Subscription vs One-Time Purchase: 63.5% of Apps Are Subscriptions-Only

The short answer: subscribe if your value renews

Subscription vs one-time purchase: which model should your app use? For most apps the answer is a subscription. 63.5% of apps are subscriptions-only, and the deciding variable is whether your value renews (RevenueCat State of Subscription Apps 2026, 115,000 apps).

If your app delivers ongoing utility that the user returns for (tracking, coaching, content, a tool they reopen weekly), a subscription matches how the value arrives. If the value is consumed or unlocked once, a one-time purchase fits better. The model follows the category, and the market has mostly already voted.

Subscription, one-time, or hybrid: what each model wins

A subscription bills on a repeating cadence and suits value that renews. It's the default for a reason: recurring revenue compounds, and 63.5% of apps run it as their only model. Once you've picked it, the cadence question (weekly, monthly, or annual) is a separate decision, covered in subscription billing period: weekly, monthly, or annual.

A one-time purchase (a single unlock, a consumable pack, or a donation) suits value that's used up or bought once. Medito runs no paywall at all, just a donation ask, and survives on under a 1% donor rate only because the free app is one node in a larger nonprofit ecosystem of retreats and grants. That's the tell: a pure one-time model works when the app isn't the whole business.

Hybrid runs both at once. It's tempting because it captures two kinds of willingness-to-pay, but only about 10% of apps actually operate it (Thomas Petit). It's a commitment, not a bolt-on.

The market already voted, and geography barely changes it

63.5% of apps are subscriptions-only, and that number is stable across the world: developer geographies land between 56.4% and 61.0% subscriptions-only (RevenueCat 2026). This isn't a Western quirk that reverses elsewhere. It's structural.

The money agrees. Non-game in-app-purchase revenue grew 21% year over year, and only $3.5B of that was generative AI (Olivia Moore, a16z). So the growth is broad-based, not an AI mirage riding one hot category. The revenue itself is a power law regardless of model, which is its own breakdown: how much money subscription apps make.

Gaming is the exception that proves the rule

Gaming is the one category that breaks the subscription default: 40.5% subscriptions-only against 27.5% consumables, the highest consumable share anywhere (RevenueCat 2026). The reason isn't a different audience. It's a different value shape.

A game's value gets consumed. Lives, gems, boosts, and skins are used up or bought once, so a per-item purchase matches the moment better than a monthly bill. That's the rule underneath the whole comparison. The model follows how the value is delivered, and gaming just delivers it in discrete, consumable units. Notice the shape of your own value before you copy anyone's pricing.

Hybrid is a hard operating choice, not a free upside

The tempting move is 'why not both.' Offer a subscription AND a one-time option, capture the commitment-averse buyer and the recurring one. In theory hybrid widens the top of your revenue. In practice only about 10% of apps run true hybrid (Thomas Petit), and the gap between theory and practice is the operating cost.

Running two models means two pricing pages to maintain, two mental models for the user to resolve at the paywall, and a live cannibalization question. Every lifetime buyer is a subscriber you didn't get. More plans can also add decision friction, which is its own tradeoff in how many plans a paywall should show.

The accessible on-ramp is the lifetime plan: a single one-time purchase sitting inside an otherwise subscription app. One in four apps offers one. Daylio does roughly $100K a month with a three-tier paywall (annual, monthly, and lifetime), using the lifetime option less as a volume seller and more as a confidence signal that the app is worth owning. That's the honest way to test hybrid. Add one one-time tier, then watch whether it grows the pie or just moves subscribers into it. If you're leaning on price to do the work, raising the subscription price usually beats bolting on a cheap lifetime escape hatch.

How to choose your model

If you're building with an AI coding agent, tasu's MCP is the conversion-expertise layer it calls while it builds your app: the monetization-mix data above, with sources, on tap inside your editor.

• Name your value shape first: renewing value (content, tracking, a weekly-use tool) points to subscription, consumed or one-shot value (game currency, a single unlock) points to one-time
• Default to subscription unless you have a specific reason not to. 63.5% of apps do, across every geography, because recurring value is the common case
• Go hybrid only when you genuinely serve both buyers AND can operate two models. It's a commitment about 10% of apps make, not a free add-on
• Test hybrid cheaply: add one lifetime tier (the on-ramp one in four apps already use) and measure whether it grows revenue or just relocates subscribers
• Separate the model question from the cadence question. Pick subscription vs one-time first, then decide weekly, monthly, or annual