Hard Paywall vs Freemium: The 5x Conversion Gap, and When Each Wins

The short answer: hard paywalls convert 5x better, but read the nuance

Hard paywall vs freemium is not close on conversion: hard paywalls convert about 5x better, at 10.7% of downloads paying by Day 35 versus 2.1% for freemium (RevenueCat, 115,000 apps). If you want subscription revenue now, a hard paywall is the default. The nuance is what decides the exceptions, so read past the headline before you rip out your free tier.

Two numbers keep this honest. Hard-paywall conversion ranges from 4.2% to 38.7%, so a badly built hard paywall loses to a well-built freemium funnel. And year-1 retention is nearly identical between the two models, so the 5x gap is about conversion, not whether users stick. The rest of this piece is when each model wins.

The numbers: 5x on conversion, 8 to 9x on revenue per install

RevenueCat's 2026 report measures Day-35 download-to-paid at 10.7% for hard paywalls and 2.1% for freemium. The revenue gap is wider than the conversion gap. Hard paywalls earn $2.32 per install at Day 14 versus $0.27 for low-priced freemium, and $3.09 versus $0.38 at Day 60 (Phil Carter's cut of the data). That is 8 to 9 times more revenue per install, because a hard paywall selects for payers and typically pushes annual plans.

The variance is the part most summaries drop. Hard-paywall conversion runs from a 4.2% floor to 38.7% at the 90th percentile. Freemium runs 0.3% to 8.2%. The ranges overlap, which means the model label predicts less than execution does. A top-decile freemium app at 8.2% converts better than a bottom hard paywall at 4.2%. Pick the model for your business, then out-execute your benchmark.

Why hard paywalls convert: they truncate the zero-willingness-to-pay tail

The mechanism is selection, not persuasion. A hard paywall does not talk anyone into paying. It removes the people who were never going to pay before they reach a purchase decision. Freemium admits everyone, including the users whose willingness to pay is zero, and those users sit in your denominator dragging conversion down.

Put plainly, conversion is the share of entrants whose willingness to pay clears your price. A hard paywall chops off the zero-willingness-to-pay tail at the door; freemium keeps it. This is the explanatory model for the 10.7% versus 2.1% gap, not a separately measured effect, so treat it as the why behind the benchmark rather than its own datapoint. It also explains the revenue gap: the users who survive a hard paywall skew toward higher intent, so they tend to buy annual and monetize better per install.

The catch is that self-selection cuts both ways. A hard paywall also turns away users who would have paid if they had felt the product first. That is why the gate works when your value is obvious upfront and leaks when your aha moment needs a few minutes to land.

The nuance the headline hides: year-1 retention is nearly identical

Here is the number that changes how you read the other numbers. Year-1 retention is nearly identical between hard paywalls and freemium in RevenueCat's data. The 5x gap is entirely on getting the first payment, not on keeping it. A hard paywall wins more payers at Day 35. It does not produce stickier subscribers a year later.

That matters for two decisions. If your problem is durability, not conversion, switching to a hard paywall won't fix it, and you should look at why users leave, whether that's a Day-0 free-trial cancel, faster AI-app churn, or involuntary billing failures on Google Play. The two models also cash in on different timelines. Hard paywalls spike again at Day 4 to 7 as free trials expire, while freemium has a long tail, with 23% of its conversions landing 6 or more weeks out.

When freemium or no paywall is the right call

The 5x number assumes you want subscription revenue now and your product can earn it at the door. Change either assumption and the answer changes. The tasu brain keeps this tension open on purpose, because the deciding variables are real.

The first variable is product maturity and whether your aha fits inside a trial. An early product often can't retain users through a free trial yet, so the advice from founders like Refeel is blunt: start with a hard paywall, pay to use, no trial, get money in. Flip it once the product delivers real value fast, and a trial wins, because longer trials convert better, at 42.5% for 17 to 32 day trials versus 25.5% for 4 days or fewer. If your aha needs minutes to land, like SuperChinese's speaking moment sitting behind a few lessons, a hard paywall in front of it turns away buyers who never felt the value.

The second variable is your business model and funding. A hard paywall is the default only if subscription revenue is the goal right now. Some apps skip the onboarding paywall on purpose: VC-funded growth plays optimizing for scale, donation and goodwill apps, and content-funnel-first products like Deepstash, which runs a 14-screen onboarding with no commercial ask and monetizes later. That's a different game, not a mistake. If your model is one of those, freemium or no paywall is correct even though it converts worse on the Day-35 chart.

So the honest rule is two questions, not one. Do you want subscription revenue now, and does your aha fit where the paywall sits? Two yeses point to a hard paywall. A no on either points to freemium, a trial, or no paywall, matched to your model.

If you stay freemium, run the reverse trial

Freemium's conversion gap is not fixed. The counter-play with the biggest lift in the report is the reverse trial, and it flips the dismissal into the sell. When a freemium user closes your paywall, don't fight the no. Grant them temporary full premium access, no card required, right at that moment.

Now they use the premium features instead of imagining them. When the access is about to end, cancelling feels like losing something they already own, not declining something they never had. Loss aversion and the endowment effect do the work. Steve Young of App Masters reported this lifting freemium conversion from 0.4% to 4.5% with no change to traffic or pricing.

The condition is engagement, not exposure. It only works if they actually use the premium features during the window, so the more they use them, the stronger the pull to keep them. A reverse trial handed to a user who never opens the premium features converts nothing. It's the freemium sibling of a hard paywall's trial expiration: ownership on dismissal instead of a wall on entry.

How to decide for your app

• Ask two questions first: do you want subscription revenue now, and does your aha moment fit where the paywall sits. Two yeses mean a hard paywall
• If your product is early and can't retain through a trial yet, start with a hard paywall, no trial, and get money in. Add the trial once the product earns it
• If your aha needs minutes to land, put the value before the gate, or you convert your best users worst. Move the paywall, don't just redesign it
• If your model is growth, donation, or a content funnel, skip the onboarding paywall on purpose. The Day-35 chart is not your scoreboard
• If you stay freemium, run a reverse trial: grant premium on dismissal, no card, and let loss aversion close it. It lifted one app from 0.4% to 4.5%
• Whichever you pick, execution beats the label. Hard paywalls span 4.2% to 38.7%, so benchmark against your category and out-build the median

The label matters less than the execution

Hard paywall versus freemium is a real 5x gap on conversion, and for most subscription apps that want revenue now, the hard paywall is the right default. But the model you pick matters less than how well you build it, and the exceptions (early products, slow ahas, growth and content models) are real, not excuses. Decide by your business and your aha, then out-execute your benchmark.

If you want these benchmarks on tap while your AI coding agent builds your paywall, that's what the tasu MCP is: the conversion-expertise layer your agent calls as it works. It returns sourced claims, not code, so your agent does the building.