Find Onboarding Friction with Data

Company Context

FinFlow is a mid-stage personal finance app with 6 million registered users and a freemium subscription model. The company has strong top-of-funnel acquisition through paid channels, but activation is lagging behind competitors because too many new users abandon the onboarding flow before reaching their first key value moment.

Problem

FinFlow's onboarding flow has 7 steps: account creation, email verification, goal selection, bank connection, income setup, notification preferences, and dashboard tutorial. Of users who start onboarding, only 42% complete it, and just 24% link a bank account within 24 hours. Early retention is highly correlated with successful bank connection: users who connect an account in their first session are 2.3x more likely to still be active after 30 days.

Leadership wants to know where the biggest friction points are and what should be fixed first. You are the PM responsible for onboarding. Your task is not only to identify drop-off points, but also to determine which friction is most meaningful for users and the business, and how you would validate improvements.

Deliverables

1. Define how you would use quantitative and qualitative data to identify friction in the onboarding flow.
2. Segment users and explain which segments you would prioritize when diagnosing onboarding issues.
3. Recommend the top 2-3 onboarding changes you would prioritize first, with rationale.
4. Define success metrics and an experiment plan to validate whether your changes reduce friction.
5. Discuss the key trade-offs you would make between reducing friction, collecting user data, and preserving downstream product value.

Constraints

• You have one product analyst, one designer, and two engineers for the next 6 weeks.
• The bank-connection vendor cannot be replaced this quarter.
• Compliance requires FinFlow to collect certain identity and consent information before users can access budgeting features.
• You may change step order, copy, UI, defaults, and progressive disclosure, but cannot remove legally required disclosures.