🎓
This one's password protected!
Drop me a message and I'll share the password with you. I'd love to walk you through it! 💛
🔍
📚 🎓 💡
📱 EdTech · Mixed Methods · 3 months

From Ghost Town
to 30,000 Users

A student app had 80,000 people who could use it. Only 5,000 did. The CEO faced a tough call: pump more money into marketing, or invest in research. He chose research. Then he hired me, and I built the team that figured it out.

My Role
Lead UX Researcher
Timeline
3 months
Methods
Surveys · Interviews · SQL · R · Usability Testing
80K
Students who
had access
5K
Actually
showing up
723
Survey
responses
30K+
Users after
we shipped 🎉

The Problem

A free study app nobody was using 🤔

Universities were paying for this app for its students so they could use it for free. But most students who tried it didn't stick around past a couple taps.

Why this was a big deal
Schools were paying real money for this app. If students weren't using it, schools wouldn't renew their contracts. The whole business was at risk, not because the product was bad, but because students weren't getting far enough to find out if it was good.
What the team thought was wrong
The team had a hunch: maybe students didn't know about the app. Or... maybe it was confusing? Slow? Missing features students needed? Nobody knew for sure, and that's where I came in.
🎯 My job
Talk to students. Find out what was really going on. Turn those insights into changes that would actually make people want to use the app. I also hired and led a team of 3 junior designers and 2 UX-focused iOS engineers to help bring those changes to life.

The Team

I got to build something special 👩🏽‍💻

One of the best parts of this project was getting to hire and lead a team of junior designers. Watching them grow was genuinely one of the highlights of my career. Several of them went on to land 6-figure design roles. Pretty cool.

Team Responsibility
1 CEO Key decision maker
1 VP of Engineering Scoping and feasibility
1 VP of Success Connect to users and budget
3 UX Designers Hired by me Craft wireframes and prototypes
2 Success Managers Support end-to-end research
2 iOS Engineers (UX-focused) Hired by me Build designs in mobile and web app
1 UX Researcher (me!) Lead end-to-end research

Where I Started

First, I mapped the stakeholders 🗺️

Before I talked to a single student, I needed to understand the people around this product. Who makes the decisions? Who has a stake in the outcome? Who am I actually designing for? Mapping this out helped me see the whole picture and ask better questions from day one.

🎯 Why this matters
A stakeholder map isn't just a pretty diagram. It helped me understand what each group cared about, where their priorities might clash, and who I needed to keep in the loop. It also helped me scope the research and sharpen my user questions, so I was spending my time on what mattered most to the people who'd act on the findings.
Tertiary Secondary Primary ↑ Outside org ↓ Inside org
University
Students
Primary users
Community
College
Students
Primary users
High School
Students
Primary users
CEO
Key decision maker
College
Admins
Institutional buyers
School
District
Leaders
Institutional buyers
VP of
Engineering
Scoping & feasibility
Success
Managers
Connect to users + budget
VP of
Success
User outreach & retention
VP of
Marketing
Adoption & growth
UX
Designers
Tertiary stakeholder
iOS
Engineers
Tertiary stakeholder
Primary
Secondary
Tertiary

Hover over each stakeholder to see their role 💛


Research Process

3 months, 8 research phases, real answers 🔬

I used a mix of methods to understand the full picture. Some methods helped me see the big patterns. Others helped me understand the real stories behind those patterns.

Phase 1
Stakeholder Interviews
I started by talking to the people who built the app, the CEO, engineers, and success team. I wanted to know what they already knew, what they were guessing, and what data they already had sitting around (like session recordings in HotJar and usage stats in Metabase).
Phase 2
SQL & User Segmentation
I queried the app's database using SQL to look at how students were actually using the app, how often they logged in, which features they touched, how long sessions lasted. This helped me sort users into three groups: Power Users, Active Users, and Inactive Users.
Phase 3
Survey Research
I sent a survey to 817 students and got 723 usable responses. I asked about how they study, what features they use, what frustrates them, and what would make them come back more often.
Phase 4
Data Analysis in R
700+ survey responses is a lot of data! I used R to clean the data, remove duplicates, and find the patterns that mattered. I also ran some stats to make sure the things I was seeing were real signals, not just noise.
Phase 5
User Interviews
I talked to 20 students, across all three user groups. These conversations went deeper than a survey ever could. I got to hear the real stories: what they were doing instead of using CircleIn, what made them give up, and what would actually bring them back.
Phase 6
Design Sprint & Co-Creation
Together with a junior product designer (her first real project!), we took all the research and turned it into new designs. We focused on three things: fixing onboarding, improving video chat, and adding a new feature students had asked for.
Phase 7
Usability Testing (2 Rounds)
Round 1 was moderated, I guided students through tasks and watched where they got stuck. Round 2 was unmoderated, students completed tasks on their own, which helped confirm the fixes were actually working.
Phase 8
Heatmap Analysis
Using HotJar, I analyzed heatmaps and session recordings from real student interactions. This showed exactly where students were clicking, scrolling, and dropping off, giving us a visual layer of evidence to back up what students had told us in interviews.

Methods

A lot of tools for a complex problem 🛠️

🗄️
SQL & Data Querying
Pulled raw usage data from the database to understand real behavior, not just what students said they did, but what they actually did.
📊
Survey + R Analysis
Broad survey to 817 students, then R to clean and analyze 723 responses. Found the patterns that showed up across different types of students.
🎙️
User Interviews
20 in-depth conversations with students across all three user groups. The "why" behind every number in the data.
🔭
HotJar Session Analysis
Watched real session recordings to see exactly where students got confused, clicked the wrong thing, or just gave up.
🎨
Design Sprint
Co-created new designs with a junior designer, turning research insights directly into prototypes ready for testing.
🧪
Usability Testing
Two rounds of testing, moderated and unmoderated, to make sure the new designs actually solved the problems we found.

Survey Results

What 723 students told us 📊

The survey was my widest net. It helped me see the big patterns across all kinds of students before I zoomed in with interviews. Here's some of what came back.

Survey results chart showing student preferences
Add your caption here
Survey results chart showing feature usage
Add your caption here

Who We Were Designing For

Not all students study the same way 🎯

The SQL data and interviews showed three very different types of students using the app. Each group needed something different, and if the app wasn't serving at least one of these needs really well, students left.

🤝
The Collaborator
3 in 5 students · Most common type
What keeps them coming back
  • Study groups and shared notes
  • Connecting with classmates in their course
  • Chat and real-time collaboration
  • Not feeling alone while studying
🏆
The Competitor
1 in 5 students · Loves to win
What keeps them coming back
  • Leaderboards and badges
  • Seeing how they rank against classmates
  • Streaks and challenges
  • Public recognition
🗂️
The Organizer
1 in 5 students · Quietly consistent
What keeps them coming back
  • Flashcards, notes, and to-do lists
  • Scheduling and deadline tracking
  • A clean, easy-to-navigate interface
  • Customizable study tools

Key Findings

Five things that explained almost everything 💡

1
Onboarding
Students had no idea what the app was for when they first opened it.
The app just... dropped students into a dashboard with no explanation. No tour, no tips, no "here's how to get started." Students said they wanted short video tutorials that showed them the best features right away. Without that, most people just left before they ever found anything they liked.
2
Features
The features students wanted most weren't working well enough.
Students really wanted better flashcards and a reliable chat. Those were the two most-requested things. They also wanted to customize their experience, like filtering content by class or professor, but the app was too rigid for that.
3
Motivation
Students came back for competition, connection, or organization, not just features.
The data was clear: students who kept using the app were getting something specific out of it. Either the competitive streak kept them logging in, or they felt connected to classmates, or the organizational tools fit their study style. Apps that didn't deliver at least one of these things deeply got abandoned fast.
4
Usability
Small broken things were driving students away for good.
Login issues on mobile. A navigation structure that didn't make sense. Features that were hard to find or didn't work reliably. None of these felt "small" to the students experiencing them. One bad moment was enough to make them give up and go back to Quizlet.
5
Competitor Insight
Almost 80% of students were already using Quizlet, mostly for flashcards.
The survey made this really clear: flashcards were the number one study tool students relied on, and most of them were going to Quizlet to get them. CircleIn had a flashcard feature, but it wasn't good enough to pull students away from what they already knew and loved. That finding went straight into the product roadmap. We upgraded the flashcard experience and made it a priority.

In Their Own Words

What students actually said 💬

"I didn't even know what half the features were supposed to do. I just kind of gave up and went back to Quizlet."
→ Students were leaving before they even got a chance to like the app.
"If I could see where I rank against my classmates I'd probably check it every day. That would actually motivate me."
→ Gamification wasn't a "nice to have", for some students, it was the whole reason to come back.
"The chat keeps cutting out and I can never find where my notes went. It's honestly more stressful than helpful."
→ Reliability issues were breaking trust faster than any missing feature could rebuild it.
"I want to study with my friends in my actual class, not random people. Can I filter by course or professor?"
→ Social features only felt useful when they connected students to people they actually knew.

Heatmap Analysis

Seeing exactly where students got stuck 🔥

Surveys and interviews told me what students thought. Heatmaps showed me what they actually did. Using HotJar, I could see where students clicked, how far they scrolled, and where they dropped off. It gave us a visual layer of proof to back up everything students had told us.

HotJar heatmap showing student click and scroll behavior
Add your caption here

The Designs · Onboarding

Fixing the very first thing students saw 👋

Students were getting dropped into the app with zero guidance. So onboarding became priority number one. We redesigned it with a friendly welcome, helpful tooltips, and a clear path to the features students actually cared about.

Redesigned onboarding flow for CircleIn
Add your caption here

The Designs · Flashcards

Giving students a reason to stay 🃏

Since almost 80% of students were leaving to use Quizlet for flashcards, we knew this was a make-or-break feature. We upgraded the flashcard experience to be something students would actually choose to use inside CircleIn.

Redesigned flashcards feature for CircleIn
Add your caption here

Interactive Prototype

Bringing it all together in Figma 🎨

Here's the interactive prototype we built and tested with students. Click through to see the new onboarding, flashcards, and features in action.

Replace the src with your Figma embed link


What We Did About It

Research that actually shipped 🚀

The findings went straight into designs, got tested with real students, and shipped as new features, all within 3 months.

What We Recommended (and Built!)
1. Fix onboarding first, Add video walkthroughs and feature tours so students know what they're getting before they hit the dashboard.

2. Make flashcards and chat actually great, Invest in the two things students asked for most, and make them reliable.

3. Add a gamification layer, Leaderboards and badges tied to specific courses, not just the whole platform.

4. Fix the broken stuff, Login bugs and confusing navigation had to go before we could add anything new.

5. Let students find their people, Course-level filtering so students could connect with classmates in their actual classes.
What Happened 🎉
One week after launch, with a marketing email series I helped design with the VP of Success, the app jumped from 5,000 users to over 30,000. That's a 6x increase. The research gave the whole team a shared, clear picture of who students were and what they needed. No more guessing.
📱
Product Type
EdTech Mobile + Web
🔬
Research Phases
8 Phases
👥
Participants
743+
📈
User Growth
5K → 30K+ 🎉

Reflection

Glows & Grows ☀️🌵

☀️
Glows
One of the best parts of this project was mentoring a junior designer, straight out of college!, through her first real project. Watching her grow and co-create solutions with me was genuinely one of the highlights of my career. I also got to write my first formal UX research budget proposal, which came in super handy for recruiting and incentives.
🌵
Grows
My first survey email went to 3,000 students and got 10 responses. Ouch. I quickly added a $250 AirPods incentive in a follow-up email, and within 12 hours, we had 500 responses. Lesson learned: always, always think about what motivates participants to show up. Incentives aren't optional, they're part of good research design.
How did this land?

When you listen to students,
you find out what they actually need.

This project showed us that low adoption wasn't a marketing problem. It was a "we don't understand our users yet" problem. Once we understood who students were and what was getting in their way, the product decisions got a lot clearer, and the results spoke for themselves.

← Back to all work