
App onboarding is the first experience a user has with your mobile app: the sequence of screens, questions, tutorials, and interactions that takes them from "just installed" to "understood the core value." Done well, it sets up every future session. Done poorly, it's the biggest single reason users uninstall on day one. Industry data from Localytics and AppsFlyer consistently puts day-one uninstall rates in the 20-25% range for mobile apps, and the single most predictive variable is whether the user completed a meaningful first action during onboarding. A 2024 Adjust mobile app trends report put average day-30 retention for consumer apps at roughly 5%, and the gap between top-quartile and bottom-quartile performers is almost entirely explained by first-session behavior.
I review onboarding flows constantly, both as a reference when designing new ones and because session replays from UXCam's customer base show me how real users actually interact with them. The 12 apps below represent the best patterns I've seen, with specific callouts on what each one does right. None of them are perfect, but each gets something right that the others can learn from. The onboarding screen, onboarding page, and in-app onboarding patterns across these 12 examples cover most of the decisions a product team will face when designing their own flow.
Good onboarding is short (3-5 screens, not 10), personalized (based on 1-2 quick questions), and punctuated by a meaningful first action so users feel value before they're asked for commitment.
The biggest mistake I see: asking for permissions, account creation, or payment details before the user understands what the app does. Every barrier before value destroys completion rates.
Progress indicators matter. Users who see "step 2 of 4" complete at materially higher rates than users in a seemingly endless flow.
The best onboarding flows get users to a first meaningful action (logged workout, created playlist, sent first message) within the first 60 seconds of the first session.
Measure activation rate (percentage of new users who hit your defined activation event) weekly. Watch session replays of users who don't activate to find the specific friction. That loop is how onboarding improves over time.
Wise (formerly TransferWise)

What works: Instagram's onboarding is famously short. A username, a password, a bit of profile info, and you're in. No tutorial walls. No "let us show you around." The app assumes (correctly) that anyone installing it has seen Instagram before and knows what it's for.
The key pattern: minimum required information to start. Each additional field in the signup form drops the completion rate. Instagram asks for what's necessary to create an account and defers everything else to in-context discovery.
Where teams copy it wrong: Instagram's model works because the app has massive brand recognition. Users know what they're signing up for. For lesser-known apps, skipping value explanation entirely is a mistake. You need some lightweight "here's what this app does" context.


What works: SoundCloud's onboarding is a masterclass in getting users to the "aha moment" quickly. Within the first minute, users are listening to music. The app asks them to pick genres they like, then immediately serves a customized feed of tracks.
The key pattern: personalization leading to first meaningful action leading to aha moment, all within a single onboarding session. The "first meaningful action" is listening to music, which is why people installed the app in the first place.



What works: LinkedIn's onboarding turns a chore (building a professional profile) into a progressive series of small wins. Each step produces something visible: a completed section, a connection suggestion, a network size that grows as you confirm contacts.
The key pattern: progressive disclosure with visible progress. LinkedIn could ask for your entire work history on one screen; instead it splits the task into pieces and shows progress after each. Users feel like they're accomplishing something rather than enduring a form.
Where teams copy it wrong: the progressive approach only works if the user has motivation to continue. For apps with weaker intrinsic motivation (utilities, transactional apps), forcing 5-6 steps of onboarding before the core feature is backwards.

What works: Wise (formerly TransferWise) handles one of the hardest onboarding problems: identity verification (KYC) required for financial services. They break it into clearly-explained steps, show estimated time at each stage, and let users save and resume.
The key pattern: transparent expectation-setting for long-form compliance flows. Telling users "this step takes 3 minutes" and "you can save and come back later" dramatically reduces abandonment in regulated contexts.
What teams can learn: if your onboarding has required friction (identity verification, payment setup, compliance), don't hide it. Show users what's coming and let them plan around it.

What works: MyFitnessPal's onboarding asks a series of personal questions (goal, current weight, activity level) and uses them to calculate a personalized daily calorie target before the user has done anything else. The personalization pays off immediately.
The key pattern: questions that produce a personalized output. Users are willing to answer 5-6 questions if the output is obviously personalized to them. They're not willing to answer 5-6 questions if the result is generic.
What fails: apps that ask personalization questions but then show the same home screen to everyone. If you're going to ask, the answers should visibly shape the experience.

What works: Binance handles crypto-specific onboarding (KYC, 2FA setup, funding) with a dashboard that shows what's done and what's pending. Users can complete the setup piece by piece across sessions.
The key pattern: multi-session onboarding for complex products. Not every app needs to be fully onboarded in a single sitting. Some products are complex enough that breaking setup across multiple sessions (with clear progress tracking) works better than trying to cram it into one flow.

What works: Strava gets users to their first recorded activity fast. The onboarding skips secondary features (social feed, clubs, challenges) and focuses on the core loop: record an activity, see it, share it.
The key pattern: prioritize the single most important action. Strava could surface 10 features during onboarding. Instead it surfaces one: start recording. Everything else is discovered later through natural use.
What works: Tinder's onboarding is famously efficient. Upload a few photos, answer a handful of questions, start swiping. The time from install to first interaction is often under 2 minutes.
The key pattern: remove every step that isn't strictly necessary for the core action. Tinder doesn't need your professional background or your life story to let you start swiping, so it doesn't ask.
What teams can learn: audit your onboarding screens ruthlessly. For each one, ask: "does the user need this to take the first meaningful action?" If not, defer it.

What works: Expensify's onboarding is transactional (expense management is a chore, not a delight) and embraces that. It asks for the minimum info needed, explains what will happen next, and gets the user to the "scan a receipt" action quickly.
The key pattern: match the tone of the onboarding to the tone of the product. For utility apps, efficiency beats delight. For lifestyle apps, vice versa.
What works: Duolingo's onboarding is famous for starting the user in their first lesson almost immediately. Within 30 seconds of opening the app, users are answering their first question. Account creation is deferred until after the first meaningful action.
The key pattern: inverted onboarding. Most apps do "sign up, then use the product." Duolingo does "use the product, then sign up." For categories where users are naturally curious, the inversion dramatically lifts early engagement. Duolingo's team has publicly shared on their design blog that this inversion was one of the single highest-impact retention changes they've made.
What teams can learn: if your app lets users try the core experience without logging in, consider inverting the flow. Users are much more willing to create accounts after they've experienced value.
What works: Headspace onboarding uses narrative framing. Users are greeted with "welcome" screens that set expectations, then guided through a first meditation session (typically 3-5 minutes). The calm tone of the onboarding matches the product's entire value proposition.
The key pattern: onboarding as a preview of the product experience. If Headspace's onboarding was fast, transactional, and noisy, it would contradict what the app is for. The onboarding is the product's promise in miniature.
What works: Notion's onboarding starts users with a pre-populated template workspace rather than an empty canvas. This avoids the "blank page problem" that kills productivity-app retention. Users see what the product can do by interacting with an already-functional workspace.
The key pattern: demo-like first session. For powerful tools with a learning curve, a pre-populated example beats a blank starting state. Users learn by modifying something that already works.
Pulled from hundreds of session replays and funnel reviews, these are the specific patterns that separate onboarding flows that work from flows that leak users at every step. Treat this as a checklist against your current flow.
Show users what the app does before asking them to register. Duolingo, TikTok, and Pinterest all let users sample content before committing. The data from Appcues and similar sources consistently shows deferred signup lifts activation by 10-30%.
Ask for one or two pieces of information at a time, spread across sessions. LinkedIn does this by nudging profile completion over weeks. The alternative, a 15-field signup, is a proven drop-off generator documented in Nielsen Norman Group form research.
Offering Apple Sign In and Google sign-on shaves seconds off the critical first screen. For consumer apps, social sign-on conversion runs 2-3x higher than email-and-password flows.
Every screen a new user can reach needs an empty state that suggests what to do next. Notion's template gallery and Figma's starter files are the canonical examples.
Ask for push, location, and contacts only after the user understands why. iOS's permission priming pattern (a custom screen that explains the value before the system prompt fires) lifts opt-in rates by 20-40% in my experience.
Users in a "step 3 of 5" flow complete at higher rates than users in an unbounded flow. This is well-documented in work from the Baymard Institute on checkout usability, and the same principle applies to onboarding.
Onboarding flows that don't let users go back and change answers feel punishing. Always allow edit-in-place, especially for personalization questions that shape the rest of the experience.
A tutorial wall (five swipeable screens explaining the app) is skipped by most users. Tooltips that appear in context when the user reaches a feature are read. Tools like Pendo and Appcues exist to build these without shipping code.
Pre-fill sensible defaults (currency based on locale, units based on region, notification preferences based on category norms) rather than making every user decide everything. Defaults are a form of onboarding.
When the user completes their first meaningful action, acknowledge it. Strava's kudos animation, Duolingo's XP explosion, and Headspace's session completion screen all reinforce that the user just did something worthwhile.
One or two questions about intent (weight loss vs. muscle gain, leisure vs. business travel) let the product tailor the entire experience. Calm and MyFitnessPal both do this well.
Showing a paywall before the user has experienced value kills conversion. The pattern across RevenueCat benchmark data is clear: apps that let users feel the product before the paywall convert at higher rates.
If a step isn't strictly required for the core experience, offer a visible "skip for now" option. Forced steps generate the worst abandonment.
For products like Wise, Notion, and Spotify that span web and mobile, the onboarding should remember progress across platforms. A user who started signup on the web shouldn't start over on mobile.
For e-commerce and marketplace apps, the post-onboarding home screen with zero items in cart, zero saved searches, zero recommendations is a silent killer. Seed the experience with recommendations based on onboarding answers.
The right onboarding flow depends heavily on the category. Here's how the principles shift across verticals.
Regulated flows require KYC, identity verification, and disclosure screens. The trick is separating regulated steps from optional ones, then completing the regulated minimum before showing the product. Wise, Revolut, and Chime all handle this by explicitly labeling mandatory vs. optional steps. Expect 40-60% completion rates in first session for regulated flows, versus 70-85% for unregulated consumer apps. A pattern I recommend is a "dashboard of pending tasks" that lets users chip away at verification between product sessions rather than forcing it all up front.
Personalization is the whole game. Users arriving at MyFitnessPal, Strava, or Whoop expect the app to know their goals, body metrics, and context before it starts giving advice. 5-7 questions is acceptable here because each answer visibly shapes the plan the user receives. The companion principle is privacy framing: health apps that explain why each question is needed convert at meaningfully higher rates than apps that just ask.
The first session should end at a product, not an empty home screen. Wayfair, Etsy, and Depop all use category-picker onboarding to seed the home feed with relevant items. The activation event is almost always "viewed first product" or "added to wishlist," not signup. Guest checkout should be a first-class option, because forcing account creation before purchase is one of the most studied conversion killers in Baymard's research.
Team invites, workspace setup, and integrations dominate the flow. The best B2B flows (Notion, Linear, Figma) start individuals in a template workspace and defer team setup. This lets the inviting user experience the product before turning it into a team setup chore. Activation rates for B2B run lower than consumer because the activation event often requires a teammate, so measuring "invite sent" as a leading indicator matters.
First-session engagement trumps information collection. Supercell games, Netflix, and Spotify all drop users into content within 30 seconds. Account creation and personalization get layered in after the user is hooked. The industry standard in games is the "first 30 seconds of gameplay" metric, and the flows that hit it consistently outperform flows that front-load tutorials.
Location permissions and payment setup are the critical barriers. Uber, Lyft, and Airbnb use progressive flows that let users browse before they book, deferring payment until it's actually needed for the first transaction. The pattern to copy: let users search and see results before any account or payment step, then require auth only at the conversion moment.
Building onboarding flows from scratch is rarely the right call. The tooling landscape covers analytics, in-app guidance, A/B testing, and deep-link attribution.
Behavioral analytics and session replay. UXCam for mobile apps and web with session replay, heatmaps, and funnels. Amplitude and Mixpanel for event-based funnel tracking. Heap for autocapture-style event analytics on web.
In-app guidance and tooltips. Appcues, Pendo, and Chameleon for no-code tooltips, checklists, and walkthroughs. UserGuiding and Userflow for onboarding checklists specifically.
A/B testing. Statsig, Optimizely, and LaunchDarkly for flow experimentation. Most teams underestimate how many onboarding A/B tests they should be running at any given time.
Attribution and deep linking. Branch, AppsFlyer, and Adjust for connecting acquisition channel to activation. Critical if you're running paid acquisition, because different channels activate at very different rates.
Customer messaging. Braze, Iterable, and Customer.io for onboarding email and push sequences that re-engage users who drop off.
Subscription and paywall optimization. RevenueCat, Adapty, and Superwall for paywall A/B testing and subscription analytics, which become critical once you start moving paywalls past the aha moment.
These are the failure patterns I see in UXCam session replays across categories. Most onboarding improvements come from removing one of these, not from grand redesigns.
Asking for push notification permission on screen one. Users have no context yet. iOS opt-in rates collapse below 30% when the prompt fires too early.
Mandatory account creation before value. Every forced signup wall is a leak point. Defer where legally possible.
Tutorial walls with 5+ swipe screens. Users skip these. The information is wasted.
Asking personalization questions that don't change anything. If the post-onboarding experience is identical for every user, the questions are decorative and annoying.
Hiding the skip button or not offering one. Users feel trapped and abandon entirely rather than pushing through.
Over-reliance on modal popups on the home screen. The first session gets interrupted by three tooltips, a promo, a permission prompt, and a rating request. Pick one.
Forgetting the empty state. A home screen with zero content and no guidance tells the user "there's nothing here for you."
Paywalls before the aha moment. Users who haven't experienced value won't pay for it. RevenueCat's subscription benchmarks confirm this across categories.
Not measuring drop-off at each step. Without a funnel, you're guessing at where the flow breaks. Every onboarding flow needs step-level instrumentation.
Redesigning the whole flow instead of fixing one step. The highest-impact wins are almost always "fix step 3," not "rebuild everything." Watch session replays, find the step, ship the fix.
Most teams I work with fall into one of four maturity tiers. Knowing where you are tells you what to do next.
You've built an onboarding flow but you don't know where users drop off, what percentage activate, or what the aha moment even is. Next step: instrument the funnel. Install UXCam or an event-based analytics tool, define the activation event (the single most predictive action for day-7 retention), and start measuring activation rate weekly.
You have a funnel and you can see drop-off rates, but you haven't watched sessions or formed hypotheses about why users leave. Next step: pair funnels with session replay. For the step with the highest drop-off, watch 20 sessions of users who bailed. Patterns will emerge within the first 10.
You run 1-2 onboarding A/B tests per quarter and have a documented hypothesis backlog. Next step: increase experiment velocity. The best teams run continuous onboarding experiments, 4-6 per quarter minimum, with clearly defined primary metrics. This is where Recora achieved a 142% increase in activation and Inspire Fitness saw 460% growth in a key engagement metric by treating onboarding as an always-on optimization target.
You have different onboarding flows for different user segments (acquisition channel, geography, intent), and activation rates are tracked per segment. Next step: use AI analysis. Tools like Tara can surface segment-level friction patterns that manual analysis misses, and the insights compound over time.
If you're starting fresh, here's the playbook I recommend. Week one: instrument the funnel and pick an activation event. Week two: watch 50 session replays of users who didn't activate, categorize the friction into three to five themes. Week three: ship the single highest-impact fix and start a 2-week A/B test. Week four: review results, queue the next experiment, and start the loop again. Most teams see a 15-30% lift in activation within the first 60 days of running this playbook consistently.
The patterns across these 12 examples cluster into five principles:
Minimum required to start. Every field in your signup form drops completion. Ask only for what you truly need before the first meaningful action.
First meaningful action within 60 seconds. Get users to the core value (log a workout, create a playlist, send a message, complete a lesson) as fast as possible.
Personalization that visibly shapes the experience. Ask 1-3 questions if the answers materially change what the user sees next. Don't ask if they don't.
Visible progress in multi-step flows. Progress indicators reduce abandonment measurably. Users in "step 2 of 4" complete at higher rates than users in "step 2."
Match tone to product. Utility apps can be efficient; lifestyle apps can be delightful; financial apps need to be trustworthy. The onboarding sets the expectation for everything that follows.
App onboarding is the series of screens, questions, and interactions that introduces a new user to a mobile app and gets them to their first meaningful action. It typically includes account creation, permission requests, personalization questions, and a first task completion. Good onboarding is short, personalized, and focuses on getting to value quickly.
First-session experience predicts long-term retention more reliably than any other variable. Users who complete a meaningful action on day one retain at 2-3x the rate of users who open the app, scroll, and close. Getting onboarding right is the highest-leverage retention investment most teams can make.
For most apps, yes. Exceptions: apps with massive brand recognition (Instagram, WhatsApp) where users already know the core value, and utility apps where the core action is obvious from the home screen. Even these often benefit from light personalization or permission flows.
A mobile onboarding flow is the ordered sequence of screens a user sees from install to first meaningful action. It usually covers account creation, permission requests, personalization, and an initial task. The specific sequence varies by app, but the best flows I've seen share patterns: short, personalized, and punctuated by a first action.
Keep it under 5 screens for most apps
Ask for permissions (push, location, contacts) only after showing value
Use a progress indicator so users see how far they've come
Defer account creation where possible (let users try the product first)
Ask personalization questions only if the answers change the experience
Watch session replays of users who drop off during onboarding to find specific friction
UXCam is a product intelligence and product analytics platform that automatically captures every user interaction on mobile apps and websites, no manual event tagging. Funnel Analytics shows you the exact step where users drop off during onboarding, and session replay shows you what they were actually experiencing when they bailed. Tara, UXCam's AI analyst, processes sessions to surface the specific friction patterns causing onboarding drop-off and recommends actions.
Costa Coffee increased registrations by 15% after UXCam revealed a password-error issue in their loyalty signup. Housing.com grew feature adoption from 20% to 40% by watching 50-100 onboarding sessions daily. Recora lifted a core activation metric by 142%, and Inspire Fitness saw 460% growth after rebuilding their onboarding around behavioral data. These are the kinds of wins that show up when you pair onboarding funnels with behavioral observation.
Installed in 37,000+ products, working across mobile apps and the web. Request a demo to see it for your app.
Frequently asked questions
A good user onboarding flow is short (3-5 screens), personalized (based on 1-2 questions that visibly change the experience), and gets the user to a first meaningful action within 60 seconds. It avoids asking for permissions, payment, or commitment before the user understands the value. It uses progress indicators in multi-step flows to reduce abandonment.
App onboarding is the sequence of screens and interactions that introduces a new user to a mobile app from install to first meaningful action. It typically includes account creation, permission requests, personalization, and an initial task. The goal is to get users to the "aha moment" (where they understand the app's value) as quickly as possible.
Audit your current flow with a funnel analytics tool like UXCam to find the step with the highest drop-off. Watch 10-20 session replays of users who drop off at that step. Form a specific hypothesis about the cause. Ship a targeted fix. Re-measure over 2-4 weeks. Most onboarding wins come from removing one specific friction point, not from redesigning the whole flow.
Follow the principles: minimum required information before first action, personalization that visibly shapes the experience, progress indicators in multi-step flows, first meaningful action within 60 seconds, tone that matches the product. Study the 12 examples above for specific patterns you can apply to your category.
For most apps, yes. The exceptions are apps with massive brand recognition (Instagram, WhatsApp) where users already know what to expect, and simple utility apps where the core action is obvious. Most apps benefit from at least lightweight onboarding (2-3 screens) to set up personalization and permissions.
Depends on your category. For lifestyle and consumer apps, SoundCloud and Duolingo are gold standards. For professional apps, LinkedIn. For compliance-heavy flows, Wise. For utility apps, Expensify. For productivity tools, Notion. Study 3-4 examples that match your category rather than trying to copy a single "best" example.
An onboarding screen is one of the individual screens in an onboarding flow. A typical flow has 3-5 onboarding screens covering value proposition, personalization questions, permission requests, and the first meaningful action. Each screen should have one clear purpose; screens that try to do too much drop completion rates.
Track activation rate: the percentage of new users who complete your defined activation event (logged workout, first message, etc.) in session one. Healthy activation rates range from 30-50% depending on category. Segment by acquisition channel, because paid users typically activate at lower rates than organic. Watch session replays of users who don't activate to find specific friction points.
For most consumer apps, 3-5 screens completed in under 90 seconds. Regulated products (fintech, healthtech) require longer flows, often 5-10 minutes, and the key is to break them into savable sessions with clear progress. The goal is never "short for its own sake." It's "short enough that the user reaches value before giving up."
Contextual tooltips, almost always. Tutorial walls (the 4-5 swipeable intro screens) get skipped by the majority of users. Tooltips that appear when the user reaches a relevant feature have much higher read and retention rates. Tools like Appcues, Pendo, and Chameleon make contextual guidance easy to ship.
After the user has experienced value, not before. A first-session permission prompt with no context typically converts at under 30% on iOS. A prompt that fires after the user has completed a meaningful action and with a clear value statement ("get notified when your package arrives") can convert at 60-70%.
Detect the reinstall using attribution data from Branch or AppsFlyer and offer a shortened re-onboarding that restores their previous state. Forcing returning users through the full new-user flow is a common oversight that drives immediate re-abandonment.
30-50% for consumer apps is typical, measured as the percentage of new users who complete your defined activation event in their first session. B2B SaaS is often lower (15-30%) because activation usually requires team setup. Track it weekly and segment by acquisition channel. The absolute number matters less than whether it's moving in the right direction over time.
Mostly the same in structure, but platform conventions matter. iOS users expect Apple Sign In and iOS-native permission prompts. Android users expect Google sign-on and Material Design patterns. Don't ship a web-wrapped flow on either platform, because it reads as low-effort and hurts trust from the first screen.
Web flows can be longer because users are often at a desk with less time pressure and no small screen constraint. Mobile flows need to be tighter, with larger tap targets and fewer fields per screen. For products that span both (Notion, Spotify, Wise), the critical capability is syncing progress across platforms so users don't re-onboard when they switch devices. UXCam covers both mobile apps and the web so you can measure and optimize each flow in one place.
Silvanus Alt, PhD, is the Co-Founder & CEO of UXCam and a expert in AI-powered product intelligence. Trained at the Max Planck Institute for the Physics of Complex Systems, he built Tara, the AI Product Analyst that not only analyzes user behavior but recommends clear next steps for better products.
Métricas de customer experience, las 12 que vale la pena monitorear, fórmulas, benchmarks, agrupaciones de percepción vs. comportamiento...
Founder & CEO | UXCam
Métricas de customer experience, as 12 que vale a pena acompanhar, fórmulas, benchmarks, agrupamentos por percepção, comportamento e operação, e como a...
Founder & CEO | UXCam
Customer experience metrics — the 12 worth tracking, formulas, benchmarks, perception vs behavioral vs operational groupings, and how AI session analysis...
Founder & CEO | UXCam
