Your paywall is a 60-second decision that decides your whole business

A subscription app has exactly one screen that makes money. It is seen for about a minute, usually once, and most teams design it in an afternoon after months on everything else. Here is how we think about ours.

When it appears

After the first personal result. The user has told the app about their child, their city, their coffee order, and has seen the app do its one thing with that information. Now the paywall is not asking them to pay for a promise. It is asking whether they want to keep what they just saw. That is a much easier question.

What it says

Lead with what the person keeps, in their own terms. Not "unlimited generations" but "a new story for Lily every night". Not "premium voices" but "told in your voice". The feature list belongs below, in smaller type, for the people who want it. The headline is about them.

How many plans

One recommended plan, clearly marked. A second, usually annual, for people who want the saving. A third option is where decisions go to die. If you have a lifetime plan, think hard about whether it is a product decision or a cash-flow decision.

Price and trial, honestly

The price per period, the trial length and what happens when it ends, all visible without scrolling. Apple requires this and users check for it. A trial that hides the price converts worse, not better, because people assume the hidden number is bad news.

The close button

Visible, immediately, in the usual place. A paywall that traps people converts a few extra percent for a week and then produces refund requests, one-star reviews that mention the trap by name, and attention from App Review. The people who closed it are not lost; they are the audience for the second paywall, after another good result.

Dark patterns that look like wins

• Countdown timers that reset on every open.
• "Trial" toggles that silently switch the plan.
• Prices shown per week to look small, charged per year.
• Decline buttons written to shame ("No, I don't want to help my child sleep").

Each of these moves the number for a short time and damages the thing the number depends on: whether people trust the app. We do not use them in our apps and we do not build them for brands.

Test it like the most important screen, because it is

Headline, order of plans, trial length, placement in the flow: one change at a time, measured on trial starts and, more importantly, on trial-to-paid and first renewal. A paywall that wins on trial starts and loses on renewals has merely moved the refund to a later date.