How Many Onboarding Screens Should a Mobile App Have? 6 to 42, and Both Win

Screen count is a symptom, not a lever

How many onboarding screens should a mobile app have is a question that sounds like it wants a number. It does not have one. In the tasu library, Liftoff runs 42 screens, Oniri runs 34, and Catzy runs 23. In the same language-learning category, Airlearn runs 6 screens against Duolingo's 38. Both extremes convert.

So screen count is a symptom of a deeper choice, not a lever on its own. Every screen should do one job. It deposits a micro-commitment, it deepens the user's conviction that they have a problem worth solving, or it moves them toward the value moment. A screen that does none of the three is pure drop-off risk. One practitioner heuristic puts the stakes high, "the onboarding is 70% of the app" (@Jahjiren), though that 70% is a heuristic with no dataset behind it, not a measured funnel split. The direction is the useful part. Onboarding is where conversion is mostly decided, so the screens are worth getting right.

Long flows win when the user has to convince themselves

Liftoff runs 42 screens at $500K per month, 1M+ users, and a 4.7 star rating. Oniri runs 34 screens and is a 10-year-old app that shows no trial on its paywall. Catzy runs 23 screens at $35K MRR, bootstrapped, with a 4.9 star rating. These flows are long on purpose. Each screen deposits a small yes, and by the paywall the user has talked themselves into the purchase. This is effort justification. The work the user puts in becomes the reason they buy.

The category is the tell. Self-care, habit, and health apps sell an identity, not a utility. The user is buying a version of themselves, and that decision takes screens to build. The six jobs of onboarding are what those screens are doing: personalization, commitment, social proof, and the rest. Length is fine when every screen earns its place with commitment, not tutorial.

Short flows win when the product shows value fast

Airlearn runs 6 screens and puts its paywall on screen 6, with a 14-day trial. Duolingo runs 38 in the same category and converts roughly 10% of about 60M monthly active users to paid at $1B ARR. The gap is not sloppiness. Airlearn treats every screen past the minimum as drop-off risk for a user who is already motivated to learn a language. It gets out of the way.

The extreme version is a pre-converted user. Fishing Points runs 11 screens at 10M+ installs and a 4.7 star rating. Serious anglers arrive already sold, so a thin flow is correct, and the binding constraint moves off the onboarding and onto the price, which sits near $10 per year. When the user is converted before install, more screens buy nothing.

The deciding variable: conviction needed vs value speed

This is a genuine tension, and the tasu brain keeps it open rather than flattening it. One side: more screens mean more micro-commitments and more self-conviction before the paywall, so long flows convert for products that need the user to talk themselves into it. The other side: every extra screen is another drop-off point, so short flows convert for products that deliver value instantly.

The deciding variable is how much the user needs to convince themselves versus how fast the product can show value. Emotional and identity purchases, like self-care, habit, and health, tolerate long flows that build commitment. Utilities that deliver instantly should get out of the way. A deeper read of the same variable is how converted the user already is at download. A passionate, high-intent niche arrives pre-converted, so a thin flow is right and the price becomes the real leak. Length must earn its place with commitment, never with tutorial.

How to set your own screen count

• Label every onboarding screen with its job: a micro-commitment, a conviction beat, or a step toward the value moment. A screen with no job is the one to cut
• Selling an identity (self-care, habit, health)? A long flow can earn its length. Build conviction screen by screen before the paywall, the way Liftoff and Catzy do
• Delivering an instant utility? Get to value fast and put the paywall close behind it. Airlearn reaches it by screen 6
• Serving a pre-converted niche? Keep it thin and look hard at the price. When the flow is not the leak, the price usually is
• Do not copy a competitor's screen count. Copy their structure, then set length from your own conviction-versus-value math (see the six jobs of onboarding)