---
title: "Online Course Completion Rates: Free vs Paid vs Cohort — Wayflo"
url: https://wayflo.com/blog/online-course-completion-rates
description: "Free MOOCs finish 3% to 13%, paid courses around 46%, cohorts about 53%. Online completion rates traced to its source, with the denominator named."
lang: en
---

← All posts (https://wayflo.com/blog)

# Online Course Completion Rates: Free vs Paid vs Cohort (Real Data)

Free MOOCs finish 3% to 13%, paid courses around 46%, cohorts about 53%. Every completion rate traced to its source, with the denominator named.

Image: Justin Ferriman (https://wayflo.com/media/posts/8be32609-4b42-40ea-ab1b-ad23561b7ab8.jpg)

Justin Ferriman (https://wayflo.com/author/justin-ferriman) · November 12, 2026 · 12 min read

Image: online-course-completion-rates-wayflo (https://wayflo.com/media/posts/6327685a-f9c6-4f24-8db3-9400a2de0e09.png)

Looking for benchmark course completion rates before you plan your course strategy? Here's a secret: the average online course completion rate depends on where you count it.

- Free MOOCs finish roughly 3% to 13% of the people who sign up.
- Learners who pay finish around 46%.
- Scheduled cohort courses on one course platform finished 53%.

I see these numbers mixed up all the time, especially when nobody says where they were measured.

I've often seen a "5% completion rate" and a "33% completion rate" quoted for the same set of courses. That made me curious, and I noticed that different "course completion rate" statistics posts used different baselines.

So I decided to write a definitive guide to course completion statistics (and why they differ). Let me start with the completion rates I found.

## **TL;DR: Online course completion rates across the internet**

Here's a summary of the studies I referred to and how each one counted the completion rate (which denominator it used).

Image: benchmarks-course-completion-statistics (https://wayflo.com/media/posts/db4da9ea-0834-431e-9afe-a7610321df21.png)

Read the "Counted out of" column before you compare any two rows. Most marketing pages hide this, and that's how people end up quoting the wrong statistic.

| Course type | Completion rate | Counted out of | Study/Research |
| --- | --- | --- | --- |
| Free MOOCs (median course) | 12.6% | Everyone who enrolled | Jordan, 2015: https://www.irrodl.org/index.php/irrodl/article/view/2112 |
| Free MOOCs, HarvardX and MITx (2017-18) | 3.13% | People who opened the course | Reich & Ruipérez-Valiente, 2019: https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf |
| Free MOOCs, learners who planned to finish | 15.4% | People who said they intended to finish | Reich & Ruipérez-Valiente, 2019: https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf |
| Paid verified-track learners | 46% | People who paid | Reich & Ruipérez-Valiente, 2019: https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf |
| Paid self-paced courses (open access) | 41.9% | Paid enrollments on one platform | Ruzuku, 2026: https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study |
| Paid scheduled cohort courses | 53.4% | Paid enrollments on one platform | Ruzuku, 2026: https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study |

## **Why does every completion rate look different?**

The numbers above vary widely because "completion rate" means different things to different people (and to different coaches and platforms).

A completion rate is a fraction, so when the bottom number (the denominator) changes, the percentage changes too.

- Some studies divide completions by everyone who clicked "enroll."
- Others divide by people who opened the course.
- Some count only people who paid.

Image: same-course-four-rates (https://wayflo.com/media/posts/6a0c0f30-4c22-4fd1-a137-0d0f3cf65229.png)

The clearest example comes from 44 Coursera courses studied by Evans, Baker and Dee (https://scispace.com/pdf/persistence-patterns-in-massive-open-online-courses-moocs-1l24o4iqn3.pdf). The completion rate swings from 5% to 33% depending only on who you count.

- About 5% of everyone who registered earned a certificate. (Evans, Baker & Dee, 2016 (https://scispace.com/pdf/persistence-patterns-in-massive-open-online-courses-moocs-1l24o4iqn3.pdf); 44 MOOCs, 2.1M student observations)
- Counting only people who watched at least one video, the rate rose to 8.5%. (Evans, Baker & Dee, 2016 (https://scispace.com/pdf/persistence-patterns-in-massive-open-online-courses-moocs-1l24o4iqn3.pdf))
- Counting only people who watched more than a fifth of the videos, it rose to 20%. (Evans, Baker & Dee, 2016 (https://scispace.com/pdf/persistence-patterns-in-massive-open-online-courses-moocs-1l24o4iqn3.pdf))
- Among people who watched more than four fifths of the videos, about one in three earned the certificate. (Evans, Baker & Dee, 2016 (https://scispace.com/pdf/persistence-patterns-in-massive-open-online-courses-moocs-1l24o4iqn3.pdf))
- At HarvardX and MITx, 52% of people who registered never opened the course at all. (Reich & Ruipérez-Valiente, 2019 (https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf); 565 course runs, 12.67M registrations)

**Takeaway from this analysis**

If half your sign-ups never open the course, "completed out of enrolled" mostly measures your sign-up page.

The authors of that study called the all-registrants rate misleading themselves. So always track both numbers, but base your decisions on "completed out of started."

## **Free MOOC completion rates: 3% to 13%**

The famous "only 5% finish online courses" line comes from MOOC completion rate studies.

It's a real number, but it describes only free, open courses, where signing up costs nothing and takes one click.

Many people register just to look around. The range depends on the dataset.

- The median MOOC completion rate was 12.6% of enrollees, ranging from 0.7% to 52.1%. (Jordan, 2015 (https://www.irrodl.org/index.php/irrodl/article/view/2112); 221 courses, 129 with completion data)
- The median of 12.6% (https://www.irrodl.org/index.php/irrodl/article/view/2112) counts everyone who enrolled across hundreds of early MOOCs.
- Reich and Ruipérez-Valiente's (https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf) 3.13% counts only people who opened the course. It's lower, because it covers later years when completion kept falling.
- Longer courses and courses using peer grading had lower completion rates. (Jordan, 2015 (https://www.irrodl.org/index.php/irrodl/article/view/2112); 121 courses)
- Completion among HarvardX and MITx learners who opened their course fell from 4.96% in 2013-14 to 3.13% in 2017-18. (Reich & Ruipérez-Valiente, 2019 (https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf); 12.67M registrations)
- In the first year of HarvardX and MITx, 43,196 of 841,687 registrations earned a certificate, about 5%. (Ho et al., 2014 (https://papers.ssrn.com/abstract=2381263); 17 courses)
- Even learners who said they intended to finish only completed at 15.4%. (Reich & Ruipérez-Valiente, 2019 (https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf); 42,005 learners; High)

**Takeaway from this analysis**

Intent isn't enough. Five out of six people who planned to finish a free course never did. That's the gap your course design has to close.

Image: week-one-dropout (https://wayflo.com/media/posts/5f22588c-336f-4fb5-ae95-38eda14fe6f4.png)

## **Paid course completion rates: around 46%**

Putting free MOOCs aside, let's look at what almost everyone cares about: how many people who pay for a course actually finish it.

The most credible information comes from about five years of HarvardX and MITx data: learners on the paid verified track finished at 44% to 56%.

- Learners who paid for the verified certificate track completed at 46% in 2017-18. (Reich & Ruipérez-Valiente, 2019 (https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf); 46,326 learners)
- Across 12 courses, paid ID-verified learners earned certificates at 59%, against 5% for everyone else. (Ho et al., 2015, via MIT News (https://news.mit.edu/2015/mit-harvard-study-moocs-0401))
- In a $3,900 MIT xPRO program, employees given free access completed at 50%, against 77% for paying learners. But in edX courses for teachers, verified learners with a free coupon completed more often than those who paid (51% vs 38%). (Littenberg-Tobias et al., 2019 (https://webs.um.es/jruiperez/papers/conferences/2019_L@S_FreeCertificates.pdf))

**What I learned from this research (and an insight for you)**

If you sell your course, don't benchmark yourself against the 5% MOOC figure. Your real comparison is the 40% to 50% range.

## **Latest data on cohort course completion rates: around 53%**

The most recent and largest public dataset on cohort courses comes from Ruzuku (https://www.ruzuku.com/learn/articles/course-completion-rates), a course platform.

Image: cohort-vs-self-paced (https://wayflo.com/media/posts/c91f557c-3a44-4621-991e-6aafba37728f.png)

Its co-founder and CEO, Abe Crystal, analyzed about 1.3 million enrollments in paid courses.

- Scheduled cohort courses recorded 53.4% completion, against 42.9% for drip-released courses and 41.9% for open-access self-paced ones. (Ruzuku, 2026 (https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study); 1,253,716 enrollments across the three formats)
- Cohorts of 11 to 25 students averaged 65.2%, higher than cohorts of 1 to 10 (51.5%) or 100+ (51.3%). (Ruzuku, 2026 (https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study); 9,559 courses)
- In some subjects, including therapy and yoga, self-paced courses finished more often than scheduled ones. (Ruzuku, 2026 (https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study))
- A separate investment note claimed cohort courses reach over 75% completion, against 3% to 7% for MOOCs, with no data cited. (Andreessen Horowitz, 2021 (https://a16z.com/announcement/investing-in-maven/))

**Note:** The cohort-size figures are averaged per course, so they don't compare directly with the 53.4% format figure. Ruzuku says so in its own methodology. If you quote "65%," quote it as "cohorts of 11 to 25 on one platform."

## **Video completion rates are different from course completion rates**

Course completion and video completion get mixed up a lot.

Video completion is how much of a single video people watch. In edX lecture videos, more than half of viewing sessions ended before the video did.

The biggest edX study found that short videos held attention best. Yet some studies I read noted that students were about as likely to watch a 20-minute lecture as a 5-minute one.

Motivation is the driver here. Learners taking courses for credit watch roughly twice as long.

- 55.2% of viewing sessions on edX lecture videos ended before the video finished. (Kim et al., 2014 (https://cs.rochester.edu/hci/pubs/pdfs/edX-MOOC-in-video-dropouts-peaks_LAS-2014.pdf); 862 videos, 127,839 students)
- Predicted dropout rose from about 53% for a 5-minute video to about 71% for a 20-minute one. (Kim et al., 2014 (https://cs.rochester.edu/hci/pubs/pdfs/edX-MOOC-in-video-dropouts-peaks_LAS-2014.pdf))
- Median viewing time topped out at about 6 minutes, however long the video was. (Guo, Kim & Rubin, 2014 (https://pg.ucsd.edu/publications/edX-MOOC-video-production-and-engagement_LAS-2014.pdf); 6.9M sessions)
- For videos of 3 minutes or less, viewers watched more than three quarters of the video in 75% of sessions. (Guo, Kim & Rubin, 2014 (https://pg.ucsd.edu/publications/edX-MOOC-video-production-and-engagement_LAS-2014.pdf))
- The lead author of the 44-course study reported that students were just as likely to watch a 5-minute video as a 20-minute one. (Vanderbilt News on Evans, Baker & Dee (https://news.vanderbilt.edu/2016/02/26/what-makes-students-stick-with-a-mooc))
- In for-credit Stanford courses, median sessions on long lectures ran 12 to 13 minutes, double the MOOC figure. (Lagerstrom, Johanes & Ponsukcharoen, 2015 (https://peer.asee.org/24895); 30,141 sessions)

**Tip for your course planning**

Don't cut a good 20-minute lesson to hit a number. Break it into segments with a question or task between them.

Those breaks are where learning happens, as I covered in what the research says about interactive video (https://wayflo.com/blog/interactive-video-learning-research).

## **How to increase course completion rates in 2026**

No single study has yet proved what makes people finish a course.

What I can do is line up the patterns that hold across several datasets, and be clear about which ones are correlation only.

Image: increase-course-completion-rates (https://wayflo.com/media/posts/cc103ae8-461c-412f-9310-a42e7847832f.png)

Here's what I'd change, in the order I'd change it.

1. **Design week one as its own product.** Make the first lesson short enough to finish on the day someone enrolls, and end it with a small win. Dropout concentrates in week one. (Evans, Baker & Dee, 2016 (https://cepa.stanford.edu/node/1602); Ruzuku, 2026 (https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study))
2. **Put the most important video first in each module.** First-of-week and intro lectures drew more viewers, and lecture titles affected viewing too. Name videos for what learners will be able to do. (Evans, Baker & Dee, 2016 (https://cepa.stanford.edu/node/1602))
3. **Add practice, not just content.** Courses with assessments recorded 59.1% completion, against 42.4% without. That's correlation, but the learning case for in-video questions is much stronger, as I showed in the interactive video research. (Ruzuku, 2026 (https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study); 26,632 courses)
4. **Break long videos into segments.** Shorter segments hold attention, and more learners attempted the exercise after shorter videos (56% vs 31%). (Guo, Kim & Rubin, 2014 (https://pg.ucsd.edu/publications/edX-MOOC-video-production-and-engagement_LAS-2014.pdf))
5. **Keep the whole course shorter.** Longer MOOCs had lower completion. Cut anything that doesn't move learners toward the outcome you promised. (Jordan, 2015 (https://www.irrodl.org/index.php/irrodl/article/view/2112))
6. **Add a schedule if your subject suits it.** Cohort courses finished about 11 points higher on one platform, but not in every subject. Test it on your own runs. (Ruzuku, 2026 (https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study))
7. **Ask a question before the course starts.** Completing a pre-course survey was the strongest predictor of finishing in the 44-course study. That's a signal, not a cause, so use it to spot who needs a nudge early. (Evans, Baker & Dee, 2016 (https://cepa.stanford.edu/node/1602))
8. **Treat nudges as support, not strategy.** In a trial with about 250,000 learners across 247 courses, well-known nudges such as goal-setting prompts had effects about a tenth the size seen in earlier small studies. (Kizilcec et al., 2020 (https://pmc.ncbi.nlm.nih.gov/articles/PMC7334459))

## **What's a good completion rate for your course?**

There's no universal target, because none of these studies set one.

But you can pick a fair reference point once you know what kind of course you run and which denominator you use.

Image: good-completion-rate-course (https://wayflo.com/media/posts/44a7ca2c-8ea3-4744-9359-436a7241a517.png)

These are reference points from the data above, not goals proven to be achievable for your course:

- **Free course or lead magnet, counted from sign-ups:** 5% to 13% is normal for open courses. (Jordan, 2015 (https://www.irrodl.org/index.php/irrodl/article/view/2112); Ho et al., 2014 (https://papers.ssrn.com/abstract=2381263))
- **Paid self-paced course, counted from buyers:** roughly 40% to 46%. (Reich & Ruipérez-Valiente, 2019 (https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf); Ruzuku, 2026 (https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study))
- **Paid cohort course, counted from buyers:** roughly 50% to 65% on one platform. (Ruzuku, 2026 (https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study))

## **Advice for your course planning and analytics**

When you analyze completion, always report all three numbers, not one:

- Completed out of enrolled
- Completed out of started
- The share active in week one

The third moves first, so it's the one to watch after every change you make.

One last thing: completion isn't always the goal. Plenty of learners join for one module and leave satisfied.

Measure completion against what your course promises, not against a benchmark built on someone else's learners.

## **FAQ**

### **What is the average online course completion rate?**

It depends on the course type and on who you count. Free MOOCs finish about 3% to 13% of learners. Paid courses finish around 40% to 46% of buyers. Scheduled cohort courses finished 53% on one platform, though that figure is vendor data.

### **Why are MOOC completion rates so low?**

Mostly because signing up is free and takes one click. About half of MOOC registrants never open the course, so a rate counted from sign-ups starts low. Even among people who planned to finish, only 15% did.

### **Do cohort-based courses have higher completion rates?**

On the one large dataset available, yes, by about 11 points: 53.4% against 41.9% for self-paced courses. That data comes from a course platform. Other public claims of 75% or more come with no published data that I can find.

### **Does paying for a course make people finish it?**

Paying learners finish far more often, around 46% against single digits. But people who pay were likely more committed to begin with, so payment alone doesn't explain the gap.

### **When do most students drop out of an online course?**

In the first week. Participation falls fastest early and flattens later. On one platform, students who finished a lesson in their first seven days completed at 78%, against 17% for those who didn't.

### **How do you increase course completion rates?**

Fix week one first, add practice and in-video questions, segment long videos, and keep the course shorter. Don't rely on reminders or nudges alone; at scale, their effects were small.

## Sources

1. Evans, B. J., Baker, R. B., & Dee, T. S. (2016). Persistence patterns in massive open online courses (MOOCs) (https://cepa.stanford.edu/node/1602). _The Journal of Higher Education_, 87(2), 206-242. Working paper version 2015. Online appendix (https://scispace.com/pdf/persistence-patterns-in-massive-open-online-courses-moocs-1l24o4iqn3.pdf).
2. Jordan, K. (2015). Massive open online course completion rates revisited: Assessment, length and attrition (https://www.irrodl.org/index.php/irrodl/article/view/2112). _IRRODL_, 16(3).
3. Reich, J., & Ruipérez-Valiente, J. A. (2019). The MOOC pivot (https://webs.um.es/jruiperez/papers/journals/2019_Science_MOOCPivot_postprint.pdf). _Science_, 363(6423), 130-131. Author postprint with supplementary tables.
4. Kizilcec, R. F., et al. (2020). Scaling up behavioral science interventions in online education (https://pmc.ncbi.nlm.nih.gov/articles/PMC7334459). _PNAS_, 117(26), 14900-14905.
5. Crystal, A. (2026). What 27,156 courses tell us about cohort vs. self-paced completion (https://www.ruzuku.com/learn/articles/cohort-vs-self-paced-completion-study). Ruzuku. Vendor data, not peer reviewed; updated November 2026.
6. Ho, A. D., et al. (2014). HarvardX and MITx: The first year of open online courses (https://papers.ssrn.com/abstract=2381263). HarvardX and MITx Working Paper No. 1.
7. MIT News (2015). Study on MOOCs provides new insights on an evolving space (https://news.mit.edu/2015/mit-harvard-study-moocs-0401). Reporting Ho et al. (2015).
8. Littenberg-Tobias, J., Ruipérez-Valiente, J. A., & Reich, J. (2019). Impact of free-certificate coupons on learner behavior in online courses: Results from two case studies (https://webs.um.es/jruiperez/papers/conferences/2019_L@S_FreeCertificates.pdf). _Learning at Scale 2019_.
9. Koller, D., et al. (2013). Retention and intention in massive open online courses (https://er.educause.edu/articles/2013/6/retention-and-intention-in-massive-open-online-courses-in-depth). _EDUCAUSE Review_.
10. Guo, P. J., Kim, J., & Rubin, R. (2014). How video production affects student engagement (https://pg.ucsd.edu/publications/edX-MOOC-video-production-and-engagement_LAS-2014.pdf). _Learning at Scale 2014_.
11. Kim, J., et al. (2014). Understanding in-video dropouts and interaction peaks in online lecture videos (https://cs.rochester.edu/hci/pubs/pdfs/edX-MOOC-in-video-dropouts-peaks_LAS-2014.pdf). _Learning at Scale 2014_.
12. Lagerstrom, L., Johanes, P., & Ponsukcharoen, U. (2015). The myth of the six-minute rule (https://peer.asee.org/24895). _ASEE Annual Conference_.
13. Vanderbilt News (2016). What makes students stick with a MOOC (https://news.vanderbilt.edu/2016/02/26/what-makes-students-stick-with-a-mooc).
14. Chen, A. (2021). Investing in Maven (https://a16z.com/announcement/investing-in-maven/). Andreessen Horowitz. Cited as an unsupported claim.

Justin Ferriman (https://wayflo.com/author/justin-ferriman)

Founder of Wayflo, creating the next era of online courses. Previously founded and sold LearnDash. Let's connect!

Justin Ferriman on X: https://x.com/justinferriman
Justin Ferriman on LinkedIn: https://www.linkedin.com/in/justinferriman/
Justin Ferriman's website: https://justinferriman.com/

## Structured data

```json
{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "BlogPosting",
      "@id": "https://wayflo.com/blog/online-course-completion-rates#article",
      "headline": "Online Course Completion Rates: Free vs Paid vs Cohort",
      "description": "Free MOOCs finish 3% to 13%, paid courses around 46%, cohorts about 53%. Online completion rates traced to its source, with the denominator named.",
      "mainEntityOfPage": "https://wayflo.com/blog/online-course-completion-rates",
      "image": "https://wayflo.com/media/posts/6327685a-f9c6-4f24-8db3-9400a2de0e09.png",
      "datePublished": "2026-11-12T12:11:24.135+00:00",
      "dateModified": "2026-11-12T12:11:24.135+00:00",
      "author": {
        "@type": "Person",
        "@id": "https://wayflo.com/author/justin-ferriman#person",
        "name": "Justin Ferriman",
        "url": "https://wayflo.com/author/justin-ferriman"
      },
      "publisher": {
        "@type": "Organization",
        "@id": "https://wayflo.com/#organization",
        "name": "Wayflo",
        "url": "https://wayflo.com/"
      }
    },
    {
      "@type": "BreadcrumbList",
      "itemListElement": [
        {
          "@type": "ListItem",
          "position": 1,
          "name": "Home",
          "item": "https://wayflo.com/"
        },
        {
          "@type": "ListItem",
          "position": 2,
          "name": "Blog",
          "item": "https://wayflo.com/blog"
        },
        {
          "@type": "ListItem",
          "position": 3,
          "name": "Online Course Completion Rates: Free vs Paid vs Cohort (Real Data)",
          "item": "https://wayflo.com/blog/online-course-completion-rates"
        }
      ]
    }
  ]
}
```