Users were signing up but not sticking. An 8-week 0→1 build that turned activation from a problem into a strength.
The app had healthy top-of-funnel numbers. Downloads were strong. Sign-ups were converting at a reasonable rate. But the D7 retention curve told a different story — most users who signed up weren't coming back after day one.
The core hypothesis from the growth team: users weren't finding their "aha moment" fast enough. But nobody had tested what that moment should be.
I was brought in to own the 0→1 feature build from problem definition to launch. Team of 6: myself, 2 engineers, 1 designer, 1 data analyst, and 1 QA. Timeline: 8 weeks to an MVP that we could measure against.
I started with cohort analysis. Users who returned on Day 7 had one behaviour in common in their first session: they'd completed a specific personalisation step that the majority of users skipped. The skip rate was 73%.
Qualitative interviews confirmed it — users who skipped it didn't understand why it mattered. The step felt optional and abstract. For users who completed it, the app immediately felt relevant to them.
The insight: The aha moment already existed in the product. The problem was that we were making it too easy to skip. The solution wasn't to build something new — it was to make the existing value unavoidable.
A personalisation feature that surfaced during onboarding and made the app's value proposition immediately personal. Key design decisions:
Figma design
In Figma: Share → Copy link → paste into src above.
Change embed_host=share&url= to your link.
The personalisation onboarding flow. Key screens: entry prompt, real-time preview as user makes choices, and the completion state showing their personalised home view.
8 weeks is tight for a 0→1 feature. I kept scope ruthlessly narrow: we were solving one problem (low personalisation completion during onboarding) and measuring one thing (D7 retention for users who completed the new flow vs. those who didn't).
We ran a 3-week beta with 15% of new sign-ups before full rollout. The D7 retention lift was significant enough to justify full release within the original timeline.
Feature now drives the majority of successful onboarding completions.
For users who completed the personalisation flow vs. control.
Including 3-week beta and full rollout.
The most common 0→1 mistake is building something new when the value you need is already there. In this case, the aha moment existed — we just weren't surfacing it well. Discovery before building saved us from building the wrong thing.
Keeping scope narrow also mattered. An 8-week timeline forces good prioritisation. We cut three ideas that were interesting but not essential, and the feature is better for it.