Onboarding Flow

Context

Our app provides personalized skincare routines based on an AI Skin Scan. The initial onboarding flow required first-time users to perform this scan immediately after the welcome screen. However, data showed significant friction in this journey:

• 13% of users skipped the scan entirely.

• 30% dropped off right after tapping "Start Skin Scan" and never returned. This indicated that the flow was forcing a high-effort action before establishing trust, negatively impacting both user retention and data completeness.

Problem

We discovered that asking users for a Skin Scan immediately(before they understood the app’s value) created psychological resistance. The upfront, heavy tutorial also overwhelmed users with too much information at once, failing to clearly communicate why the scan was necessary for their personal benefit.

Objectives

• Reduce onboarding friction for first-time users by lowering the barrier to entry.

• Deliver immediate value by helping users understand the app’s core benefits before asking for data.

• Improve Day-1 retention by stabilizing the initial user funnel.

Approach

• User Research & Persona Refinement Conducted interviews and analyzed drop-off behavior to identify key friction points. I updated user personas to better reflect "Trust Thresholds" and "Urgency," allowing us to design for different levels of user readiness.

• Redesigned "Value-First" Flow I redesigned the linear onboarding into a flexible, 4-step experience. By decoupling the Skin Scan from the mandatory entry flow, users could now explore the dashboard first, building trust with the UI before committing to a scan.

• Contextual Mini-Tutorials (Just-in-Time Learning) Instead of a one-time, bulk tutorial at the start, I embedded short, contextual guidance into key feature screens. This supported "learning-by-doing" and reduced cognitive load.

• Cross-functional Collaboration Worked closely with product and engineering teams to ensure the new modular flow was technically feasible and allowed for independent iteration of onboarding steps.

Outcomes

• Day-1 retention increased from 17% → 23% (a 35% improvement compared to the previous baseline).

• Bounce rate significantly decreased immediately after onboarding, as users were no longer forced into a high-friction task.

• Increased User Agency: Allowing user-led exploration led to more qualified scan completions, as users understood the "why" behind the action.

Platform Learnings

• Modular UX Architecture: Designed the onboarding as a layered experience, allowing for flexible updates to feature guidance without disrupting the core entry flow.

• Scalable Guidance Patterns: Established a reusable pattern for in-app tutorials that can be applied to future feature rollouts.

• Impact of Progressive Disclosure: Confirmed that revealing complex features (like AI scanning) at the right moment—rather than all at once—is crucial for maintaining user trust in a SaaS environment.