Unstuck
A mobile app that helps people who are good at their jobs, but no longer want them, find which parts of their work can still change so they stop feeling stuck.
Try the App
Year
2026
Platform
Mobile
Role
UX Designer
The Problem
Many stay for years because they can't tell if the job will ever improve.
By the end of the workday, they're too drained to figure it out.
Skilled but Unhappy
Most people I surveyed had been good at a job they no longer wanted and most of them stayed a year or more.
Worn Out by Work
Half said work often leaves them worn out.
Advice That Skips the Present
Most career tools focus on the next job, not the one people have now.
The Solution
Talk Through Your Week
Each session opens with one question: what took up your time at work last week? Users can talk or type their answer.
Rate the Effort
Users rate how much effort each responsibility takes. Unstuck builds that list from their answers.
Find the Biggest Gap
Results show the work that takes the most effort but matters least to the company. That's the best place to start.
Turn Results Into Steps
Users plan small steps to make their hardest work easier. When they're out of ideas, AI suggests a few to start from.
Background
Unstuck is a mobile app I designed. I handled the research, design, and testing on my own. I also built a working version of the app. Testers used it to talk through real situations from their current jobs.
My first idea was to point out each user's strengths at work and suggest other jobs where those strengths could transfer. My research pointed somewhere else. Before looking for a new job, workers needed to know if their current one could still improve.
One person I interviewed put the problem simply:
"Just because you're good at something doesn't mean you enjoy it."
Interview participant
Research
I wanted to know why people stay in jobs they no longer want. I looked at the career tools already out there, interviewed 5 people, and ran a survey with 26 responses.
Competitive Analysis
I compared 5 tools used for career help, including Teal, CareerExplorer, and Claude.
•
What They Do Well: Teal keeps a job search organized. CareerExplorer suggests careers based on a long quiz. Claude can give helpful advice on almost any work question.
•
What They Miss: Teal assumes you already know the job you want. CareerExplorer asks over 100 questions. Claude's answers stay broad unless you know exactly what to ask. Teal and CareerExplorer focus on the next job, not the current one.
•
Where Unstuck Fits: Unstuck asks very little from someone already burned out. It helps them understand their current job first and see if it can still improve.
User Interviews and Survey
•
Who I Talked To: 5 participants at different points in their careers. I asked about real moments when they felt stuck at work and what made them stay or leave.
•
What Stood Out: Talking about work quickly turned emotional. Many wanted to vent, which showed me where the app could help.
•
How I Checked It: To see if the patterns held up in a larger group, I shared a 5 minute survey on Discord, Reddit, and with people I know. I sorted every answer into an affinity map and made an empathy map for each person I interviewed.
What I Found
•
Good at It, Done With It: 20 of 26 had been good at a job they no longer wanted. 18 of those 20 stayed for a year or more.
•
Too Drained to Change: 13 of 26 said work often or almost always leaves them worn out. One person told me, "When I had a job I was too tired to search."
•
Strengths Only Count When You Find Them: 22 of 26 had been told about a strength they never noticed. Only 4 believed it right away. 24 of 26 found a strength through the work itself.
•
No Clear Way Out: The top reasons for staying were not knowing what else to do, money, and being too tired to start.
Define
I created 3 personas based on why people stay stuck. Marcus, the Stayer, has been at the same company for years and never looked elsewhere. Andre, the Depleted One, knows roughly what he wants but is too worn out to start. Nicole, the Blocked One, knows exactly where she would go, but her boss stands in the way.
Empathy Maps
User Personas
How Might We…
Ideation
My first idea depended on the app finding each user's strengths. But my research showed that people only trusted a strength they found on their own. If they didn't believe the strengths, the job suggestions wouldn't mean much either. So I focused on the parts of work that drain them instead. I wrote 3 problem statements and each one set a goal for the design:
✓
People can't tell if their job can be fixed, so the app shows which parts can still change
✓
They have no energy left to act, so the first step has to feel small and easy
✓
Talking to a manager doesn't help when the manager is the problem, so the app works without one
My next version split work into 7 parts, like pay, growth, and recognition. It was too big for this project. In my interviews, real moments drew more detailed answers than questions about the job overall. So I built each session around one recent work week and sorted my features into 4 groups:
•
Must-Have: A guided look at one work week, a list of responsibilities in the user's own words, a way to rate what each one takes out of them, a comparison of the results, and a plan of small steps.
•
Nice to Have: A progress tracker, a way to pick up where the last session left off, and reminders for planned steps.
•
Surprising and Delightful: Work categories that let sessions be compared over time and a way for users to group their responsibilities however they like.
•
Can Come Later: A map of who helps, who drains, and who blocks the work.
I mapped every screen in a sitemap. The shaded screens are the ones I designed for this first version. The rest show where the app could grow.
I drew 2 user flows. One follows a new user through their first session. The other follows someone coming back to check on their plan or see their progress.
I sketched every screen in low fidelity wireframes. This let me work out the order of screens and what each one asks before adding color and detail.
Design
Users may open Unstuck after a long workday. So each step asks for as little as possible, and AI does the heavy lifting where it can:
•
Talk or Type: Each session opens with one question about last week. Users can answer out loud or type, whichever is easier in the moment.
•
AI Builds the List: AI turns the answer into a list of responsibilities in the user's own words. If an answer is too short, the app asks follow-up questions first. Users can fix anything that came out wrong.
•
Rate It, Then Rank It: Users rate how much each responsibility takes out of them. Then they put their responsibilities in order by how much the company depends on each one.
•
AI Suggests, Users Decide: When users need ideas for a next step, AI offers suggestions written for that responsibility. It never tells anyone what to do with their career.
•
Built on 4 Values: Calm, honest, hopeful, and actionable. Users may open the app already drained, so it needed to feel easy on the eyes.
•
A Softer Navy: My first palette felt too dark and heavy, so I lifted it to a softer navy. I checked every color for accessibility and fixed the ones that failed.
•
One Warm Accent: Orange marks buttons and actions. A soft blue helps users compare information.
•
A Split Wordmark: I explored several logo directions and chose a wordmark with a line splitting "un" and "stuck." The line stands for what keeps someone stuck at work, and the app helps remove it.
Testing & Iterations
I tested the working app in 2 rounds over video. Testers used it on their own phones and thought aloud as they went. Round 1 had 3 testers. In round 2, the same 3 came back to see if my changes worked, and 2 new testers joined with fresh eyes. Testers worked through these tasks:
•
Answering questions about last week by talking and typing
•
Checking that the list of responsibilities matched what they said
•
Rating each responsibility and making sense of the results
•
Planning a next step and updating their to-do list
•
Feedback: The app first asked how much "energy" each responsibility takes. Some round 1 testers read energy as something work gives them. Time with their team left them feeling good, so they almost rated it as high energy.
•
Final Iteration: I renamed it effort and described each level by the time and focus it takes. In round 2, all 5 testers understood effort the way the app means it.
•
Feedback: Talking didn't work for one tester. Her company manages her laptop and a call was already using her microphone. Nothing on screen told her, so it looked like she hadn't answered.
•
Final Iteration: The app now says when it can't hear someone and offers typing instead. Many of the users Unstuck is for work on managed laptops like hers.
•
Feedback: Some questions ask users to pick a responsibility they already named and say why. I listed those answers under the question, but 2 of 5 round 2 testers missed the list and went straight to the answer box.
•
Final Iteration: A line under the question now points to the list. I also removed a fade that hid the last answer.
Tested Version
Updated Version
•
Feedback: Some responsibilities can't change right now. One tester had no say over 2 of her 3. Others said they would write a step like "Nothing can be done," which could never be checked off.
•
Final Iteration: A "Can't change right now" option puts that responsibility on hold. It no longer looks like a step they failed to finish.
Tested Version
Updated Version
•
Feedback: One tester liked several AI suggestions but could only use one. Another said some suggestions were too broad to fit into a workday.
•
Final Iteration: Users can now choose as many suggestions as they like. Each one becomes its own step they can edit and "Show more" brings up others.
Tested Version
Updated Version
•
Feedback: Testers wanted to know if their responsibilities were becoming easier, but nothing in the app showed change over time.
•
Final Iteration: Home now shows if each responsibility takes less effort than last month. A new Progress screen tracks effort and impact month by month, which also gives users a reason to come back.
Tested Version
New Progress Screen
•
Feedback: 2 testers weren't sure how to explain their results to a manager. One said her manager would ask how she measured effort and impact.
•
Final Iteration: I added a Common Questions section with answers to questions like "Where do these numbers come from?" and "How can I bring this up with my manager?"
Tested Version
Updated Version
Results
Success meant testers could find the part of their job worth changing on their own and explain it in their own words. The results screen got the strongest reaction in every session.
•
Round 1: All 3 testers found the part of their job that took the most effort while mattering little to their company. 2 of them said something about their week out loud that they hadn't said before.
•
Round 2: Testers read their results quickly. Most of their feedback was about what comes after, like putting work on hold or seeing progress.
Final Prototype
Final Thoughts
When I first asked people about their jobs, many of them got emotional. They became passionate and upset and wanted to vent. One participant called me in tears because she felt stuck and didn't know what to do next. That showed me where Unstuck could really help. My biggest lesson was how much depends on asking the right questions. Once a conversation reaches that deeper level, people share things that can spark new features and ideas.
Next Steps
•
Test the Latest Changes: Run the screens I changed after round 2 with new testers. Each change came from feedback, but none have been tested yet.
•
Track Real Progress: Test the Progress screen with returning users after a month of real use, so it shows changes from their own work instead of sample data.
•
Show Where Each Answer Came From: One tester asked to see which part of his answer each responsibility came from. That would help users see why something came out wrong.
What I'm Most Proud Of: Unstuck started as just an idea about work with no clear direction. Everything changed once interviews reached that emotional level. I leaned into it and each round of research and testing led to the next idea. In many ways, the users shaped the app and I designed around nothing but what they shared. I also built a working version on my own, which led to more honest reactions in testing. Even if someone only uses Unstuck to vent, that still matters. Sometimes the first step is simply saying it out loud.