Personal project · B2C · Mobile · UX Design · 2024
July 2024 - Aug 2024
Tools
Figma
Vinted — Redesigning the bundle feature to stop drop-off
Vinted's bundle feature lets sellers combine multiple items into a single shipment — saving on postage and increasing average order value.
The problem: most users who started the flow never finished it.
60% of regular users called the process inefficient despite using it regularly. 46% specifically asked for search and easier editing. Users weren't failing because the concept was wrong — they were failing because the interface gave them nothing to orient around.
This was a three-week personal project. Solo end-to-end — research, information architecture, wireframes, and high-fidelity prototype.
Three problems, three targeted interventions.
Endless scrolling with no way to find items — sellers with large wardrobes had to scroll indefinitely to find what they wanted to bundle. Fixed with a persistent search bar and category filters directly within the bundle flow.
Broken selection states — the deselect mechanic was unclear. Users couldn't tell what was selected and what wasn't. Fixed with explicit, visually distinct Select and Selected states on each item card.
No sense of progress — users described a "did that actually work?" anxiety throughout the flow. Fixed with a sticky bundle summary bar anchored to the bottom of the screen, showing live item count and running price total at all times.
Lo-fi wireframes → user flow validation → hi-fi prototype covering the full
happy path plus edge cases — empty states, item removal, price recalculation.
Accessibility baked in from wireframe stage: WCAG 2.1 AA contrast, 44px touch targets, ARIA labels in handoff specs., three targeted interventions.
Lo-fi wireframes → user flow validation → hi-fi prototype covering the full happy path plus edge cases — empty states, item removal, price recalculation.
Accessibility baked in from wireframe stage: WCAG 2.1 AA contrast, 44px touch targets, ARIA labels in handoff specs., three targeted interventions.
30%
Reduction in task completion time
Accessibility isn't a final check — it's a constraint that improves every decision it touches. Building it in from wireframes made the handoff cleaner and the prototype more honest.
The sample size is small. Five users tells you direction, not magnitude. The next step would be A/B testing the sticky summary bar in production — it's the most opinionated change and the one most worth validating with real conversion data.
Personal project · B2C · Mobile · UX Design · 2024
Vinted — Redesigning the bundle feature to stop drop-off
Vinted's bundle feature lets sellers combine multiple items into a single shipment — saving on postage and increasing average order value.
The problem: most users who started the flow never finished it.
60% of regular users called the process inefficient despite using it regularly. 46% specifically asked for search and easier editing. Users weren't failing because the concept was wrong — they were failing because the interface gave them nothing to orient around.
This was a three-week personal project. Solo end-to-end — research, information architecture, wireframes, and high-fidelity prototype.
Endless scrolling with no way to find items — sellers with large wardrobes had to scroll indefinitely to find what they wanted to bundle. Fixed with a persistent search bar and category filters directly within the bundle flow.
Broken selection states — the deselect mechanic was unclear. Users couldn't tell what was selected and what wasn't. Fixed with explicit, visually distinct Select and Selected states on each item card.
No sense of progress — users described a "did that actually work?" anxiety throughout the flow. Fixed with a sticky bundle summary bar anchored to the bottom of the screen, showing live item count and running price total at all times.
Three problems, three targeted interventions.
Context
Process
Lo-fi wireframes → user flow validation → hi-fi prototype covering the full happy path plus edge cases — empty states, item removal, price recalculation.
Accessibility baked in from wireframe stage: WCAG 2.1 AA contrast, 44px touch targets, ARIA labels in handoff specs., three targeted interventions.
30%
Reduction in task completion time
What this taught me
Accessibility isn't a final check — it's a constraint that improves every decision it touches. Building it in from wireframes made the handoff cleaner and the prototype more honest.
The sample size is small. Five users tells you direction, not magnitude. The next step would be A/B testing the sticky summary bar in production — it's the most opinionated change and the one most worth validating with real conversion data.
