Same Course, Two Versions. One Finished at 65%. The Other at 43%.

The bar for counting as a "community course" in this data set is one student comment.

Not a live cohort. Not a Slack channel with a moderator. One comment.

That's the line Ruzuku drew when it segmented its production database for its 2026 reports, and the gap on either side of it is the most interesting completion number published this year. Courses with active community discussion hit 65.5% completion. Courses without hit 42.6% — a 54% improvement.

The scale behind it is real. The report covers 32,000+ published courses, 853,572 unique students, 1,826,838 enrollments, 9,140,636 comments and $78,866,538 in platform revenue.

Why this number lands differently

Because the completion figures creators throw at each other are mostly borrowed from a different business.

Ruzuku's own report puts MOOC completion at 3-6%, citing Class Central, and self-paced courses on independent platforms at 10-20%. Zenler, covering the same ground, cites a Class Central-summarised MOOC average of roughly 6.5%. Elsewhere Ruzuku quotes a median MOOC completion rate of 12.6%.

So the industry's favourite doom stat ranges from 3% to 12.6% depending on which page you land on — and none of those courses are the thing you sell. Zenler's read is that many MOOC learners enrol casually, pay nothing and never intend to finish, which means MOOC completion shouldn't be the benchmark for a paid specialist course.

Strip that out and the 43% figure becomes the real number to be uncomfortable about. That's a paid course, on a creator platform, with more than half the buyers walking away.

Completion is a design problem

The platform's own framing is blunt about what doesn't move the needle. Ruzuku says courses with discussion features see 65% completion versus 43% without, calls it the single biggest lever found in fourteen years, and argues course length matters less, video quality matters less, even price matters less — add a discussion element and completion roughly doubles.

That is an awkward finding for anyone who has just spent six weeks upgrading their lighting.

The niche-level data points the same way. In yoga courses, cohort-based formats beat on-demand by 24 percentage points in median completion, 69.2% versus 44.9%.

And the money follows. Ruzuku says creators offering community features earn roughly 2x more than those without.

The caveat that matters

This is one platform reporting on its own customers, and it sells community features. Read it with that in the frame.

To its credit, the report says so itself. The segmentations are described as observational, not experimental — associations in the data, not randomized controlled trials. It also flags that its creator base skews toward professional service providers: coaching, health and wellness, creative arts, professional development.

Causality is genuinely murky. A course with a commenting student may simply be a course with a better instructor, a tighter promise and a more committed buyer. The comment could be the symptom, not the cure.

But the direction is hard to argue with, and the cost of testing it is close to zero. Open the discussion. Answer in it yourself. Watch week one, where one 2026 analysis puts 50% of dropouts.

Refunds don't start at the refund request. They start the week a student stops showing up and nobody notices.