Onboarding Funnel for Mobile Apps: Stages and Events
An onboarding funnel for mobile apps is the sequence of steps that takes a new user from first touch (an ad, a store page, a link) to the moment they get value from the product and, for subscription apps, start paying. To build one, split the path into stages, define one event per meaningful action, and measure how many users move from each stage to the next. Optimization then means finding the step with the largest avoidable loss and testing one change at a time.
This guide covers the stages, the events to track at each stage, where drop-off tends to hide, and how to optimize. It applies to in-app onboarding and to a web onboarding funnel, where the quiz, paywall and payment happen in the browser before the user installs the app.
The stages of an onboarding funnel
Names differ between teams, but most subscription app funnels have the same stages:
- Acquisition: the user arrives from an ad, a creator link, search or a store listing.
- First screen: the first thing they see, a landing page, a splash screen or the first quiz question.
- Qualification: questions about goals, level, habits or constraints. This is where personalization data is collected.
- Plan or result: a personalized plan, score or preview that shows the user what they will get.
- Paywall and payment: the offer, plan selection and checkout.
- Activation: the user installs or opens the app, logs in and completes the first action that predicts retention.
Think of a funnel as a chain of conversion rates, not one overall number. A single end-to-end rate tells you something is wrong. It does not tell you where. Each stage needs its own entry event and exit event so you can compute the step conversion: users who reached step N+1 divided by users who reached step N.
What events to track for an onboarding funnel
Track fewer events than you think you need, but name them consistently and attach the same properties to all of them. A workable minimum looks like this:
Acquisition and first screen
funnel_vieworapp_first_open, with source, medium, campaign, creative ID and country.onboarding_start, fired on the first interaction, not on page load. The gap between view and start is your bounce measure.
Qualification
question_viewandquestion_answered, with question ID and position. Per-question events show which question loses people.onboarding_complete, fired when the last question is answered.
Plan and paywall
plan_viewfor the personalized result screen.paywall_view, with paywall version and offer shown.plan_selected, with the product and price point.
Payment
checkout_startedandpayment_success, with payment method.payment_failed, with the failure reason. Without it, a payment problem looks like lack of interest.
Activation
app_installorapp_first_openafter purchase.login_success, which links the funnel user to the app account.- One activation event you choose yourself: the first workout, first lesson, first generated result. Pick the action that your own retention data shows separates users who stay from users who leave.
Give every user a stable ID early and carry it across steps, devices and the web-to-app handoff. Send the key events to your analytics tool and, for paid acquisition, to the ad platforms as server-side events, so campaign optimization sees purchases and not only clicks. Our analytics guide goes deeper on setup.
Where drop-off hides
The obvious drop-offs, such as a long quiz or a high price, are easy to spot. These are the ones teams usually miss:
- The pre-start gap. Users who load the first screen but never interact. Slow load times, content that renders badly inside social in-app browsers, and a mismatch between the ad promise and the first screen all show up here.
- One bad question. Average completion hides a single question that asks for too much too early, such as an email, a body measurement or a sensitive answer. Compare the per-question exit rate.
- Paywall view without checkout. This usually means price, plan layout or trust issues, not interest. Compare by traffic source before concluding anything.
- Checkout started without payment. Failed cards, unsupported local payment methods, and extra friction in the payment form. Break it down by country, device and method.
- Paid but never activated. In a web onboarding funnel the user pays in the browser, then must install the app and log in. If the link, the store redirect or the login step is unclear, you lose people who already paid. This is the most expensive leak and the one most often unmeasured.
- Attribution gaps. If the purchase is not tied back to the campaign, you will cut campaigns that work. Check that source parameters survive every step.
Always segment before you decide. The same funnel can convert well for one source and badly for another, and a blended average hides it.
How to optimize an onboarding funnel
Funnel optimization is mostly discipline, not clever ideas. A sequence that works:
- Find the biggest loss. Compute step conversion for every stage over a period long enough to include a full week, and rank steps by how many users you lose, not by percentage alone.
- Segment it. Split by source, creative, country and device. Fix the segment with the most spend first.
- Form one hypothesis. For example, "moving the email request after the plan preview will reduce exits at that question."
- Test one change. Run it as a split test against the current version and wait until you have enough users per variant to see a real difference. Decide the success metric before you start. See our notes on A/B testing funnels.
- Judge on downstream results. A shorter quiz may raise completion and lower purchase intent. Measure paid conversion and retention, not only the step you changed.
- Keep a log. Record what you changed, why and what happened, so the team does not repeat old tests.
Common levers, roughly in order of effort: first-screen message match to the ad, question order and count, the personalized result screen, paywall layout and offer, payment methods, and the post-payment handoff to the app. Localization is a lever too: translating the whole funnel, not just the paywall, tends to matter in new markets.
In-app vs web onboarding funnel
Many subscription apps now run the onboarding and payment in the browser and only then send the user to the app. These are web2app funnels. Neither approach is always better, so here is how they compare on the factors that matter.
Iteration speed
- In-app onboarding: changes usually go through an app release and store review, unless you use remote configuration.
- Web onboarding funnel: changes are published on the web, so tests can run without a release.
Payments
- In-app onboarding: purchases go through the app store billing system, which charges a commission (Apple charges up to 30%, or 15% under its Small Business Program).
- Web onboarding funnel: payment runs through a web payment provider, where store rules allow it. Fees and rules differ by provider and region.
Measurement
- In-app onboarding: events are in one place, but attribution from ads depends on the platform's mobile tracking.
- Web onboarding funnel: you control the full path from ad click to payment, but you must carry the user identity into the app to link everything.
Risk and rules
- In-app onboarding: fewer policy questions, as everything stays inside the store model.
- Web onboarding funnel: store policy on external purchases varies by region and changes over time. Check the current App Store Review Guidelines and Google Play policy before you launch, and read our web2app funnel compliance guide. Nothing here is legal advice.
Activation
- In-app onboarding: the user is already in the app when onboarding ends.
- Web onboarding funnel: you add a handoff step (install and log in) that needs its own tracking and a clear instruction screen.
Tools for building and measuring a web onboarding funnel
You can build a web funnel with a developer and your own stack: a front end, a payment provider, an analytics tool and a way to pass the user ID to the app. This gives full control and costs engineering time for every change.
The alternative is a no-code builder. web2wave is one option for web2app funnels: you assemble the quiz or onboarding flow, personalized plan and web paywall visually, run A/B tests, track UTM and attribution, and send events to analytics tools and ad pixels with server events. It also connects to subscription tools such as RevenueCat, Adapty, Qonversion, Superwall and Apphud, and supports AI localization into 40+ languages. The quiz builder page shows what the flow editor covers. If you want to compare builders first, see the builder comparison guide.
Whichever route you take, the checklist is the same: events named and documented before launch, user identity carried into the app, payment failures logged with reasons, and one owner for the funnel who reviews step conversion every week.
FAQ
What is an onboarding funnel for mobile apps?
It is the series of steps from first contact to activation, and for subscription apps to the first payment. Each step has an entry and exit event so you can measure how many users continue.
What events should I track for an onboarding funnel?
At minimum: first view, onboarding start, each question answered, onboarding complete, paywall view, plan selected, checkout started, payment success, payment failed with reason, app install, login, and one activation event that predicts retention in your own data.
How do I find where users drop off?
Compute step conversion (users reaching the next step divided by users reaching the current step) for every stage, then segment by source, country and device. Look at per-question exits and at the gap between payment and first login.
What is a web onboarding funnel?
It is an onboarding flow that runs in the browser: quiz, personalized plan, paywall and payment. After paying, the user installs the app and logs in. Check store policy for your regions before using it.
How often should I test changes to the funnel?
Test one change at a time, and only as fast as your traffic allows you to reach a clear result per variant. Running several overlapping tests on the same step makes the results hard to read.
Want to see how this works on your own funnel? Book a demo or start free.
.webp)
