Helping Students Find Their Fit During a Pandemic

Problem Statement

Discovery Was Front-Loaded, and Students Couldn't Catch Up

It took me a while to find my place during my first year of college. I became an expert in both joining and leaving communities before I found where I was happiest. Finding my place took a lot of luck and perseverance.

The one mechanism built to fix that was an in-person club fair during the first weeks of fall semester. It worked well for the students it worked for, but it front-loaded all discovery into a moment when most first-years were distracted by classes, new friendships, and the general chaos of starting college.

It also left no good path back in. Students who didn’t find their fit the first time, or whose plans changed later in the year, had nowhere to turn. When COVID hit, the fair was cancelled outright, and the informal, serendipitous ways students normally found community (walking back from the library, a flyer in a campus-wide group chat) had no virtual equivalent.

We knew this problem intimately from our own experiences, and validated it by talking to dozens of other students.

User Research and Analysis

Finding Your Fit Was Left to Chance

I’d lived a version of this problem myself. I met my eventual co-founder through track, and we’d both gone through that spring scramble of trying to find a new community after our first-semester groups hadn’t worked out. So going into this, I already had a strong hunch about what was broken. But a hunch isn’t research.

Over about a month at the start of the school year, I ran around a dozen exploratory interviews with first-year students, formal interviews with club leaders, not to mention the countless informal conversation I had with just about anyone to better understand the issues. I also conducted a competitive analysis against GroupMe, which despite being a messaging app, was the dominant way that clubs advertised for events. We synthesized all of it as a team through How Might We sessions, dot-voting to prioritize which problems to actually go after.

Discovery Was More Serendipity Than System

Most students found clubs through word-of-mouth and mutual friends, not anything resembling a deliberate search. Students who made close friends in their first couple weeks had an easier time, because those friends became a pipeline of invitations and reassurance. Students who didn’t get that early win tended to stay stuck. Showing up to something alone is a much bigger ask than showing up with someone you already know.

Where Discovery Broke Down

Across the interviews, the same few problems kept coming up. Each one made it harder for a first-year to find a club, or for a club to find them.

Design and Prototyping

Aligning Discovery to How Students Actually Behave

We didn’t land on the final design right away. We went through four real directions before settling on one, usability testing each with three to four students per round.

, open full size

Design sprint, 2020

, open full size

Wireframes to expo build

Leading With Clubs

Events were often the way in, but clubs were where the community actually lived. So we built V1 around them. A club-first layout felt familiar to club leaders, and we needed their buy-in. It also mirrored how students already found things: through a group first, then its events.

Clubs didn’t struggle to reach their own members. GroupMe and private group chats already kept members up to date on events. The hard part was getting students to know a club existed at all. Campus became the front door that funneled students into those chats, and let the event channels clubs already had do the rest. We planned to add event-driven discovery that led back to clubs later, but this was our starting point.

Rollout and Adoption

Partnering for the Club Fair

While we were iterating, we started talking with our student engagement office, sharing research and comparing notes on how to improve the first-year experience.

When the COVID pandemic hit, and our schools campus closed for in person events for over a year, this relationship proved invaluable. Up until this point we had been planning for a launch that centered around in-person community and giving students the knowledge and confidence to show up to events. COVID forced us to pivot overnight, as any semblance of in person community was evaporated. The office of student engagement kept us in the loop early about how the administration was handling the pandemic shared research with each as we both scrambled to adapt.

They’d already been evaluating outside vendors for a virtual club fair, and we eventually won that bid, which became our actual launch moment. It gave club leaders a concrete reason to sign up and guaranteed a huge amount of visibility and students signing up.

Hank, who joined as our third co-founder in 2020, designed many parts of the fair experience, like the livestream.

, open full size

Expo planning board

The fair ran on Monday, Aug 24, 2020, from 4:15 to 6:15 PM. Before it started, students signed up and told us what they were into, and clubs set up their own pages, which our team reviewed before they went public.

Bringing an In-Person Feel to a Virtual Fair

The best part of an in-person club fair was all the wacky things club leaders did to show off their club’s personality and pull people over to their booth. We wanted to keep that feel, and the livestream was one of the ways we did it.

For the two hours of the fair, we ran a livestream co-hosted with the Office of Student Engagement. It streamed on YouTube and played picture-in-picture in the app, so students could keep browsing while they watched. Clubs got their moment on stream, and a lot of them leaned into it. The Outing Club ran a virtual hangboarding challenge, and other clubs came up with their own goofy bits to make it as fun as possible.

Alongside the livestream, every club ran its own Zoom room, linked from its page in the app, so students could jump in, say hello and get to know people in the club. Clubs could also post short videos, which played in a scrolling “virtual walk” feed inside the app. This feed mixed clubs matching a student’s interests with ones slightly outside them, to recreate some of that in-person serendipity on purpose.

Outcomes and Impact

A Virtual Club Fair That Moved the Needle

The club fair worked. We had over 2,000 signups and 220 clubs registered in the app, with over 1,500 active users at the time of the fair. For a school of about 5,500 undergrads, that was a large share of the first-year class, who the app was mainly marketed to.

Club leaders’ Zooms were packed for the whole livestream, and students found communities they might not have known about before.

But Zoom Fatigue Was Real

Despite our best efforts, the in-person shutdown made it really hard to carry the momentum of the club fair forward. Zoom fatigue became a real issue: after spending all day in classes on Zoom, students didn’t want to hang out online too, and many clubs started to struggle with retention.

Most of our plans to re-engage students felt futile in an environment that didn’t let community exist in the first place. For safety reasons, many students simply retreated into their own insulated bubbles and strict campus guidelines forced in-person gatherings underground.

Key Takeaways

A Lesson in Validation Speed

We began working on campus late into the spring of my freshman year. Between classwork, many rounds of design and iteration, and the task of programming the app it took our team of two nearly a full year to be ready to launch with a product we were confident in. And then COVID came out of nowhere and completely changed our plans.

While I’m proud of how our team adapted to the pandemic and found a way to still make an impact, I’m disappointed that we couldn’t make an impact sooner. Ultimately, campus never fully returned to open and welcoming culture that had existed before the pandemic by the time we graduated. So we never got to see how our app might have enabled human connections that might never have occurred before it. We were too slow.

What I’d Do Differently: Run Campus as a Group Chat First

Everyone already had GroupMe. Clubs shared news there, and there were chats like the free food chat. It was a clunky medium, but its casualness fit the word-of-mouth culture, and the barrier to entry was very low.

So I’d run the app’s functions manually through a group chat. Our team would gather the events and post them. Then we’d try different framings. Post days in advance, or last minute? Lead with need a study break? Here are three low-stakes things you can do this afternoon? Because people could act on each post live, we’d see which framing actually got them to go. Then we’d launch the app to the people already in the chat.

We could have done this weeks before we ever opened Figma or VSCode. That would have given us real behavioral data, which was impossible to get from usability testing.

Use the right research methodology for the question at hand

ln testing, I asked students to find an event they’d want to go to, and one outside their usual comfort zone. They could find events, and some said they could see themselves going. But the question we were really trying to answer was whether a student would check the app when bored, and then actually go. A usability test can’t answer that.

The only reliable way to know what people will actually do is to observe them. Simple real world experiments would have told is more in a week than months of testing would have.