Coursera Case Study
A challenge to increase the number of Coursera power users.
I'm Albert, a design technologist.
I'm currently advising early-stage startups on their design strategy through Dorm Room Fund. Previously, my work has mostly revolved around supporting first-time founders with tools such as VCWiz and FounderKit.
On the side, I do freelance work for brands including Royal Carribbean, Alpha Bridge Ventures, and LimeBike.
My core design focuses are in digital product design with an emphasis on product ideation, user experience (UX), and sprint frameworks.
Sector-wise, I'm deep into fashion, wellness, and AR/VR
Hi, I’m Albert and I’m a MOOC addict.
Massive Open Online Courses, or MOOCs for short, are something that have been a big part of my life for a while, my interest starting when I first took Introduction to Genetics and Evolution in high school. Nowadays, I often take multiple courses at once on topics that spark my interest.
However, after chatting with a few friends who were also taking MOOCs, I realized something interesting — I was one of the few that ever took more than one MOOC at the same time. This became the inspiration of my design challenge:
How do we increase the number of courses a user takes on Coursera?
So why is this important? My initial hypothesis was: more courses taken per person = more engaged users & more conversions from free-to-paid courses.
Timeline: 5 days | Tools: Sketch, Balsamiq
I started off by trying to get a deeper understanding of Coursera’s platform and business objectives -- going through the entirety of the site, reading various blog posts by Coursera, and watching interviews on Andrew Ng & Daphne Koller talking about the product.
From there, I went on to get other users’ opinions.
Through these interviews, I discovered a few critical things that would heavily influence the direction of this project.
I wanted to learn more about why many of them felt confused navigating the site, so my first step was to sketch a quick sitemap to better understand the site flow.
The sitemap was a start but I still needed to know more.
To address that issue, I messaged three of the previous interviewees to do a quick usability test. I tasked them with finding courses that matched certain criteria and had them speak their thoughts out loud as they perfomed the task.
Through these three interviews, the key takeaways were:
Since I was working on a limited timeline, I decided to move on towards brainstorming possible solutions.
If I had more time however, I would have liked to:
After a day of brainstorming potential solutions and some competitor research, I came up with three ideas to improve Coursera’s user experience.
Simplified Site Flow
Initially, I outlined the site flow as a way to help me learn more about the platform, but after thinking through it further, I hypothesized that a simplified site flow will benefit the user as well.
As such, I redesigned the flow with an aim to to reduce the amount of friction it takes to search for a course and reduce the cognitive overhead when navigating between and within courses.
How would this help? Why would you want to take more courses if there's a lot of friction in finding new ones and if taking new ones messes with your existing workflow? I hypothesized that by simplifying the user flow throughout the site, users would be more open to adding new complexities (courses) to their user flow.
How can we measure success? Measuring the impact of this redesign alone would be difficult due to the multiple confounding factors resulting from the other changes necessary to implement this new sitemap. However testing is possible if each individual alteration to the sitemap is tested independently from each other but as a package with their corresponding necessary visual redesigns.
Empowered Search Experience
My first attempt to redesign the search experience was creating a watchlist feature, where users could save courses for future reference and choose to receive notifications when a new session would start for each course. After tinkering a bit more with the platform, I realized that because there is no feature available to save courses, users are forced to enroll in a course in order to save it. Because they do this while browsing for courses, it spreads out the number of steps between discovering a course and paying for one.
As a result, I scrapped that idea and included a “+” and an “Purchase” button to the course card. The “+” button enrolls the user in the course and adds it to their dashboard in one click. The “Purchase” button will prompt a pop-up giving them the option to purchase the course and recieve an certificate upon its completion. In the case of a specialization, the "+" button is removed and the "Start Trial" button, which starts their 7-day free trial for specializations, replaces the Purchase button.
In addition, using the insights gained from listening to the users, I revamped the list of available filters as well as redesigned what information could be gained with a glance at any course’s card:
How would this help? If it becomes easier to discover and save interesting courses, then the average user will take more of them. Spreading out the amount of friction an user has to face in the process of finding a course to starting a course will increase the completion rate of that process.
How can we measure success? A/B testing the old course search page vs. redesigned course search page with at least the 2 metrics:
I also had some hesitations regarding the convoluted dashboard screen and the lack of consistency on it. Why does only the most recent specialization course get the detailed screen while more recent non-specialization courses and non-recent specialization courses get the bare minimum? As a whole, the dashboard was lacking a well-defined hierarchy with the specializations visually overpowering single courses and making it confusing to see where one course ends and another starts.
The redesigned dashboard works with the previously mentioned features to entice the user to hoard courses that they may potentially take in the future. The structure is also more organized with a lesser visual disparity between specialization and non-specialization courses. Purchased courses are also emphasized and higher in the stack order to give them a feeling of more importance as well as making it more convenient for the user to access. I hypothesize that visually highlighting this difference would lead to greater conversions from free to paid courses as well.
How would this help? Simplifying the dashboard hierarchically and visually for multiple courses would make the user more open to having multiple courses on their dashboard and less likely to remove them. Visual clutter = increased perceived complexity. Taking multiple courses should not feel like a complex task.
How can we measure success? A/B testing the old dashboard vs. redesigned dashboard with at least the 3 metrics:
A Quick Note...
Something else I've noticed with my time on the Coursera platform is that the team is testing new features and new UI designs very frequently. I hypothesize that this has lead to many inconsistencies and over-complexities throughout the platform. However, I am hesistant to make a comment on that practice without ample data to support my stance. Perhaps in the long-term, the marginal benefit of each iterated decision outweighs the detriment of the short-term inconsistencies.
This was a fun challenge albeit its difficult due to the number of inconsistencies throughout the platform. As such, I tried to view each feature as its own independent entity that will be user tested as that's the mentality I imagine the designers at Coursera to have. Key points I learned through this process were: