Short onboarding was never the point
On this page
Recently Mobbin published a video digging into onboarding screens. I was already poking around the same stuff on my own, so I figured why not turn the takeaways into a blog. Easier to skim later, and saves me from rewatching the whole thing when I need a reference.
A lot of advice says keep onboarding short. From the data in that video, the average app still has 25 onboarding screens. Longest categories: finance, health and fitness, education. Seven out of ten of those longest ones are finance apps.
Weird bit: some of the longest flows belong to the most successful apps. Shortest ones? Three of them are AI products.
So maybe short was never the goal.
Screens below are from Mobbin. They don't allow iframes (page is locked to same-origin), so each block is a row of real screens + a link to open the full flow on Mobbin if you want to click through it yourself.
The pattern that keeps showing up
Best flows usually go like this:
- Sign up
- Set up your account
- Hit the aha moment (when you actually feel the product working)
Airbnb: first booking. Netflix: find a show and press play. Mobbin: find a screen you like and save it.
So what are these apps doing on the way there?
Stop listing features. Sell the outcome.
Best screens I kept seeing are not feature lists. They sell the outcome.
Tiimo just shows the product in action. Calendar import, plan with AI, the whole setup.
Open Tiimo onboarding on Mobbin
Some apps do it with motion. Open the app and you already get what it does, no copy needed.
Alma goes further. You can try the core experience before signing up. Rare for AI products. Most lock everything behind an account.
Sometimes it is just copy. Superhuman turns a boring signup into a pitch, with logos on the side as proof.
Open Superhuman onboarding on Mobbin
Some apps skip the pitch and just feel human
One Year drops a founder note into onboarding. Handwritten signature, little flower drawing. Cute. Feels made by people.
Tinder notices when your birthday is close. Small human detail in a long signup.
Open Tinder onboarding on Mobbin
Airbnb (not even in onboarding) shows a video from the CEO when you list your first space. Founder touch at the aha moment.
Click to enlarge
Basecamp does a personal CEO note after account creation. Same vibe: this was built with intention.
Personalization that is actually worth your time
23% of apps personalize during onboarding. AI apps only 7%. Feels like AI tools skip the questions and just learn from usage later.
Headspace realized people show up with more than one problem. So they let you pick multiple goals. Small change, 10% lift in free trial conversion.
Open Headspace onboarding on Mobbin
FocusFlight lets you pick a map style. App already feels yours before you start focusing.
Open FocusFlight map style flow on Mobbin
Dollar Shave Club just made the quiz copy more conversational. That alone got a 5% bump in subscriptions. Here is that quiz vibe:
Dollar Shave Club quiz screens on Mobbin
Don't just collect answers. Show what they unlocked.
Some apps take the quiz answers and show you the payoff immediately.
Endel: a few questions, then "personalization complete." You have barely used it, but it already feels set up for you.
Open Endel onboarding on Mobbin
BitePal builds your personal plan and tells you the date you'll hit the goal.
Open BitePal onboarding on Mobbin
Speak asks language + goals, then one screen: in two months you can communicate while traveling in France. Simple graph. Speaking vs reading. And the steps before that already had you speaking out loud.
Open Speak onboarding on Mobbin
Paywalls during onboarding
Out of 900+ apps and sites, 22% put a paywall in onboarding.
Some pair it with personalization. Beside: quiz + one-time offer. Tiimo: social proof first, then the wall. FocusFlight: the offer looks like a boarding pass. Phone vibe and all. Weirdly fun for a paywall.
Grammarly recommends pricing based on quiz answers. Almost 20% more plan upgrades.
Grammarly plan screens on Mobbin
Long does not have to feel long
Duolingo is one of the longest. Flow is basically: start → pick language → learn about you → first lesson → feel the win → then create an account.
Dozens of screens before signup. Does not feel like dozens.
Open Duolingo onboarding on Mobbin
Bump keeps something happening the whole time. Even verification loading gets a fun treatment.
BitePal has 61 screens. You name the raccoon. They keep reminding you of the value. Then the paywall.
Don't dump all the education up front
Cake Equity takes dry equity stuff and walks you through it with calm copy and tooltips. Password field that checks rules as you type. Tiny stuff, but you get stuck less.
Open Cake Equity onboarding on Mobbin
Todoist skips the blank empty state. You land in a getting-started list. No tour. No popup. Just a nudge.
Open Todoist onboarding on Mobbin
Mural swapped popups for a six-step checklist and got a 10% relative bump in one-week retention. Checklist sticks around after the first flow ends.
Open Mural onboarding on Mobbin
Ask for notifications properly
A lot of apps show a custom screen before the system popup. Accept rates go up.
Brilliant: "I'll remind you to learn so it becomes a long-term habit."
Open Brilliant onboarding on Mobbin
Some apps even preview the notification you will get if you allow it. Same idea, different apps:
Atoms · Aaptiv · Tempo · Duolingo Math
Also helps explain why web onboarding is 21% shorter than iOS. Mobile has more permission and paywall screens baked in.
Friction is not always bad
House split their signup form across multiple screens and saw 15% more conversions. Friction in one place can remove friction somewhere else.
Culture matters too. Heavier UIs feel fine in some markets and cluttered in others. So copying a "winning" flow blindly does not always work.
Do we even need onboarding?
No single answer.
The flows that stuck with me did not feel like onboarding. They got people to value fast.
Sometimes that means making a long flow fun. Sometimes personalization. Sometimes just getting out of the way.
Mobbin itself is a good example. Product speaks for itself. Same for a lot of AI chat apps. First prompt is the aha. For those, maybe the best onboarding is almost no onboarding.
Kind of comes down to the product.
Used LLMs to correct grammar, typos etc 🤖