On this page
Signing up for a free online course takes one click. Finishing it takes weeks of choosing it over everything else. If you have a trail of half-watched courses behind you, the numbers say you are in the majority. This post looks at what the research on massive open online courses (MOOCs) found about who finishes, why simple tricks often fail at scale, and a plan to complete the next one.
How many people actually finish online courses?
Very few. Katy Jordan collected public enrolment and completion data on early MOOCs in a 2014 study in the International Review of Research in Open and Distributed Learning. The average course enrolled about 43,000 students, and 6.5% of them completed it. Completion rates fell as courses got longer.
Later data were no better. Reich and Ruipérez-Valiente (2019) analysed six years of MIT and Harvard courses on edX in the journal Science. They reported that most learners never returned after their first year and that low completion rates had not improved. As Inside Higher Ed summarised, 3.13% of all participants completed their courses in 2017-18. Among learners who paid for a verified certificate, 46% completed.
Why is the completion gap so large?
Part of it is that many people never meant to finish. Some sign up to watch one lecture, or to look around. That is fine, and it inflates the drop-out numbers.
But the verified-learner figure shows something else. People who paid, and so had a clear reason to finish, completed at a much higher rate. Paying probably does not cause finishing on its own; it is more likely that those learners already had a goal and a deadline. The useful lesson is what that goal and deadline do: they turn a course from something you might watch into something you have committed to.
Do planning tricks help people finish?
Less than you might hope, which is worth knowing so you do not rely on one. Kizilcec and colleagues (2020) tested well-known behavioural interventions in PNAS across 247 online courses from Harvard, MIT and Stanford, with about a quarter of a million students. Self-regulation activities, such as writing a plan at the start, raised engagement in the first few weeks but did not raise final completion rates. Activities that helped learners connect the course to their own values raised completion for learners in developing countries, but only in courses where those learners were falling behind. The authors note that scaling up such interventions reduced their average effect by about an order of magnitude.
The message: a one-time plan at the start fades. You need structure that keeps working in week five.
A plan to finish your next course
Choose a short course first. Jordan's data showed completion falls as courses get longer. A four- or eight-week course is a better first target than a twelve-week one.
Write down why, in one line. "I want to understand linear algebra before my first semester" is a reason you can come back to in week four.
Book fixed slots, not "free time". Two or three named sessions a week, at the same time, with the course already open. Our guide to weekly time blocking shows how.
Do the weekly assignments on time. Many courses publish weekly deadlines. Treat them as real, even when nobody is checking.
Test yourself, do not just watch. After each video, close it and write the three main points or solve one problem. Watching feels like progress; recall is what you keep.
Commit to the exam. On NPTEL, for example, a certificate requires weekly assignments and an in-person proctored exam; the NPTEL FAQ says most courses need at least 40 out of 100 on both the assignment average and the final exam. Having that date fixed gives the course an end.
Find one other person. Agree to finish the same course and compare notes each week.
For free Indian options with exams and certificates, see our guide to free official learning resources in India.
What if the course moves faster than your background?
One common reason to stall is a lecture that assumes something you never learned. That is a gap in prerequisites, not a lack of discipline. Fermi Courses are university-level courses from working scientists, paired with a one-to-one AI tutor that works out what you already know, so you can fill a missing idea before moving on.
Pick one short course today, book its first three sessions in your calendar, and write your one-line reason on the first page of your notes.
