At-A-Glance
Forming Compatible Student Teams for Better Outcomes
Students often struggle with mismatched project teams, which can lead to unclear goals, uneven workloads, and poor collaboration. This project tackled that challenge by creating a smarter team formation process that matches students based on shared goals, working styles, and availability, while also supporting stronger course outcomes for instructors and institutions.
Using Lean UX principles, I conducted secondary research, built rapid prototypes, and validated ideas directly with users through multiple iterations. This approach helped shape a solution that maximizes both user outcomes and business value.
My Role
UX Research
UX Design
UI Design
Skills
Secondary Research
Rapid Prototyping
User Validation
Iterative Testing
Context
8 weeks
Spring 2025
Solo Project
Academic Projects
Business Problem
Starting a team project can feel like a gamble..?
Students often get assigned to teams randomly or based on superficial criteria like social ties, which frequently results in poor collaboration, unclear expectations, and uneven workload distribution. Instructors also lack time and data to form balanced groups, and existing tools fail to prioritize compatibility in working preferences and project goals.
Business Outcomes
Higher Course Satisfaction and Improved Academic Outcomes
The business goal for this project is to help students experience better collaboration and achieve stronger project outcomes, which in turn raises overall course satisfaction. In the long run, this contributes to a stronger academic reputation, higher student retention, and improved institutional results. We’ll know we’re successful when we see indicators such as higher team satisfaction, more balanced workload distribution, better project outcomes, and fewer requests for team changes among our target users.
The solution
A Transparent Process for Compatibility-Based Team Formation
Our product fills this gap by offering a thoughtfully designed team formation platform that matches students based on individual goals, skills, collaboration styles, and role preferences. It is grounded in an understanding of user needs, context, and collaborative dynamics. By doing so, it aims to improve collaboration, fairness, and learning outcomes through a user-centered, data-informed approach.
Literature Review
Team Formation Practices Overlook Compatibility and Balance
To define the problem space, I reviewed academic literature on team formation in education. The key takeaway was that team compatibility, especially in terms of collaboration preferences, goals, and availability, plays a significant role in project outcomes and student satisfaction. This led me to focus on the idea that early-stage team formation should align mindsets rather than rely only on skill matching. That insight shaped the decision to use lightweight data inputs such as collaboration styles and project goals during onboarding.
“
Letting students freely choose teammates too early can lead to poor matches. Forming groups with balanced skills can reduce friction and improve collaboration.
Lopez et al., 2021
“
Current group formation is often inefficient and relies too much on chance. Using goals, strengths, and availabilitycould help automate better matches.
Lopez et al., 2021
“
Group strategies must consider skills, goals, and personality. Factors like collaboration style and availability strongly influence team cohesion.
Odo et al., 2019
proto-personas
Students Who Need Better Group Work
To ground the solution in real student needs, I developed two proto-personas: Alex, a third-year computer science undergraduate, and Rachel, a second-year cognitive science graduate student. Despite their differences in academic level and goals, both struggle with mismatched teams, unclear expectations, and uneven workloads. Alex wants to build strong portfolio projects with like-minded peers, while Rachel seeks meaningful research partnerships. Together, these personas show how the solution can support diverse students and highlight the need for compatibility, transparency, and student agency in team formation.
Business Opportunity Statement
Opportunity, Focus, and Success Metrics
Current State
The current state of group formation in academic settings has focused mainly on instructor assignment or student self-selection, rarely considering goals, collaboration styles, availability, or work preferences.
Gap
What existing products like CATME fail to address is the need for a balanced and transparent process that matches students by interpersonal fit and project goals.
Solution
Our product will address this gap by helping students form better-matched teams that support fair collaboration, clear roles, and stronger learning outcomes.
Our Focus
Our initial focus will be higher education students in project-based courses.
Success Metrics
We’ll know we are successful when we see better-matched teams, smoother teamwork, and stronger project results.
User Outcomes & Benefits
Why Students Use This Product
This product helps students form groups that are aligned with their goals, work styles, and availability. By matching students intentionally, it makes group work more effective and less stressful.
- Align goals and interests to boost motivation and reduce conflict.
- Combine strengths so students feel more confident contributing.
- Match projects to personal or professional interests.
- Set clear expectations upfront to avoid confusion.
- Improve collaboration for better performance and satisfaction.
- Save time and reduce stress by preventing last-minute team reshuffling.
Students benefit academically, socially, and practically when they start with the right team from the beginning.
Validation approach
Stage-by-Stage Validation Planning
I used Pirate Metrics in the early stage of the process to validate whether the MVP was necessary. At this stage, I focused on acquisition and understanding the user funnel, looking at how students discovered, signed up for, and engaged with the platform.After observing retention and consistent usage, I incorporated Outcome-to-Impact Mapping, I incorporated Outcome-to-Impact Mapping to explore the broader educational and institutional value. This phased approach helped me test user behavior early and assess long-term impact after building a foundation of engagement.
| Stage | What I’m Measuring | What I’m trying to Learn | How I might Measure it | Success Criteria (Target) |
|---|---|---|---|---|
| 🧲 Acquisition | Student interest and willingness to try the platform | Do students see value in a better team formation and want to try it in the first place? | Landing page test with call-to-action (e.g., “Match me with a better team”) | At least 30% of invited students click the CTA |
| ✅ Activation | Whether students experience any value from the product | Does the onboarding process result in teams students feel good about working with? | Post-match survey (Likert scale: “How well does your team match your preferences?”) | At least 70% say they’re satisfied with their team after 1 week |
| 🔒 Retention | Whether students stick with their assigned team and continue using the system | Are students satisfied with the team and engaged with what they are working on? | Multiple peer-reviews to track group experience | At least 80% of teams remain unchanged by the end of the course |
| 💰 Revenue | Higher course evaluations and program performance through well-formed teams | Does using the platform lead to better collaboration, satisfaction, and project outcomes? | Compare project scores, peer reviews, or team performance against past cohorts | Average peer review score ≥ 4/5 and project scores 10% above average |
| 🔁 Referral | Willingness to recommend or reuse the platform in other courses | Do students want to use the platform again in future teams? | End-of-course feedback survey or quick NPS form | At least 60% say they’d use it again in another course |
How Might We
How might we help students find compatible teammates with shared goals, interests, and working styles?
Design Studio Exercise
Explore Possible Solutions
Based on the HMW question, I used a design studio exercise with a six-up template to quickly generate multiple solution ideas. This helped me explore different ways to address the defined problem and compare which ones best meet student needs.
Hypothesis Mapping
Connecting Features to Business Goals and User Goals
This table connects each potential solution to specific user outcomes and business goals. While all ideas address the same core problem of helping students form better-matched teams, each one differs in how it affects the student experience and creates value for the institution. This mapping clarifies what we expect each solution to achieve and supports clear prioritization of which ideas to test first.
| Business outcome | Persona | User outcome | Feature |
|---|---|---|---|
| Higher course evaluations and program performance through well-formed teams. | Student like Alex and Rachel | I can join a group where everyone shares similar goals and interests | Compatibility via past projects, onboarding survey |
| Improved institutional outcomes & academic rankings | Student like Alex and Rachel | I enjoy the group experience and feel it improved my learning | Prompt-based grouping, Compatibility via past projects |
| Increased student internship and job placement rates | Student like Alex and Rachel | I’ll be able to work on what’s aligned with my interests and strength | Interest-based channels |
| Higher student retention and experience | Student like Alex and Rachel | I know what to expect from teammates and understand team expectations upfront | Course-embedded analysis |
| Improved academic performance in team-based projects | Student like Alex and Rachel | I feel confident that I’ll be able to contribute meaningfully | Prompt-based grouping |
Hypothesis Prioritization Canvas
Prioritize What to Test and Build
After mapping possible solutions to both business goals and user needs, I prioritized them based on their potential value and associated risk to decide which hypothesis to validate first. This helped me focus on an idea that balanced impact with feasibility and offered the best chance for meaningful early results.
Hypothesis Risks and Criteria
Prioritized Hypothesis
When prioritizing the hypotheses, I identified risky assumptions and how to validate each one. This helped me start small, test critical assumptions early, and define next steps to build solutions that truly matched student needs. After evaluating all concepts, I decided to prioritize the Onboarding Survey because it presented low risk and high value, making it suitable for achieving a quick win during the MVP stage. This approach could leverage existing data sources such as students' skills, grades, and past project performance to improve team compatibility and reduce reliance on self-reported preferences.
| # | Hypothesis | Value | Risk | Rationale | Risky Assumption | What Would Invalidate It |
|---|---|---|---|---|---|---|
| 1 | Higher team satisfaction through onboarding and past project compatibility | High | Low | Uses existing data such as skills, grades, and past projects to match students more effectively. | Data from onboarding and past performance accurately predicts satisfaction and improves the group experience. | Students skip or rush through onboarding, or the data does not align with actual team satisfaction. |
| 2 | Stronger academic performance through prompt-based grouping | High | High | Needs testing to confirm that the prompt approach supports better academic outcomes. | Prompt responses meaningfully capture team skills and working styles. | There is no measurable performance difference between prompt-grouped and randomly assigned teams. |
| 3 | More students land jobs through interest-based project channels | Medium | Medium | It is challenging to directly link group interests to future job placement without additional evidence. | Projects aligned with student interests result in stronger portfolios and job opportunities. | Interest-aligned projects do not produce meaningful work for portfolios or job applications. |
| 4 | Improved institutional rankings through better group experiences | High | High | Long-term benefit, but more data is needed to prove the link between group experiences and rankings. | Better team experiences lead to higher course evaluations and stronger overall performance, which help improve rankings. | There is no clear correlation between improved group satisfaction and institutional ranking metrics. |
| 5 | Better student retention supported by course-embedded analytics | Medium | Low | Leverages existing course behavior and participation data to support better matches. | Patterns in student behavior predict team compatibility and satisfaction accurately. | Retention and satisfaction rates do not improve with analytics-based team formation. |
Hypothesis
We believe we can achieve higher course evaluations and stronger program performance through well-formed teams if students like Alex and Rachel join groups where everyone shares similar goals and interests, supported by compatibility data from past projects and an onboarding survey.
Experiments
Risky Assumptions
I listed the risky assumptions related to it to clarify the next steps. This helped me start small, test critical points early, and iterate effectively.
| # | Risky Assumption |
|---|---|
| 1 | Students care about being matched with compatible teammates. |
| 2 | Students are willing and able to complete the onboarding process. |
| 3 | Compatibility based on goals, past projects, and skills leads to well-formed teams and improved satisfaction. |
| 4 | Past project performance is a reliable predictor of future team fit. |
| 5 | Students trust the team formation process. |
Experiments
Validate Risky Assumptions
Based on the hypothesis prioritization, I decided to start with an idea that offered high perceived value and low risk, aiming for a quick win. I chose to first validate a foundational assumption: “Students care about being matched with compatible teammates.” This assumption is subjective but critical as it underpins all other project ideas and hypotheses. If students are not concerned with team dynamics, investing further in compatibility-focused features may be irrelevant. Validating this early allows us to reduce risk and build confidence before exploring additional solutions.
Assumptions to Test
| Hypothesis | Risky Assumption | Experiment | What I’m trying to learn | Valid if | Invalid if |
|---|---|---|---|---|---|
| Higher course evaluations and program performance through well-formed teams. | Students care about being matched with compatible teammates | A landing page test | See whether students are interested in a better way to form teams | At least 30% of unique visitors click the CTA (e.g., “Match me with a better team”) or fill out the form | Less than 10% click the CTA or complete the form |
Methodology
To test early interest and engagement, I built a simple MVP landing page with a clear CTA: “Start the team match survey.” The button linked to an embedded onboarding survey to simulate the experience. Instead of just asking if users were interested, I tracked real behavior to get unbiased insights. I also included a short survey to gauge interest, collect basic demographics like major and student level, and check if they had used similar tools before. I shared the landing page with students in my HCI program and a few product and design communities to gather diverse input. This helped me spot opportunities, clarify expectations, and see how familiar they were with existing solutions.
Results & Next Steps
unique visitors
unique clicks
click-through rate
I tested early interest and engagement with a simple MVP landing page and a clear CTA, which brought in 13 unique visitors and 8 unique clicks, resulting in a 60% click-through rate. All participants reported they had never used similar tools before, and 5 out of 8 shared they had struggled to form groups in the past and were somewhat interested in this idea. However, their mixed answers to “What do you think the website is trying to solve?” and feedback asking for more context showed that the value proposition needed to be clearer.
As the next step, I created a clickable prototype with the actual onboarding questionnaire to test whether students would complete the process and better understand the concept and value behind the project.
Prototyping
MVP Development
The landing page test showed early interest in the idea, but it also revealed that a simple static page wasn’t enough to communicate how the product would actually work. To address this, I created an interactive prototype that demonstrated the onboarding and matching flows. This step not only made the concept more tangible, but also allowed me to validate a second key assumption: Students are willing and able to complete the onboarding process. With the prototype, I could observe how students interacted with the flow and whether the team formation experience felt meaningful in practice.
Prototype Walkthrough
Student Onboarding → Team Matching → Team Overview
A student first completes onboarding at the start of the course by filling out a short questionnaire that captures their goals, preferred collaboration style, strengths, and project interests. The system uses this information to suggest team configurations that prioritize shared goals and compatible working styles. Once a team is confirmed, the student can view a team overview with brief profiles of their teammates, including their goals and working styles, to help everyone get to know each other and start the project smoothly.
Design Rationale
1. Onboarding Content Grounded in Insights
Survey responses and research by Lopez et al. (2021) and Odo et al. (2019) highlighted challenges like mismatched work styles, unclear direction, and poor timing. These informed the onboarding quiz design, which includes questions about project goals, collaboration preferences, and typical team roles. The goal was to make the process lightweight yet meaningful enough to support more compatible and balanced team formation.
2. Need for Clearer Concept Communication
When asked what the site was trying to solve, participants gave answers like “teammate finding system” or “help students create project teams.” One response revealed confusion, indicating the need for clearer communication. This insight directly led to developing an interactive prototype to show onboarding and team matching in action.
3. No Familiarity with Similar Tools
None of the participants had used a team formation tool before. While this didn’t signal high demand, it reinforced the need to design a simple, intuitive experience that works without requiring prior tool familiarity.
Methodology
Unmoderated Online Testing
Participants
I tested with seven graduate students who have participated in team-based projects, such as those in computer science or product design, and are familiar with the team formation process.
Objectives
In the previous phase, all eight participants expressed moderate interest in the concept. While the sample size was small, the feedback was generally positive and encouraging. The previous round focused on concept validation through a landing page and survey. In this phase, I wanted to move beyond initial interest and assess whether users could understand and engage with the value, flow, and functionality of the product. Specifically, I aimed to test if users could complete the onboarding flow, understand the team matching process, and feel confident moving forward after joining a group.
Risky Assumptions to Test
- Students are willing and able to complete the onboarding process.
Research Questions
-
Do users understand the purpose and value of the
onboarding flow?
I wanted to see whether the prompts (e.g., collaboration style, project goals) feel relevant and easy to answer, or if users need more context or examples. -
Is the process of joining a group clear and
satisfying?
I observed whether users feel confident in selecting a group and whether the confirmation and handoff to the team overview provide a strong sense of closure. -
Does the team overview help reduce awkwardness or
uncertainty?
I tested whether surfacing teammate info like roles and availability encourages early engagement, or if users still feel hesitant to initiate contact.
Tasks Given to Users
- Complete the onboarding form by entering your project goal, collaboration style, preferred roles, and availability.
- View and choose a recommended group to join.
- Explore the team overview and reflect on how ready you feel to meet this group.
Results
Key Takeaways
1. All Participants Completed the Full Task
Observation
All seven participants from the target population successfully completed the onboarding questionnaire tasks. The System Usability Scale (SUS) score averaged 82.5, indicating strong perceived usability.
Interpretation
The fact that all participants were able to complete the onboarding without dropping off or needing assistance suggests that the onboarding flow is intuitive and well-structured. This outcome supports the second risky assumption: that students are both willing and able to complete the onboarding process when it helps them form better-matched teams. The high SUS score further confirms that the current design supports ease of use and user satisfaction.
Design Recommendations
Given the high SUS score and 100% task completion rate, the current onboarding experience appears to be effective for the MVP stage. No critical usability barriers were identified, so the basic navigation and interaction patterns can be retained.
2. 🧠 Students Want Insight into Teammates' Working Style, Personality, and Communication
Observation
Survey responses showed students want to share more about how they work in teams, not just surface-level facts. Comments included “Ask if the person procrastinates,” “a question about work ethic,” and suggestions for including “personality types” and “preferred communication style.”
Interpretation
Students want the form to capture real behavior, such as how teammates meet deadlines, communicate, and adapt during projects.
Design Recommendations
I added scenario-based questions to reflect key traits like work ethic and conflict style. This method is supported by research on Situational Judgment Tests and allows one concise question to reveal multiple behaviors without lengthening the form.
3. 👥 Students Want More Context About Teammates Before Joining
Observation
Participants said they wished they knew more about their teammates in advance. They asked for info like “how ambitious people are,” “skills and past experience,” and simply “who the members are.”
Interpretation
Students are more comfortable joining a team when they understand teammates’ skills, goals, and personalities beyond just names or roles.
Design Recommendations
I added a team preview screen that shows teammates’ goals, roles, and availability while keeping names hidden for privacy. This gives enough context to support informed decisions.
Iterations
Two Major Iterations
Scenario-based Questions for Richer Insights in Fewer Steps
To reduce redundancy, I replaced two direct questions with one scenario-based prompt. This new format captures key compatibility factors such as work ethic, communication style, and personality in a more efficient way. Informed by SJT research and group work studies, the scenario surfaces real behavioral traits in a single, lightweight step.
Show Member Details Early to Support Confident Group Selection
Allowing users to view key teammate information before joining a group helps them feel more informed and confident in their decision.
Reflection & If We Had More Time
Quick and Dirty, Then Iterate
I tackled this project by developing new concepts and features step by step. Throughout the design process, I started small, validated ideas and functions with users, and then moved on to the next iteration. This approach helped me iterate effectively and shape the solution to what users actually need, quickly.
Business Point of View
As I designed, I aimed to maximize business outcomes as well as user outcomes while addressing real problems people face. This point of view helped me not only solve problems for users but also support business growth through design.
Future Iterations
I would run another round of usability testing to check whether the changes I made were effective. After that, I’d validate the risky assumption that “compatibility based on goals, past projects, and skills leads to well-formed teams and improved satisfaction.”
