Cut onboarding from 30+ quiz screens to 11 steps built around a first real entry, and doubled the share of users reaching the paywall (34% → 69%).
Daytiko is a private journaling and self-care app for iOS and Android. People write, record voice or video, check in on their mood, breathe, and burn what weighs on them, all behind a PIN. It runs on a subscription with a 7-day free trial.
The old onboarding was a long quiz: about 20 questions about feelings, habits and goals, then science claims, a self-care plan, a reminder picker, reviews, a theme picker and finally a 50% off paywall. Nobody used the product before being asked to pay.
What the data said: Over 28 days, 418 people opened the app and only 143 reached the paywall (34.2%). 65.8% left before they ever saw it. The problem wasn't the paywall, it was everything before it.
I led the onboarding redesign end to end as the Senior Product Designer. I audited the old flow against the client's analytics, rebuilt it around a first real entry, and designed every screen and state in Figma from the component library. I defined an analytics event for each step, so the new funnel could be read screen by screen, and handed it off to the founder and developer.
Why: The old flow asked about 20 questions and never showed the product. Now people write a real first entry inside onboarding, with text, voice or video, stored on their device only. All three lead to the same saved card.
Trade-off: The hardest step to design, with nine states. And only a real entry starts the streak, not the sample.
Why: Goals and experience shape the space, and the button stays off until something is chosen. Almost half the people who answered were brand new to journaling, so newcomers see a short primer before they write.
Trade-off: Less personalisation than 20 questions, but both can be skipped and people are writing by step 5.
Why: Each commitment explains itself: why it works (a cited study), when it happens (a daily reminder) and who can see it (a PIN). Each one has a clear exit: Skip for now, Maybe later, Don't allow.
Trade-off: No forced opt-ins, and people still said yes: 81% set a PIN and 66% set a reminder.
Why: Real user reviews land right before the ask. The paywall explains the trial in three steps: everything unlocks today, a reminder comes on day 5, and the trial ends on day 7, so cancelling before then costs nothing.
Trade-off: No discount banner and no countdown, so the ask relies on trust instead of urgency.
The new onboarding was measured from Aug 1 to Sep 30, 2026, for new installs in the US, Canada and the UK. The old numbers come from the client's report for May 7 to Jun 3, 2026 (28 days, app opens, all app versions). Each sample is a few hundred people and the two aren't perfectly like-for-like, so read the comparison as a direction, not a precise lift.
"I started the free week just to look around. That was six months and 180 entries ago." (Olena K., one of the reviews on the social-proof screen)
The new flow fixed reach and activation: more than two thirds of people now get to the paywall, and most of them have already written a real entry by then. They arrive knowing what they'd be paying for.
The new onboarding is live. It's about a third of the length, it puts the product first, and twice the share of people reach the paywall.
Onboarding is the first promise a subscription app makes. The old one asked for a lot and gave nothing back. The new one lets people feel the product first, then asks honestly, with a way out every time. Twice the share of people now reach the paywall, most of them with a real entry already written, and with every step measured, the team can see exactly where people stay and where they leave.