Marketing an app on a college campus (2500+ users in 2 weeks for $6)

If you’re like most creators, you develop software products with the hope that people will actually use and derive value from them. I sense that for many teams, driving trial and adoption can be as, if not more challenging than developing a product itself. This was certainly a concern of mine ahead of launching ClassAI to thousands of students at the University of Michigan. ClassAI uses machine learning to predict course availability well in advance of a student’s course registration appointment, enabling them to plan with more information and save time and energy.

Let’s get started with what a lot of people say: You need a good product. In my opinion they should say need an excellent product because I’m not sure ‘good’ always makes the cut. I don’t have a clue as to how to sell and market a mediocre product, but I assume it’s tough. From a small group of testers, we were feeling confident we had a great product, so every decision and action we took was with that assumption.

Cultivating a Campus Brand is Cheap but Valuable

Companies deploy varying commitments to branding and cultivating an identity. Some believe a brand to be more important to ‘selling’ than others and for good reason. Developing a national or international brand is incredibly challenging and bloody expensive, even if the impact of having one is substantial.

On the other hand, on a college campus, the path to developing a brand is incredibly accessible and here’s a few reasons why. On a college campus all students can be reached by being present at just 2–3 locations. Tens of thousands of students walk through Michigan’s Diag each day. Campuses also have incredibly high word-of-mouth (WOM) activity. Also, colleges are comprised of a pretty similar group — educated young adults, almost all of whom are within a few years of one another. Although you can still (and should) employ practices of segmentation, targeting, and positioning, whatever segments you create will still undoubtedly be much more similar than they are different. As a result, it’s incredibly easy to strike meaning with the large group, and for that effort to reach all corners of campus.

The Diag at the University of Michigan

The benefits of a brand on a campus yet, are very similar to those of a national brand. A visibly present brand (posters, leaflets, chalk, t-shirts) does wonders for awareness and can spur general interest. A present brand also helps publicly define and reinforce a simple and compelling message that travels fast through boosted WOM. Last, a visibly present brand ensures better recall and recognition for when the time comes that a person decides to try your product.

Focusing on a Simple, Consistent Message

We worked through many options before arriving at a simple, direct, and memorable message that would define ClassAI’s brand and the product’s core functionality.

“What’s the chance you get into {class X}?”.

The 7 word question on its own wasn’t all that effective, but when combined with different course options, it immediately became understandable to observers. The question was highly effective for many reasons, not the least of which was because it’s a question that has been considered by most students at one point or another as they plan their schedule preferences as their registration appointment approaches.

Opposed to using a single course, by alternating the course, in Slack’s famous style, the key function of the app was suddenly explained.

Landing page of http://class.ai

This method obviously doesn’t work on paper, so for our paper promotions around campus we simply put posters in clusters that displayed different classes to communicate that you could throw in any class.

Snippets from posters posted around the University of Michigan campus

This theme continued to many of the other channels we leveraged to increase awareness and build a brand centered around our signature question for students. For example, here’s the header of one of the email templates we employed.

Snippet from an email sent from ClassAI

Another key leg of our marketing campaign was creating ‘ClassAI Hall’, filled with ClassAI’s message, logo and fun smaller messages poking fun at the some of the other less-than-stellar scheduling tools students use. Sticking to the same message, but posed differently (to sufficiently cover the space), we covered the cross-sections of the stone ground with different classes. By the end of the 20 second walk, students had the same experience as viewers of our landing page.

ClassAI Hall, a 2.5-hour chalking project created near the heart of Michigan’s campus in a highly trafficked corridor

Segmentation, Targeting, and Positioning

We saw firsthand the boost in impact we were able to have in convincing students to try ClassAI by identifying several segments, strategically targeting them, and alternating the positioning of our product in those targeting efforts accordingly.

Our two broad segments were underclassmen and upperclassmen. Since registration appointments are assigned roughly in order of degree credits earned, upperclassmen typically have the earlier appointments, while underclassmen are less privileged in that they are at the disposal of the thousands of students who register for their chosen classes before them. Underclassmen, typically as a result have to reshuffle their schedules a bit depending on popularity of their field of study as course sections get filled up in the days before their appointment.

While underclassmen were our predominant focus, due to the fact that they’d derive the most value from using ClassAI, we also targeted upperclassmen with a slightly different message.

Upperclassmen, as a result of their earlier registration times, more frequently have the the mindset — justified or not — that “of course, I’m going to get into this class”. So appealing to this end, we positioned ClassAI somewhat distinctly to these users, asking them “wouldn’t you like to make sure that you can definitely get into this course?” Through this message, we positioned ClassAI as a tool that could validate and confirm their existing knowledge of the system, compared to say freshman, who want as much help as they can get and lack institutional knowledge about how registration works.

This messaging was definitely more welcoming to this segment and reduced barriers to trial for this group, which due to the high WOM, we think helped spread ClassAI even further (a junior tells their freshman sister about the app).

Additional Channels

Our marketing strategy consisted of several channels, in addition those mentioned above. Here’s how we used some of them.

Social Media—Our user-centered social media efforts focused on being an source of information for our ClassAI’s most loyal and interested fans. Through social media, we communicated major milestones and feedback from users about the ways that ClassAI had empowered them.

We also deployed a special “I schedule with ClassAI” overlay to friends’ profile photos, which through tagging, and sharing, boosted reach and awareness.

A selection of graphics from ClassAI Facebook posts

In Person Conversations—Over two weeks we held nearly 75 conversations with students to communicate to and spread awareness about ClassAI and help onboard new users. These conversations were incredibly open-ended and lasted anywhere from 1 to 30 minutes, for the more curious folks. Generally we think that the students who met us in person had even more interest in the project and further boosted WOM. Aside from driving more WOM, it was tremendously valuable to hear from users directly and see how they responded to different messages, and to observe their instincts and pain points as they registered and explored the app.

This is certainly the type of work many call ‘customer discovery’ and do before they dive in to build a product. For our three month timeline, we didn’t have time to do that, but it was nonetheless very helpful to do at launch.

Campus Newspaper — After about a week after launching, The Michigan Daily, the university’s main student run newspaper heard about ClassAI and covered the app in an article. Referral analytics show that over one hundred students signed up through the newspaper’s article to our landing page, which certainly didn’t hurt.

Intersection of Marketing and Launching

Another cause for ClassAI’s success in my view was our strategy to rolling out and steering adoption by students. This can be examined through Everett Rogers’s categories of adopters, as described in his Diffusion of Innovations. In short, Rogers identifies the five types of customers that comprise the adoption cycle for new technologies: Innovators, Early Adopters, Early Majority, Late Majority, and Laggards.

While I won’t jump into fitting ClassAI users into each of these segments, I will mention one, which we feel was incredibly important to our successful launch. A few days before launch, we seeded word about ClassAI to the hardcore, experienced members of the computer science community at Michigan. These students, hungry to try new technologies, provided feedback and proved keen on finding and reporting bugs early. This group of less than one hundred users was incredibly invested in helping us fix backend issues and make small changes to instructions and tooltips throughout the site, and thus, served as our Innovators.

Oh, and if you’re still wondering what we spent that $6 on, it was actually $5.97, spent on chalk :D

Also shoutout to Tyler Laredo who was daring enough to embark on this journey with me :)

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.