MishaelGilland

01 — Problem

Users reported feeling overwhelmed by feature density. Task completion rates dropped 22% after a major feature push. The product needed to refocus on its core value: helping people enter deep work.

02 — Research

Conducted 12 user interviews and analyzed session recordings from 500+ users. Key insight: power users wanted depth, but 70% of daily users only used 3 core features. Competitive analysis of Notion, Linear, and Sunsama informed our positioning.

03 — Process

Ran a 2-week design sprint with cross-functional stakeholders. Prioritized features using RICE scoring. Created low-fi prototypes tested with 8 users before high-fidelity design. Iterated on navigation architecture three times based on tree-testing results.

04 — Wireframe

Explored sidebar vs. top-nav vs. command-palette-first patterns. Wireframes emphasized progressive disclosure — advanced features hidden behind intentional gestures. Mobile wireframes prioritized thumb-zone placement for primary actions.

05 — Design System

Extended Nexus tokens with focus-mode palette: muted backgrounds, single accent color, reduced visual noise. Created 8 new components: FocusTimer, TaskBatch, AmbientPanel, SessionSummary. Documented motion principles: calm, purposeful, never distracting.

06 — Final UI

Shipped a three-column layout that collapses to single-column focus mode. Hero metric: time-in-focus displayed prominently. Subtle grain texture and soft gradients create warmth without distraction. Dark mode as default with carefully tuned contrast ratios.

07 — Reflection

The redesign increased weekly active usage by 31% and NPS by 18 points. I learned that restraint is a feature — saying no to good ideas was harder than saying yes. Next time, I'd involve engineering earlier in wireframe validation for feasibility checks.

01 — Problem

Enterprise trial users abandoned setup at 45% completion. Average time-to-first-dashboard was 47 minutes. Sales flagged onboarding as the #1 churn driver in the first 30 days.

02 — Research

Mapped the existing 12-step onboarding funnel. Identified three drop-off cliffs: data source connection, widget configuration, and team invite. Surveyed 30 churned trials — 'too complex' and 'unclear value' topped responses.

03 — Process

Adopted jobs-to-be-done framework. Defined the 'aha moment' as seeing first AI-generated insight on real data. Reduced onboarding to 4 steps with optional advanced setup. Built interactive demo mode with sample data for instant gratification.

04 — Wireframe

Storyboarded the emotional journey from signup to first insight. Designed a wizard with contextual help and skip options. Created empty-state illustrations that guide rather than intimidate. Mobile wireframes deferred — enterprise users primarily on desktop.

05 — Design System

Added onboarding-specific components: StepIndicator, DataSourceCard, WidgetPicker, InsightPreview. Established illustration style: geometric, monochromatic with accent highlights. Wrote microcopy guidelines for supportive, concise tone.

06 — Final UI

Launched guided setup with live preview panel showing dashboard building in real-time. Celebration moment on first insight with confetti micro-animation. Progress saved automatically — users can resume across sessions.

07 — Reflection

Trial-to-paid conversion improved by 24%. Setup completion rose from 55% to 82%. The project reinforced that onboarding is product — not a formality. I'd invest more in analytics instrumentation upfront next time.