← Back to projects

Wink

A compatibility-first dating app, built to replace endless swiping with fewer, better matches

Role
UX Designer and Researcher
Timeline
8 weeks
Team
Solo Project
Tools
Figma
Overview

The Objective

Wink is a dating platform built around compatibility over quantity. Instead of a bottomless stack of strangers, its matching algorithm scores compatibility from a user's answers to hundreds of questions about interests and core values, then surfaces a smaller pool of genuinely fitting matches.

The goal wasn't just a prettier swipe screen, it was a more personalized and inclusive experience for people who'd grown tired of superficial matching, ghosting, and apps that didn't reflect who they actually are.

Constraints: a solo project with no dedicated engineering team every research, design, and testing decision had to hold up on its own before any handoff.

My role: research, personas, wireframes, visual design, and both rounds of usability testing.

Research

Discovery

Three people, three very different reasons to be tired of dating apps. Before any screens got designed, personas were built around the pain points that kept coming up: matches that go nowhere, time wasted on the wrong fits, and apps that weren't built with everyone in mind.

Research methods:

  • Persona development: Ava, Marcus, Jordan
  • Round 1: black and white wireframes, moderated with 5 testers
  • Round 2: clickable hi-fi prototype, moderated with 5 new testers
Ava persona photo

Ava, 28

UX Designer · Chicago, IL
Goal
  • Meaningful connection
  • Shared values
  • Open to serious
Pain point
  • Superficial matches
  • Ghost culture
  • Misaligned expectations
Bio

Ava is a 28-year-old UX Designer based in Chicago who brings the same thoughtfulness to dating that she does to her work. She's past mindless swiping and now wants someone who actually reads her profile, shares her values, and is open to seeing where things go.

Marcus persona photo

Marcus, 32

Marketing Manager · New York, NY
Goal
  • Committed relationship
  • Lifestyle compatibility
  • Firm on deal breakers
Pain point
  • Time wasted on bad fits
  • Unclear intentions
  • Endless swiping
Bio

Marcus is a 35-year-old Marketing Manager from New York City who knows exactly what he's looking for and has little patience for anything less. He's the type to set dealbreakers before he even starts browsing, because his time is valuable and so is his energy.

Jordan persona photo

Jordan, 24

Graduate Student · Austin, TX
Goal
  • Safe, inclusive space
  • Identity expression
  • Community & connection
Pain point
  • Binary gender options
  • Misgendering risk
  • Niche match pool
Bio

Jordan is a 24-year-old graduate student in Austin, Texas, who has spent enough time on dating apps that weren't built for them. Non binary and out, Jordan isn't just looking for a match but an inclusive space where they don't have to explain themselves.

Ideate

From paper to pixels

Playful enough to flirt with, calm enough to trust

Pink, white, and purple carry the warmth and playfulness of the brand, while a soft, rounded UI keeps the experience feeling cute rather than gimmicky.

Wink style guide showing colors, type, and UI components
Ideate

Four zones

Before wireframing, I mapped the full app architecture from onboarding through discovery, connection, and account support.

Wink app architecture map covering onboarding, discovery, connection, and account support
Ideate

Five page types

Each Page type was planned out to better understand the workflow.

Paper Sketches

Paper sketch of the onboarding screen
Paper sketch of the feed screen
Paper sketch of the messaging screen
Paper sketch of the user profile screen
Paper sketch of the settings screen

Low Fidelity

Low fidelity onboarding screen
Low fidelity feed screen
Low fidelity messaging screen
Low fidelity user profile screen
Low fidelity settings screen

High Fidelity

High fidelity onboarding screen
High fidelity feed screen
High fidelity messaging screen
High fidelity user profile screen
High fidelity settings screen
Onboarding
A sequential flow that gathers profile and preference data.
Feed
Card-stack browsing built for fast, high-interaction evaluation.
Messaging
The communication hub, unlocked only after a mutual match.
User Profile
Expanded view with interests, prompts, and photos.
Settings
Menu-driven controls for filters, privacy, and account.
Ideate

The Three tasks

Each participant stepped into a short scenario, then attempted the task cold.

  1. 01
    After swiping right on Kyle's profile, you both mutually matched. Reach out by sending him a first message.
    Start a conversation with a match
  2. 02
    Chad sent you a message that made you uncomfortable. Report the conversation so he can't contact you again.
    Block and report an inappropriate user
  3. 03
    You've matched with someone on the app and are ready to settle down. Delete your account.
    Delete account
Test

Round 1: Low Fidelity Prototype

Five testers walked through hand-drawn screens on index cards before a single pixel was designed — cheap enough to throw away, real enough to expose problems early.

Task Avg. time
1. Create account 2:37
2. Block & report 3:03
3. Browse & match 1:37
Total per tester 7:27

What informal testing found

  • The report/block flow (Task 2) took the longest — testers weren't sure where to find it.
  • Some of the learning curve was the study format itself: several testers had never done a usability session before.
  • Without color, the wireframe was genuinely hard to read for some participants.
Test

Round 2: High Fidelity Prototype

The same three tasks, now run against an interactive Figma prototype held in-hand by five new testers recruited from within the app's target age range.

Task Avg. time
1. Create account 1:15
2. Block & report 0:44
3. Browse & match 0:49
Total per tester 2:44

What formal testing found

  • The filter icon was regularly mistaken for account blocking or reporting — iconography needed to be clearer.
  • The report form didn't offer enough options to capture what testers actually wanted to flag.
  • A few prototype-only issues (unresponsive touch targets, a message icon that blended into the background) got flagged for fixing.
Design Impact

Results

9/10
Average overall experience rating
9.2/10
Average task intuitiveness rating
−63%
Drop in time on task vs. Round 1
Iteration

Closing the gaps testing surfaced

Every formal-testing finding turned into a specific design change — nothing got fixed without a tester pointing at exactly why it was confusing

  • Enlarged and added contrast to iconography so it reads clearly against the pink background.
  • Increased touch target sizes across buttons to cut down interaction errors.
  • Redesigned the messaging icon so it no longer blends into a matched profile.
  • Rebuilt the report form to be more interactive, with clear support for selecting multiple reasons.
  • Added confirmation feedback whenever a user takes an important action, like reporting or matching.
Annotated Wink screens showing a report confirmation message and an outlined compatibility score
Annotated Wink screens showing an enlarged profile arrow and higher-contrast settings icons
Iteration

What two rounds of testing actually proved

The gap between the two testing rounds ended up being the most useful data point in the whole project. It wasn't just that the hi-fi prototype looked more finished testers moved through it in a third of the time it took to get through the same tasks on paper, and rated it 9+ out of 10 on both experience and intuitiveness.

Some of that gap was fidelity: color, real photography, and a tactile phone screen are simply easier to navigate than pencil on an index card. But a real part of it was the design itself catching up to what round-one testers were struggling with especially the block-and-report flow, which went from the slowest, most confusing task to one of the fastest.

The lesson I'm carrying forward: test the safety-critical flows first. Reporting and blocking aren't a feature users browse casually they're the flow someone reaches for at their most uncomfortable moment, and that's exactly where clarity matters most.

Reflection

What two rounds of testing actually proved

The gap between the two testing rounds was the most useful data point in the project. Fidelity mattered less than expected testers moved through the hi-fi prototype in a third of the time it took on paper, and rated it 9+ out of 10 on both experience and intuitiveness.
The block-and-report flow went from the slowest, most confusing task in Round 1 to one of the fastest in Round 2 a reminder to prioritize testing safety-critical flows before anything else next time.