Schools and the Research Business
Related to the core metric optimization around user generated content, the legacy Studymode UGC upload process allowed users to associate content with a School, Course, and Professor. The issue here is that the metadata acquired was not leveraged in any meaningful way and added a point of friction to that process. Seeing the percentage of historic content already tagged and normalized we set out to answer the question of if we could layer on a campus level organization structure to our existing research business?
After focuses testing we proved that the research business only wanted one thing....a document. This traffic profile was hyper focused on acquisition of a piece of content for an immediate need and was not engaged in creating a community layer on top of the existing research business. Looking at the traffic segmentations, only a small percentage of StudyMode traffic overlaid domestic cities where we has acquired previous content. What this showed us is that these schools showed a high likelihood to upload content but relative to the total traffic network being that much of our traffic was international schools would have to be launched in a different way. Knowing that users from our organic search based research traffic didn't overlay with domestic schools as well as were solely focused on research documents we set out to prove if we could organically acquire content at scale. Content is king on college campuses and we set out to create a community to provide a sense of belonging and participation between groups of people who otherwise wouldn't engage in a meaningful way.
Looking at competitors in our space there were problems with their current model of content acquisition at schools as there was no proven cost-effective way to acquire content at scale. Course Hero, OneClass, and Koofer’s (bankrupt) all employed a margin based business where they would pay students for content then resell as part of their core businesses at a target profit margin. This was both expensive with Course Hero spending several million dollars a year on acquiring content as well as questionable quality as interns for the company would try to cheat the system to optimize their earning potential with little effort.
Accessing our Users where they are engaging
To prove our hypothesis of being able to scale content without instituting a margin based business we set out to infiltrate existing Facebook groups. Here we located Facebook groups with high user penetration such as the referenced multiple University of California, Irvine groups or freshman level groups such as the San Diego State University Class of 2018.
To grow a school we need to optimize our funnel audience as we can't rely just on school enrollment numbers but instead the compiles user counts within these networks of groups. This allowed us a great way to target different schools. Our competitors were also in some of these groups but no competitor was using it how we were. This was a way to reach out to college students for FREE and acquire content for FREE as we were not reliant on Facebook ads which the demographic cognitively ignored and was a costly endeavor given that.
In integrating ourselves into these groups we were able to gain not only eyeballs into engagement prototypes we were creating but also to receive a general understanding of what students cared about or found important. We found that they were generally engaged in conversation around:
- Classes to take
- Professor ratings
- Housing options
- Financial Aid
- Internship Opportunities
- Study Help
Concept and Questions
After gaining access to these groups we surveyed it's members to find insight around the following questions:
- Is a centralized place to contribute and view documents good or bad?
- Will users upload documents w/o any material incentives?
- How do we validate quality of the content?
- What types of documents do students need on campus?
- Can engagement be ‘sparked’ organically via word of mouth?
- Does having a campus platform under the StudyMode corporate brand hurt?
To start engaging with these groups we targeted San Diego State University as our newly hired Business Development Manager had a network of friends still enrolled in which we could target a larger percentage of users based on mutual relationships. Here we sent a message to the groups asking for user to engage in launching a shared document drive via Dropbox with no corporate affiliation.
“Hello everyone, I made a dropbox account that we can all access. I am going to start uploading my class documents there so that way if any of you are taking a class I have already taken, you'll have access to my tests, quizzes, study guides, and notes. Feel free to share your stuff on here too. The more people we have adding things, ideally it will be able to help everyone whose apart of this page. This will be the new place to upload and view documents”
This post resulted in 60 immediate likes on Facebook but came with the caveat that we were still within the summer months. We found that a majority of students weren't accustomed to Dropbox causing some confusion as well as low engagement. We migrated over to Google Drive to use the same call to action and gain a higher percentage of users.
After a second posting a few students rose to the top and shared the post with their friends via @ mentioning other peers to join the group. We now had 10-15 students contributing 70 documents of extremely high quality. A few more posts down the line and we arrived at our AHA moment. After 4 posts we had almost 400 students requesting access to the Google Drive and 200 personal messages. We went from no engagement to students begging for access and having a hard time keeping up w/demand of users wanting access.
We expanded to a few more schools and realized that engagement varied across the groups at different campuses, some were receptive others were not. Funny enough the schools that engaged were those with a high percentage of commuter students, large enrollments, and schools a large extracurricular and sports presence. We invested deeper into these schools and set up to validate our next hypothesis, how can we structure these drives for growth and ownership?
Spinning up a Google Drive
After creating these Google Drives from a blank slate we were able to see natural behavior and organic organization patterns. We updated the model in which we asked students to send us at least 1 document to gain access to the Google Drive and the following happened in sequence.
- Empty Google Drive
- Students seeded the drive with unorganized content
- Students started to speak up and began a community policing of content structure
- Folders began popping up for user organization by subject
- In various separate folders students created their own rules of the road for best practice in organizing content.
- Folders were updated with specific course categorization instead of topic level.
With this organic growth in a matter of a week we had 350 members on the San Diego State google drive with 546 documents collected in a 2 week period. 385 students requested access within a single week and our Facebook posts promoting document count snapshots were receiving 150+ likes. In one day we received over 200 files where the StudyMode business as a whole was acquiring 1200 files a day on 200k individual sessions.
Hesitation Creeps In
As the drives started to scale and receive promotion on Facebook user hesitation started to creep in. Students were concerned with their anonymity in engaging with content that some institutions may find suspect. These institutions traditionally would bully some users in stating that collaboration between students was out of bounds but was a frivolous claim. Seeing the value of engagement we set out to invest in our own platform and migrate engagement away from third party solutions to be able to optimize for our use case. Much was learned in the meantime as we started to prioritize our own solution.
Learnings to date:
- Building a community is extremely important
- It is do-able and we will make it happen
- Exists on a micro-level w/what I have been working on
- It doesn’t matter how many documents you have, it’s about having the right documents
- Quality documents are crucial
- Great quality documents do exist
- There is a need for what we are going to offer
- So many students expressed how awesome this is and how needed it is
- Having it feel that it’s organic to students from students make it feel ‘grassroots’ and valuable
Scaling Campus Promotion Teams
As we scaled up internal development we set out to replicate at scale the Facebook posting strategy with a team of local interns. Analyzing what content engaged with students as well as time of day posting was critical to replicate known growth behavior. We hired a team of interns at each of our target launch schools and am massed a mobile network of 25 students across the country to manage promotions and access to their respective schools. On top of the mobile network we also hired a team of 8 in houser interns for profile management as well as posting analysis and strategy.
Pivoting from Google Drive Theory to our Platform
After 3 weeks of development and leveraging third party document management solutions such as filepicker.io for easy ingestion and view ability of content we were live in production. At launch we were able to hold our engagement rates of users joining google drive managed folders and the below engagement rates.
- User Acquisition - 40% of Sessions converted into an account
- Content Acquisition - 30% of users uploaded Content (85% of content coming from desktop sessions)
After starting with migrating the existing google drive schools to our own solution at studymode.com/Drive we scaled from 10 schools to 67 based on access to Facebook marketing channels and existing content within our library to seed content and entice users to join.
Such as we asked users to contribute to the Google Drive prior to attaining access this business model was baked into each of the schools launched on our internal platform. Here via the onboarding experience users are asked to provide their current course schedule allowing us targeting hooks to send users messages based on new users or content uploaded in their courses as well as upload content on a term by term basis to gain access under a "membership dues" model. This allows for term by term engagement hooks for users to get the most out of the application as well as incremental growth of content on an ongoing basis.
Departments and Course Metadata
Seeing the natural organization of content by users we invested in normalizing our department and course listings for users to which we expedited and enforced the organizational structure that promoted the highest engagement within the application. This such as studymode as well is a supply and demand business as when students join Drive and provide their current courses, long term success is pairing available content to user demand based on current course load.
At the individual course level Drive allows users to join specific courses to gain access to their peers in either an anonymous engagement or user selected public model. Here we promote discovery and sharing of content as well as a communciation layer between students who otherwise would never have met.
To date Drive has over 60,ooo users across 57 schools with over 40k documents all organically sourced via social channels for a marginal cost based on intern hourly rates. We've been able to scale this platform for minimal cost outside of your standard Facebook ad model which user acquisition was not viable for scale. Along with strong user acquisition rates engagement within the application has created a new product stack of on campus tools with a high touchpoint frequency and built for scale.