How to decide how long onboarding should be

1 The bottom line

Onboarding length should follow how considered the purchase is, how motivated the traffic is, and whether each step feels easy. Test it both ways, because long flows win in some apps and short flows win in others.

2 Key takeaways

• Start long for considered purchases like education, fitness and health, and short for utilities. Users of considered products need warming up before they pay.
  • Steve P. Young and Igor Lyubimov, Headway31:10: Headway's education quiz runs about 50 screens, while utility subscription funnels are usually short.
  • Ravi Mehta, Tinder59:57: Sesame's weight loss checkout went from three pages to 25, and conversion rose 40%.
• Set length by who arrives and how they arrived. Age and channel change how much effort people will give.
  • Cedric Roberge, Pep AI28:58: users under 25 get a very short onboarding, because they buy less and drop out of long ones.
  • David Futoran, Honeydew34:50: Google searchers were bottom of funnel, while Meta users who only saw an ad behaved differently in the flow.
• Judge each screen by how hard it feels, not by the click count. More steps can win when every step is easier.
  • Houzz: three one-field screens beat a single contact form and lifted conversion by 15+%, despite more clicks.
• Add friction only when it filters for quality or gives the user something real. Fewer finishers can still mean more payers and better retention.
  • Simon Mizzi, Preply24:51: a deeper questionnaire qualified high-intent learners and tired out low-intent ones who would not have converted.
  • Adam Fishman45:37: sometimes conversion should decrease a little if the users who stay are a better fit.
• Test a big change in length, not one screen at a time. Small tweaks to a decent flow rarely move the result.
  • Brett Bauman and Zack Hargett, Coconote29:05: going from about five screens to 13 raised trial conversion by 16%.
  • Steve P. Young, App Masters10:33: one client cut onboarding from seven screens to four and subscriptions improved.

Somebody coming from Google has a very different commitment to buying than somebody coming from Meta.

David Futoran, Honeydew

2.1 Read who is arriving and what they are doing

Your audience, their mood and their traffic source set how much effort you can ask for.

• Check what users are doing when they open the app. Leon Sasson, Rise15:21
  • Most Rise users download the app on the couch or in bed, scrolling and winding down.
  • They are not solving a task in three minutes, so there is room for an entertaining onboarding.
  • Sasson says 5 to 10 minutes is worth it if the onboarding is good and entertaining.
• Build onboarding around the person who really buys. Jordan Bryant, TimeClock7:36
  • Interviews showed the buyers were office managers at construction companies with 20 to 30 employees.
  • TimeClock was their first step toward digitizing, so they wanted two to three simple steps to get fully onboarded.
  • Most downloaders were construction companies with small back offices, and the office manager made the purchase decision.
• Split by age if your audience skews young and old. Cedric Roberge, Pep AI28:58
  • Under-25s get a short flow with key points and disclaimers.
  • Older users get a longer onboarding on nutrition, food, water and sleep, plus a personalized nutrition plan.
  • Over 50%, closer to 60%, of Pep AI users are over 25, so longer education works for the majority.
• Give ad traffic a longer route than direct traffic. Patrick and Pablo, Social26:40
  • Users who arrive from an ad are already excited, so WiseSam gives them a long questionnaire that reflects their answers back.
  • The answers then feed the personalization inside the product.
  • Visitors who come straight to the website get a shorter route to the experience.
  • The team calls it "happy friction" and keeps A/B testing the length. The offer is a $1 first week, then $10 a week, on web.

2.2 Pick a starting length for your category

Long suits products people think hard about, and short suits products people just want to use.

• Go long for education and fitness, and ask in batches. Steve P. Young, RiseGuide3:36
  • Ask about six questions at a time, with breaks in between.
  • You lose some users, but the ones who finish are more likely to pay.
  • RiseGuide asks a dozen or more questions with social proof breaks before the paywall.
  • Steve Young and Lucas, Paddle9:07: longer onboarding converts better in these categories despite more drop-off.
• Treat a long flow as the aha moment when belief is what you sell. Brent Zajaczkowski, Noom9:06
  • For Noom the aha moment is believing this program will work where past attempts failed.
  • The questions, tone and understanding in the flow build that belief.
  • Users have often failed many times, so a flow that looks like a generic calorie tracker is an immediate turnoff.
  • This goes against the conventional advice to shorten onboarding to reach the aha moment fast, and it has worked for Noom.
• Go short for utility and productivity apps, and say what the app does. Steve P. Young and Chaim, App Masters8:36
  • Use one or two screens on what the app does, why, and how it solves the problem.
  • Right after install is the most critical moment, because people keep only a few apps on their phone.
  • Steve P. Young, a universal TV remote app4:16: its only onboarding screen shows a 4.6 rating and the download count, then goes to the paywall.
  • That works because the app has about 58,000 ratings and ranks number 15 in Utilities.
• For a short flow, split required from optional. Oji Udezue, Typeform40:50
  • Keep the mandatory part to three screens or fewer: what explains the product and the setup needed for success.
  • Put everything else in optional guidance people can return to.
  • At Calendly the mandatory setup was connecting a calendar and setting defaults like 9 to 5. Udezue says wizards have had limited benefit.
  • Ania Wysocka, Rootd32:12: the intro slider is optional, and fewer people drop off before sign-up.
• Match the funnel style to your pricing and growth model. Nick Candler, Calm23:54
  • Long, high-friction quizzes suit products focused on monetization.
  • Finch uses fast onboarding, a low price and a top-of-funnel, share-with-friends strategy.
  • The creative and brand positioning should follow whichever approach you choose.

2.3 Make each step easy

A longer flow works only when every screen asks for less effort than the one before it.

• Optimize for effort per step, not for fewest clicks. Houzz
  • Replace one contact form with name, then email, then phone, one per screen.
  • Test the extra-step version rather than assuming fewer screens wins.
  • Houzz's stated reason: break down information entry to reduce cognitive load, despite the added clicks.
• Split any screen that has five or six inputs. Amol Avasare, Anthropic22:39
  • At Mercury, breaking such a screen into two performed well.
  • Adding steps also performed well there, as long as each step reduced load or helped the user.
• Keep high-effort asks away from the path to the paywall. Daniel Heintzman, Sway AI21:40
  • Sway tested letting users upload profile screenshots before the paywall.
  • Users could leave the app to find screenshots and never come back.
  • The simpler flow, with the aha moment after the paywall, performed a lot better. A cheap lead-up to the aha moment would be better still.
• Set defaults so users make fewer hard choices. Adam Fishman42:58
  • Patreon defaulted creators to three to five tiers and a lowest tier of three or five dollars.
  • Creators could still choose otherwise, with more friction. The aim is to make the wrong thing hard and the right thing easy, without removing choice.
  • The defaults came from what worked across hundreds of thousands of creators.

2.4 Add friction only where it buys quality

Extra effort pays when it filters, convinces or collects data you use, and it hurts when it only slows people down.

• Use deeper questions to filter for high intent. Simon Mizzi, Preply24:51
  • The friction should give the learner real utility, such as a better match.
  • Mizzi's travel background taught him that friction is bad, yet the conversion win was like nothing he had seen before.
• Turn a short checkout into a long intake when the product is personal. Ravi Mehta, Tinder59:57
  • Sesame's intake asks about goals and medical history.
  • Users reach payment already convinced, and the company gets data for personalization.
  • Mehta says it does not work in every category.
• Use the answers, not just the effort. Steve Young and Lucas, Paddle9:07
  • Find which segments convert best and change the marketing copy to match.
  • In a mental health app, anxiety relief converted best, so the copy shifted to relieving anxiety, a painkiller rather than a vitamin.
  • Rik Haandrikman, RevenueCat44:16: answers can go back to ad networks for better targeting, and web flows give deterministic data.
  • Patrick and Pablo, Social20:13: friction can create a stronger signal for ad campaigns.
• Cut friction where it adds nothing, and keep it where it raises quality. Ravi Mehta, Tinder6:07
  • Tinder used Facebook Connect for photos and light profiles, which widened the audience beyond people willing to spend 30 minutes on a profile.
  • It added friction elsewhere with double opt-in matching, so only mutual interest led to conversations.
  • Users could go from download to swiping in under five minutes, maybe under two.

2.5 Test big and measure more than conversion

Run a large length change both ways, and read retention next to conversion.

• A/B test the screen count, because there is no right number. Steve P. Young and Chaim Sajnovsky, App Masters3:56
  • Test the length both ways.
  • If your app shows only a paywall, add screens that explain the benefits.
  • One client added three benefit screens before the paywall and saw a 234 percent increase in overall conversions.
• Make a big jump in length, and fill it with personalization. Brett Bauman and Zack Hargett, Coconote29:05
  • Testing one or two screens at a time gave small wins but little big progress.
  • Coconote roughly doubled to about 13 screens, mostly personalization such as asking what the user is studying, plus permissions and account steps.
  • Coconote in another telling48:26 puts it at about 15 screens, with personalization and social proof, compounding on an onboarding they already thought was decent.
• Watch retention and conversion together. Adam Fishman45:37
  • Extra friction can weed out poor-fit users and improve retention even if fewer people finish.
  • Track opposing metrics so conversion does not fall further than retention can offset.
  • Fishman says the biggest impact of onboarding is on retention, not conversion.
• Redesign only when you have a new hard-earned insight. Adam Fishman49:05
  • A good trigger is something fundamentally new about customers or the growth model, like the right tier structure at Patreon.
  • Micro-tuning a decent foundation rarely moves the needle, and redesigning for its own sake wastes effort.

2.6 Where they disagree

Sources split on how many screens are right, and the split mostly follows the product and the goal.

• Three to five screens versus a dozen or more.
  • Steve Young and Lucas, Paddle9:07 call three to five screens a good number for most apps, and long flows right for fitness and education.
  • Coconote29:05 doubled to about 13 screens in a consumer app and gained, so the line between categories is not sharp. Test it.
• Effort before the paywall: help or harm.
  • Tomas Hoyos, Apt34:00: a quiz of roughly 15 to 20 minutes beat shorter versions, because invested users are curious about the result.
  • Daniel Heintzman, Sway AI21:40: a screenshot upload step lost to a simple flow with no required effort, because users might leave and not return.
• Fewer screens before sign-up versus more investment before the paywall.
  • Ania Wysocka, Rootd32:12 wants as few screens as possible before sign-up.
  • Rik Haandrikman, RevenueCat44:16 argues that time invested in the first session makes users more inclined to start a trial at the paywall.
• Maximize conversion or accept less of it.
  • Adam Fishman45:37 says conversion should sometimes decrease a little if retention improves.
  • Nick Candler, Calm23:54: quiz-heavy funnels fit heavy monetization, while a fast, low-price flow fits a top-of-funnel, sharing strategy like Finch's.