Case study
TStar: UofT Cohort App
Designing a cohort-based platform to strengthen student belonging within academic cohorts
- Role
- Team Lead, UX Designer and Researcher
- Timeline
- 2 months (Oct–Dec 2025)
- Tools
- Figma, Miro, Otter.ai, Procreate
01Overview
U of T students want friends in their cohort, but class rarely gives them a reason to talk.
Students balance demanding academics with work and personal commitments, so there’s little time to build connections. Existing campus platforms are broad and aren’t designed for academic cohorts, where shared classes and goals create the strongest sense of belonging.
How might we design a low-friction way for students in the same cohort to form meaningful peer connections?
Led the team
As team lead, I ran our group discussions, delegated tasks and kept every stage of the project on schedule.
Ran the research
I wrote the interview protocol, conducted one-on-one interviews and led the affinity diagramming that turned our data into themes.
Designed the app
I sketched the first low-fidelity user flow, ran think-aloud testing with target users and built the high-fidelity screens in Figma.
02Solution
TStar turns meeting your cohort into a game you play together.
Instead of asking reserved students to walk up to someone cold, TStar gives them a reason to interact: quests to complete in real life, profiles that show common ground and a playful way to say hi.
Link the interactive Figma prototype here.
Decision 1
Gamify real-life interaction with Cohort Quests.
Quests lower the barrier to starting a real conversation and make the moment memorable, so students meet in person instead of only messaging.
Add a sentence or a test quote showing why gamification felt low-pressure to reserved students like Shy Sherry.
Decision 2
Show interests and quests on customizable profiles.
Profiles highlight each student’s interests and the Cohort Quests they’ve completed, making it easy to find common ground and start a conversation.
Decision 3
Break the ice with “Shoot a Star.”
A playful button lets students poke fun at a cohort peer, taking the pressure off the first hello in real life.
Decision 4
Build towards a shared goal.
Individual interactions add up to milestones the whole cohort works towards, so connecting feels like a group effort rather than a solo task.
03Research
Students rarely see cohort peers outside class, and faculty events don’t help.
survey responses from students across faculties
one-on-one interviews with U of T students
themes from our affinity diagram
My contributionI wrote the interview protocol, conducted interviews and led our group affinity diagramming, which organized everything we heard into nine themes.
Check these percentages against your raw data. With 31 responses, 54.9% isn't possible (17/31 is 54.8%). Consider using counts like "17 of 31".
Students barely hang out with cohort peers outside class
Only 30% of survey respondents talk or hang out with classmates outside of class. The other 70% do a few times a month, occasionally or never.
What it meansThere's a lack of third spaces for making connections.
Cohort conversations are mostly about school
64% of interviewees talk to cohort peers mainly for academic reasons and have a separate friend group for socializing.
What it meansThere are few chances to socialize casually.
Making cohort friends is hard
Large faculties, different commutes and little time or energy got in the way, and 54.9% of survey respondents described themselves as very to somewhat reserved.
What it meansStudents hesitate to start conversations when they don't feel invited to.
Faculty events don't work for making friends
None of our interviewees found faculty events effective for making friends. They described them as too professional and intimidating.
What it meansLow-pressure settings feel more comfortable.
04Process
From four insights to one student we designed for.
We followed the Double Diamond process: Discover, Define, Develop and Deliver. To narrow our scope, we created a persona, empathy map and as-is scenario, then voted on the stages with the biggest pain points.
Need statements
Shy Sherry needs a way to make cohort friends without the awkwardness.
Start conversations
Without worrying about awkward social situations.
Stop waiting
Instead of scrolling her phone until someone else talks first.
Skip faculty events
So she can still get home early.
Fit her commute
Social activities that don't clash with her schedule.
Ideation and prioritization
We dot voted every idea, then built on the home runs.
Each of us brainstormed ideas for the need statements. We dot voted and sorted them on a prioritization grid, then focused on the “home run” ideas for our final solution.
My contributionI sketched the first low-fidelity prototype of our user flow, which became the base for our think-aloud testing.
05Testing
Five think-aloud sessions showed us where people got stuck.
We tested the lo-fi screens with five target users and found three pain points to fix in the high-fidelity prototype.
Unclear labels
Labels were unclear and icons were inconsistent.
QR code scanning
There weren't enough options for scanning QR codes.
After "Shoot a Star"
The flow after shooting a star wasn't intuitive.
Show a before and after for each of the 3 think-aloud findings.
06Reflection
What I learned
Define a clear research focus early.
Broad exploration helped, but narrowing to our target user sooner would have given us findings more relevant to a cohort-specific solution.
Diverge before converging.
Time limited us to one lo-fi prototype per task. Parallel sketching across the team would have led to a stronger concept.
Design is iterative.
With one testing round, key feedback surfaced late. More rounds would have let us refine the information architecture further.