Freemium vs Hard Paywall vs Free Trial: Which Converts?

Hard paywalls convert roughly five times more installs into payers than freemium, but freemium apps reach far more people and can match them on year-one retention. The right model depends on how quickly your app proves its value, how much cash you need now, and whether you can afford to run experiments.

The seven models in one table

Superwall's taxonomy adds two soft-paywall variants, metered (free until a usage cap) and feature-gated (free app, paid features), which are freemium in this table ⁷.

Freemium

Freemium keeps a free version open indefinitely and sells the upgrade later. RevenueCat's 2026 data puts the median freemium app at a 2.1% download-to-paid rate by day 35, unchanged from 2025 ¹. The upside is reach: Spotify, Duolingo and Strava are the standard examples of freemium apps whose free users feed word of mouth and eventually a much larger paying base ⁴.

Designing the free tier is the hard part. RevenueCat's April 2026 guide describes three architectures, a "taster" (the full product with usage caps), a "split" (different features for different user types) and a hybrid, and argues that free users should make real progress without reaching the full solution ⁵. It also recommends a written "bill of rights" fixing what always stays free, so small experiments cannot erode the free tier by accident ⁵.

Hard paywall

A hard paywall blocks the core value until the user pays or starts a trial. It converts best by a wide margin: a 10.7% median download-to-paid rate at day 35 in 2026, down from 12.1% in 2025 but still five times the freemium median ^1,2. Revenue per install at day 60 is $3.09 against $0.38 for freemium, roughly eight times more ¹. Even the weaker hard-paywall apps in the 2026 dataset converted at about 4.2%, double the freemium median ⁴.

The costs are a smaller audience and a higher refund rate: 5.8% for hard-paywall apps versus 3.4% for freemium in the 2025 report ². Retention of annual subscribers after one year is now almost identical, 27% for hard paywall and 28% for freemium ¹, so the model no longer buys you stickier customers, only faster cash. That faster payback is exactly why hard paywalls dominate paid acquisition: a quick return lets you scale ad spend ⁴.

Free trials: 3, 7 or 14 days

Most hard paywalls are really trial walls. The headline 2026 finding is that longer trials convert better: 42.5% trial-to-paid for 17–32 day trials versus 25.5% for trials of four days or less ¹. Yet 46.5% of apps now run trials of four days or fewer, up from 42.1% ¹, because a short trial produces a signal in days and limits free usage of expensive features such as AI inference.

Cancellations are front-loaded. For 3-day trials, 55.4% of all cancellations land on day 0 and 84% within the first two days ¹. Adapty's 2026 report shows the same shape from a different dataset: 89.4% of trials start on day 0, the global average trial-to-paid rate is 25.6%, and trial subscribers retain 1.4–1.7 times better than people who bought directly ³. We go deeper on trial length in our trial length and conversion guide.

Reverse trials

A reverse trial gives every new user the premium product for a limited time and then drops them onto the free tier instead of locking them out. RevenueCat frames it around loss aversion: losing features you have used is a stronger motivator than never having had them, and the user stays in the app rather than churning out at the paywall ⁵. It requires a genuine free tier to land on, which is why it is a freemium variant rather than a replacement for a hard paywall. We could not find a public, dataset-level conversion benchmark for reverse trials specifically unconfirmed, so treat it as a hypothesis to test rather than a known winner.

Paid upfront

A one-off store price is still common in utilities and premium games. Apple lets non-subscription apps run a time-based trial by setting up a non-consumable at price tier 0 named "XX-day Trial", provided the app states the duration, what disappears afterwards and the full price before the trial starts ⁹. The weaknesses are structural: no renewal revenue, no win-back offers, and store commission applies to every sale. Apple drops its commission on subscriptions from 30% to 15% after a year of paid service, but a one-off purchase never reaches that tier unless you are in the Small Business Program ¹⁰.

Ads and hybrid

Ads alone rarely carry a non-game app. RevenueCat's hybrid guide notes that Spotify earns under 5% of its revenue from ad-supported users, and cites one case as low as 1% ⁶. Ads work better as the floor under a subscription: the same guide cites a 10–30% revenue uplift from pairing the two, and Impulse shows a paywall after an ad closes, then sells a $4.99 ad-removal purchase ⁶.

Hybrid adds consumables, lifetime unlocks or tiers. Around 90% of users never subscribe ⁶, and in the 2025 dataset more than 35% of subscription apps also sold consumables or lifetime purchases, rising to 61.7% in gaming ². Hybrid buyers were 7% of buyers but 25% of revenue ⁶. The risks are choice overload, extra SKUs and cannibalisation when consumables pull revenue forward and hurt renewals ⁶. Superwall's 2022 analysis of 32 million paywall opens found two products beat one by 61% and three beat two by 44%, so a little choice helps; a lot may not ⁸.

Which model fits your app

Read the tree as a starting point, not a verdict. RevenueCat's advice is to keep a hard paywall if you are bootstrapped, need cash to fund ads, solve an immediate problem, or lack the analytics to test feature gating, and to test freemium only when retention is strong and you have hit a paid-acquisition ceiling ⁴. Two of growth advisor Phil Carter's clients ran the same freemium transition: one lifted LTV per user by 75%, the other lost more than half its subscriber conversion and reverted within weeks ⁴.

Platform rules that constrain the choice

On iOS, an auto-renewable subscription must last at least seven days, work across the user's devices, and not require extra tasks such as sharing contacts to unlock paid content; apps that trick users into subscribing can be removed along with the developer account ⁹. If you move a paid app to subscriptions, users who already bought a full unlock must keep it ⁹. On Google Play, free-trial phases run from 3 days to 3 years, introductory prices must be at or below the base-plan price, and each subscription can hold up to 250 base plans and offers with 50 active at once ¹¹.

How this shows up in the charts

Monetisation model shapes where an app can rank. Freemium and ad-supported apps live on Top Free because the install is free and the funnel is wide; hard-paywall apps appear there too, but their real signal is on Top Grossing, where a 10.7% day-35 conversion stacks renewals fast. Paid-upfront apps compete on the Top Paid list, a smaller pond with far fewer entrants. Watch the iOS movers after a model change: a switch from paid to freemium often spikes Top Free rank while Top Grossing takes months to recover, because trial-to-paid cohorts need time to mature. Our guide to how App Store rankings work covers why install velocity and revenue move different charts.

A short checklist

• Measure download-to-paid at day 35 so hard and soft paywalls are comparable ¹.
• Freemium: write down what stays free before the first experiment ⁵.
• Hard paywall: budget for a refund rate near 6% and a smaller organic funnel ².
• Test trial length on your own cohorts; global medians hide category gaps ^1,3.
• Add ads or consumables only once you can see which segments never subscribe ⁶.