription Apps report, built from event-level data across more than 115,000 apps and $16 billion in tracked revenue, gives a clean answer to that argument, and it isn't the one either side usually expects.

Hard paywalls do exactly what everyone assumes on the conversion side. What they don't do is produce a worse subscriber. That distinction matters a lot more than it sounds, because it changes what paywall design is actually a decision about.

The conversion gap is real, and it's not close

RevenueCat measured day-35 trial-to-paid conversion by paywall type across its full dataset. Hard paywall apps — where a user has to subscribe before using the core product at all, past a trial or a locked demo — converted at a median 10.7%. Freemium apps, where the product stays usable indefinitely at no cost with certain features or limits gated, converted at 2.1%. That's roughly a 5x gap, and it shows up again in revenue per install: hard-paywall apps generated $3.09 per install by day 60, against $0.38 for freemium — an 8x difference.

Source: RevenueCat, State of Subscription Apps 2026.

On the numbers so far, a hard paywall looks like a strictly better decision — more conversions, more revenue per install, no contest. This is exactly where most teams stop looking, because the conversion and revenue lines are the ones that show up on a dashboard by default. The retention line is the one you have to go dig up separately, cohort by cohort, months after the decision was already made and shipped.

The twist: a year later, you can't tell the paywall type apart

This is the part that should actually change how teams argue about paywalls. Subscribers who converted through a hard paywall retained at 27% after twelve months. Subscribers who converted through freemium retained at 28%. RevenueCat's own read on the gap is that it's statistically negligible, and given the sample size behind it — tens of thousands of apps across categories — that's a hard finding to wave away as noise.

The intuitive story was adverse selection: force people to pay before they've had time to fall in love with the product, and you'll drag in a batch of reluctant subscribers who bail the moment the charge posts. That story assumes the paywall changes who converts, not just how many convert. The data says something closer to the opposite. Once someone has actually decided to hand over a card number, they've cleared roughly the same bar of intent regardless of whether that decision happened on day one behind a hard gate or day thirty after weeks of free use. The paywall changes the size of the funnel at the top. It doesn't change much about who makes it through the middle.

Where the paywall debate is actually a distraction

If paywall type barely moves retention, then the real question teams should be litigating isn't hard-paywall-versus-freemium — it's trial length, and here the data is far less forgiving of current practice.

Source: RevenueCat, State of Subscription Apps 2026.

Trials in the 17–32 day range converted at 42.5%, against 25.5% for trials under four days — roughly 70% better for the longer window. That's a far larger, more consistent effect than anything tied to hard-paywall-versus-freemium, and it's happening on the side of the funnel most teams treat as a minor settings toggle rather than a real product decision.

What makes this worth flagging is the direction the industry is actually moving. The share of apps running trials under four days climbed from 42.1% in 2025 to 46.5% in 2026 — more teams shortening trials, not fewer, in the same year the data shows longer trials converting substantially better. Some of that is chasing faster payback periods and defensible unit economics under investor pressure; a shorter trial gets a subscriber onto the books, and the revenue clock starts, sooner. But if the goal is maximizing the number of people who actually convert, the aggregate data says a meaningful chunk of the industry is optimizing the wrong variable, on purpose, in the wrong direction.

Why retention holds steady even though the funnel doesn't

The mechanism worth understanding here isn't mysterious once you separate two different questions that are easy to collapse into one: how many people convert, and how good are the people who convert. A hard paywall filters at the point of entry — you only see people who were willing to commit before extended free use. Freemium filters over a longer window — people self-select out during the free period, and only the ones who reach genuine upgrade intent convert at all. Both processes are filtering for the same underlying thing, real intent to keep using the product, they're just doing it at different points in the funnel and at different speeds. Retention comes out similar because by the time someone has actually paid, under either model, they've cleared a comparable bar. The freemium funnel just takes longer and loses more people along the way to get there.

This has a direct parallel to a debate we've covered before in trial-to-paid conversion benchmarks: opt-in trials that require a card upfront convert at a much higher rate than opt-out trials that don't, for the same reason a hard paywall beats freemium — fewer casual sign-ups, more filtered intent at the door. The RevenueCat retention data adds a layer that benchmark alone doesn't cover: filtering harder at the door doesn't obviously buy you a worse cohort downstream. It mostly just buys you a smaller, faster-converting one.

What this means if you're not shipping a mobile app

Most CancelFlow customers run web-based B2B SaaS, not App Store or Play Store subscriptions, and paywall mechanics differ enough between the two that you shouldn't import RevenueCat's exact percentages into a board deck. But the underlying finding travels fine: don't make a paywall or trial-gating decision based on an assumption about the quality of subscriber it will produce, because the strongest evidence available says that assumption is usually wrong. Decide based on funnel math you can actually measure — how much your top-of-funnel volume shrinks under a stricter gate, and whether your product can prove its value inside whatever trial window you're testing.

That second point is where the real leverage sits, for mobile and web SaaS alike. If paywall aggressiveness doesn't meaningfully change who sticks around, then the retention work worth doing lives somewhere else entirely — in the first days and weeks after someone converts, which is exactly the territory we cover in early-lifecycle churn, and in understanding what actually drives someone to cancel once they're in, which is the whole premise behind why customers cancel in the first place. A trial or paywall redesign is a funnel-size decision. Fixing the reasons a paying subscriber quietly disengages six weeks in is a retention decision. Teams that spend a quarter re-litigating paywall placement while a shaky onboarding sequence keeps leaking subscribers at month two are solving the smaller problem.

It's also a reminder that the interval you sell on shapes the shape of your churn curve independent of paywall type — something our annual cancellation timing research gets into from a different angle, using the same RevenueCat dataset. Paywall type is a top-of-funnel lever. Billing interval, trial length, and what happens after signup are the levers that actually decide who's still around a year later. If you want to see what a given shift in retention is worth to your own numbers rather than the industry median, run it through our retention rate calculator — and if the leak you're chasing is happening at the cancel button rather than the paywall, that's the exact moment CancelFlow is built to catch.

Frequently asked questions

Stop losing subscribers today

One script tag. One function call. A live cancellation flow in under 10 minutes.