Cohort or self-paced online course: what to choose and why
Fixed-date cohort or always-on course? What the few studies say (46% dropout in an experiment on 231 teachers), costs, risks and how to choose the format.
In 2026 many course creators wonder whether to open fixed-date cohorts or sell an always-available course. The honest answer is that the evidence is thin, and the choice depends more on your time and your audience than on a universal rule.
A premise: when we speak of completion data, the sturdiest source we read is an academic analysis of massive open online courses (MOOCs); for comparisons between formats we found only one controlled study, from 2009, and a few platform pages.
What is the difference between a cohort and self-paced?
A cohort has the same start and end date for everyone: students advance together, usually with live meetings and a forum. Self-paced lets people start when they like and proceed alone, without deadlines imposed by the creator.
Each has pros and cons. A cohort creates shared commitment, allows feedback and concentrates sales in launch windows; but it requires your presence on set dates, and whoever misses the moment must wait for the next cycle. Self-paced sells all year and does not depend on your calendar; in exchange, without a deadline, many students postpone. A mixed format is possible, for example a recorded course with optional live meetings.
What do studies say about completion?
Jordan's (2015) analysis of 221 massive open online courses found a median completion of 12.6% of enrolments, with values from 0.7% to 52.1%. It does not compare cohorts and self-paced, but shows how low the share finishing a course is when there is no strong commitment.
The analysis also shows the first two weeks are critical: afterwards the number of active users levels off. That suggests a good start counts more than many other choices. But the data are from 2013 and earlier, for often free courses, with differing definitions of completion: they work as an order of magnitude, not as a target for a paid course. To work on the student's path, start from how to structure an online course.
Is there proof that cohorts work better?
What we read does not prove it. Russell and colleagues (2009) compared four conditions of an eight-week online course for teachers: 231 participants, with an overall dropout of 46%. Differences between formats were not statistically significant.
Dropout by condition
| Condition | Dropout |
|---|---|
| High support (instructor, facilitator, peers) | 53% |
| Facilitated peer support | 41% |
| Instructor only | 45% |
| Self-paced, no support | 44% |
The same authors found comparable learning outcomes across the four conditions, with a surprisingly higher dropout in the most supported group, though not significant. The study has evident limits: one course, middle-school teachers who signed up on their own, an asynchronous cohort and 2009 data. We cannot conclude that a cohort is useless; we can say we found no academic proof it always helps. We did not search every database exhaustively.
And what about course platform data?
Some platforms publish comparisons of completion for scheduled courses and self-paced ones, generally favouring the former, especially where there is active discussion. They are biased data, not controlled for who chooses the format, and we cite them only as a hypothesis to verify.
One such platform, for example, presents a study of 1.3 million enrolments; but we found two of its pages with different figures and could not read them at the source, so we do not report the numbers. Figures such as “85% cohort completion” circulate in marketing articles without a primary source: we disregard them. The caution has a precise reason: people who choose a cohort are often more motivated, and if they complete more the cause may be motivation rather than format. An honest criterion is to measure your own completion over time, whatever the format.
What hidden costs does a cohort have?
The main cost is your time, on set dates. A cohort needs live sessions, forum replies, feedback and a calendar to respect, and that limits the number of people you can follow in one cycle and the price you can charge.
The ratio of hours to revenue is the right question: if a cohort of twenty people takes forty of your hours, how much must you charge to repay that time? The same calculation for a recorded course that takes ten hours of work upfront is different, because the work is done once. To reason through the numbers read how much to charge for an online course. A cohort can justify a higher price because it includes access to you, but only if your presence is truly part of the result you promise.
How do you choose the right format?
With three questions: how much time you can give in person, how much your audience needs a deadline and whether the result requires feedback. If you answer yes to feedback and have the hours, start with a cohort; if the subject is repeatable, start self-paced.
A good path for beginners is to start with a small pilot cohort, say ten people, which helps you improve the material and collect testimonials; then record the course in final form and sell it self-paced with optional meetings. That way you have your own data on completion and on where students stop. To bring people to the course in launch windows use creator newsletters. And remember a deadline can be created without a cohort: a four-week challenge with hand-in dates gives a similar structure with less work.
Key terms. A cohort is a group of students following a course with the same dates. Self-paced lets people start and proceed when they like. Completion rate is the share of enrolled people who finish the course. A mixed format combines recorded material and fixed-date meetings.
How we checked this. We read the full Jordan (2015) analysis and the Russell and colleagues (2009) study and drew the numbers above from them. We found no academic studies proving an advantage for cohorts and could not read at the source the platform pages citing biased data, and we say so. Contact: hi@heytea.app. Editorial note: we fact-check every figure against the sources listed below and correct mistakes openly.
Frequently asked questions
What is a cohort-based course?
It is a course with the same start and end dates for everyone, in which a group of students follows lessons together, often with live sessions and a forum. It contrasts with a self-paced course, in which each person starts when they like and proceeds alone. Mixed formats exist.
Does a cohort make more people finish the course?
We cannot say for certain. The only academic comparison we read, on 231 teachers, found no significant differences in dropout between formats. Some course platforms publish higher completion figures for cohorts, but they are biased data, not controlled for who chooses the format.
Does a cohort take more work for the creator?
Generally yes. You need dates, sessions to run, forum replies and a calendar to respect, and your time is part of the product. A self-paced course is recorded once and sold many times, but it needs more care to keep students motivated without a deadline.
Can you run a mixed format?
Yes. A recorded self-paced course with a cohort window twice a year, or with optional live meetings, combines the advantages of both. It makes the organising harder, though, so start with one format and add the other after the first cycle.
How do I decide which format to choose?
Start from three questions: how much time you can give in person, how much your audience needs a deadline and how much you want to charge. If the result needs feedback and your presence, a cohort makes sense; if the subject is technical and repeatable, self-paced costs less and scales better.
Sources
- Russell, Kleiman, Carey, Douglas (2009), “Comparing Self-paced and Cohort-based Online Courses for Teachers”, Journal of Research on Technology in Education 41(4). Read on 16 October 2026. files.eric.ed.gov
- Jordan (2015), “Massive Open Online Course Completion Rates Revisited”, IRRODL 16(3). Read on 16 October 2026. irrodl.org