CareerCompass
CareerCompass is a mobile career-fair companion developed through two rounds of social prototyping. Survey data identified conversation initiation as the primary barrier for 10 of 18 students. The team translated that evidence into vendor maps, meeting requests, profile integration, and direct contact tools.
ROLE
UX Research & Interaction Design

Problem & Context
Career-fair participation is a social coordination problem. Ten of 18 surveyed students identified conversations with vendors or peers as their greatest challenge. The effect of this is that wayfinding and information tools alone cannot create engagement. CareerCompass therefore pairs navigation with lower-pressure paths to contact and scheduled interaction.
RESEARCH FINDING
10 out of 18 surveyed students named conversations with vendors and other students as the biggest challenge at career fairs.
DESIGN IMPLICATION
Ten of 18 responses concentrated on conversation difficulty. In other words, the central barrier was initiation, not access to information. The design therefore reduces uncertainty before contact through maps, profiles, meeting requests, and direct messaging.
4/18
Core Challenge
Engaging in conversations
10/18
2/18
1/18
Research & Discovery
The team surveyed 18 students and conducted two facilitated prototype sessions in weeks 8 and 10. Session notes recorded participation patterns, task clarity, and discussion behavior. Furthermore, each round converted observed social friction into a specific design change.
Key Findings
FINDINGS
- Students completed assigned tasks but often remained quiet. The effect of this is that task completion did not equal social engagement.
- Week 10 participants engaged more after clearer tasks and a one-on-one Slack discussion. Furthermore, students still struggled to socialize with one another.
- Both rounds showed that the system needed explicit peer-to-peer interaction, not only vendor discovery.
DESIGN IMPLICATIONS
- Vendor maps reduce wayfinding uncertainty before an approach.
- Meeting requests structure vendor contact before the fair.
- Profile and direct-contact tools centralize preparation. Furthermore, peer interaction requires a dedicated social mechanism in the next iteration.
Design Approach
Iteration A: Task-Based Testing (Week 8)
DESIGN HYPOTHESIS
Task-based prompts and a vendor-map flow would make the first action clear.
WHAT WE TESTED
The team used three spreadsheet-based activities alongside the Figma prototype.
KEY INSIGHT
Students focused on answering questions and remained quiet. The effect of this is that the activities measured task completion more than social interaction.
NEXT ITERATION
The team connected the activities more clearly and added one-on-one discussion for week 10.
Iteration B: Enhanced Task Design (Week 10)
DESIGN HYPOTHESIS
Clearer tasks and one-on-one discussion would increase participation.
WHAT WE TESTED
The team revised the activities, clarified the project goal, and used one-on-one Slack discussion during the session.
KEY INSIGHT
Participation increased, but students still did not socialize with one another. In other words, facilitation improved response without resolving peer-to-peer interaction.
NEXT ITERATION
The next version should test structured peer prompts and live interaction between students.
Key Design Screens

MAP
The map locates vendors when students cannot find them on site.

PROFILE
The profile connects a resume, LinkedIn, and Handshake in one place.

SCHEDULE
The schedule lists upcoming fairs and attending companies.

NETWORK
The network lists attending companies, supports one-on-one meeting requests, and opens direct contact.
Outcome & Impact
Takeaways from Testing
USER TESTING RESULTS
- Week 8 tasks were organized and easy to follow, but tasks 1 and 3 remained disconnected and limited conversation.
- Week 10 tasks were clearer and more connected. Participation increased during one-on-one discussion.
- Students remained quiet in both rounds. The effect of this is that clearer tasks alone did not create peer-to-peer interaction.
NEXT STEPS FOR VALIDATION
- Test structured icebreakers between students during live sessions.
- Measure peer-to-peer participation separately from task completion.
- Evaluate whether meeting requests and direct contact reduce hesitation before vendor conversations.
Reflection
PROTOTYPING INSIGHTS
Grevet and Gilbert’s account of social computing frames norms, roles, trust, critical mass, and homophily as design constraints. CareerCompass testing demonstrated the same principle: a usable task flow did not produce social interaction. Furthermore, the second round improved participation without creating sustained peer exchange.
WHAT WORKED
Week 10 improved task clarity and participation. Furthermore, the map, schedule, profile, and contact flows organized career-fair preparation around concrete actions.
WHAT WOULD CHANGE
The next study would replace isolated worksheet tasks with structured peer prompts, reciprocal introductions, and observation of student-to-student exchange. The effect of this is that evaluation would measure social interaction directly.