Designing Cohort Based Courses with High Completion Rates

Online courses are everywhere. It all started with the first well known MOOC (Massive Open Online Courses). MOOCs allowed the democratization of excellent lectures which was earlier confined to a select few students within the walls of the Ivy League(s).

There are two ways to deliver online courses:

• Self-paced: The entire course material is provided in a platform (typically a Learner Management System or LMS) and learners can go through it at their own pace. No live interaction is present.

• CBCs or Cohort Based Courses: The course material is delivered live and in sync. The course material is taught and consumed in a fixed schedule and facilitated by a mentor.

Some points w.r.t cohort vs self-paced.

• Cohort-based vs Self-Paced. Self-paced is suited for scale but very bad completion rates.

• Cohort-based is operationally heavy but better completion rates and outcomes.

I will talk about self-paced courses in a separate blog with my thoughts. But first, a fundamental question - why are the completion rates abysmal? In a transformative journey like education, we get value in the hard parts of the course. But once the course gets hard, we give up and that's why we have very low completion rates.

We must move from a compliance mindset ingrained in us with traditional education to a completion mindset - it is not about completing assignments but grasping concepts. This is facilitated by a great community of like-minded aspirants learning together. That is what CBCs enable.

What is a true course completion metric? Consider a course on programming. Which is a better outcome - completing the programming course, building an app/finding a job as a programmer. I agree that tracking this metric is difficult at scale. But acknowledging that this is a more tangible metric is a start.

There are 3 key phases in a CBC (in fact, any online course):

• Pre-course

• During course

• After course

Pre-course design

Keep the objectives of the course very clear. Who is it for and What will it help achieve? The clearer this statement, the better we would be equipped for the right design. A good mission statement would also help to design to help learners achieve this mission.

Don’t think just get a job - it must transform their life.

We need to now convey these objectives clearly to the learners who enrolled. A proper learner manual would help. Ideally, the first class must be pure expectation setting. Many courses hurry and try to cram instruction in the very first class itself. Without expectation setting, a course is set up for failure.

This phase is also a good opportunity to collect good data from the learners who enrolled. The learner persona that emerges can be matched with our hypothetical learner persona for whom we designed this course. If the learner persona is different, we need to see if some minor tweaks need to be made to accommodate the new persona. The more homogeneous the cohort, it is considered better, but I believe heterogeneity helps in diverse opinions which enriches the course.

Proper data capture is really important but highly underrated. (This is a different upcoming blog post)

During the course

We need to observe the learner’s behaviour in the early days of the course and compare with our assumptions as a course designer. Use the feedback to fix the course and encourage the expected behaviour. CBCs especially need multiple iterations to arrive at a steady-state. We need to embrace the chaos in the early days of the course and be patient. When adverse feedback comes, we need to connect with the learned; understand the pain and fix the issues. Proactively fixing issues turn disgruntled learners into our biggest ambassadors.

The learning journey is critical - we need to enable the learners to experience the delight moments fast - 1st assignment, 1st group engagement etc. Enable ice-breaker sessions to build bonds among the learners. Engagement design is key. More the students engage with each other and the instructor, better the chances of completion. Encourage feedback post sessions. Don’t keep huge forms. Take a simple questionnaire and connect with disgruntled learners later for further elaboration of the issues.

Enabling relationships is key for proper engagement. How fast someone makes a friend in the course directly correlates with their success in the given course? At the end of the day, course builders are community builders. Strong community leads to strong outcomes - when someone is stuck, the community helps the learners. Thus, a vibrant community is the true enabler of scale for CBCs.

Every course has an inflexion point. The point till where the majority of learners struggle and find it hard to cope up. Once they cross the barrier, it is smooth sailing thereon. Identifying the inflexion point and supporting the learner here is critical.

Prioritize what the learners need to do to achieve their outcomes. We might be tempted to load a course with a lot of information. We have to restrain ourselves and design the bare minimum first. Rest can be given as additional resources. Continuous experimentation on engagement, different types of assignments and interactive sessions is the key. Find what is resonating and amplify for the rest of the learners. Identify the learners’ jobs to be done and design to achieve those.

Post course

The learners have just completed the course. Post the course, we need to keep the community alive and ensure that the learners' outcomes are met. The outcomes for a programming course could be:

• Find a job

• Build an app

• Career transition

If we can enable the true outcome, that learner is the happy learner. Now, we need to tap them to get testimonials and if possible rope them as alumni mentors to interact with subsequent batches.

Collect feedback and NPS. A small caveat: While we listen to every feedback, we must be careful not to dilute our core values. For example, a learner might say that the course was too challenging. We mustn't think about making the course easier. The course must be challenging for the learner, but not confusing to the learner.

There is a lot more about an education business, I would like to share. All that and more in subsequent posts.

Additional References:

Running an online course, David Perell and Will Mannon

Tweet threads of Andrew Barry and Wes Kao.