Skip to content

The Failure of Udacity

Well, there it is folks. After two years of hype, breathless proclamations about how Udacity will transform higher education, Silicon Valley blindness to existing learning research, and numerous articles/interviews featuring Sebastian Thrun, Udacity has failed.

No one did more of a disservice to MOOCs than Thrun through his wild proclamations (“we have found the magic combination for online learning” and “in the future there will only be 10 universities”, digital learning manifestos, and so on) and self-aggrandizing. No one will do more damage to MOOCs as a concept than Thrun now that he realizes how unfounded his statements actually were.

Amazingly, after Udacity and Thurn’s “bull in a China shop” run through higher education, he proclaims that he has seen the light: “”We were on the front pages of newspapers and magazines, and at the same time, I was realizing, we don’t educate people as others wished, or as I wished. We have a lousy product…It was a painful moment.”"

The Udacity pivot, showcased (a latin term meaning “spin”) as a good thing in the Fast Company article, is the equivalent of Obama doing an Affordable Care is Working media tour. Make no mistake – this is a failure of Udacity and Sebastian Thrun. This is not a failure of open education, learning at scale, online learning, or MOOCs. Thrun tied his fate too early to VC funding. As a result, Udacity is now driven by revenue pursuits, not innovation. He promised us a bright future of open learning. He delivered to us something along the lines of a 1990′s corporate elearning program.


  1. dkernohan wrote:

    I wonder – did he actually turn up for the Fast Company interview, or did he skype it in?

    Friday, November 15, 2013 at 11:15 am | Permalink
  2. Ilya Kipnis wrote:

    Honestly, I think the reason that MOOCs have such a high non-completion rate is that there is no cost to signing up, and basically no reward for completing the course beyond a little certificate that no employer cares a hoot about.

    The greatest reward that a MOOC (or series of them) can possibly have is to have a pot of gold at the end of the rainbow, because most people certainly don’t have those coveted six-figure jobs.

    Friday, November 15, 2013 at 12:11 pm | Permalink
  3. Rolin Moe wrote:

    I thought it was interesting how heavily he tied everything into completion rates, especially after your keynote noting the folly of hyperfocus on completion rates. Obviously the problem was not a group of Intro to Statistics students failing to log in despite Mr. Thrun’s smiles.

    Friday, November 15, 2013 at 12:19 pm | Permalink
  4. Sandi Boga wrote:

    This is a very interesting, yet not altogether surprising turn of events. Coursera has also been talking about developing offerings for the corporate world. After all, in the pursuit of revenue, it makes sense to partner with industry.

    I can’t help but wonder if Coursera’s pilot project in Africa (New Economy Skills for Africa Program-Information and Communication Technologies (NESAP-ICT), offered in conjunction with the World Bank, will be the catalyst for a similar shift. Two of the objectives of that program are to provide:

    •Globally-benchmarked, industry-rated skills assessment, training and certification
    •“Bridge” style training programs that align what students are taught and what industry requires

    Sounds like a great opportunity for a large US-based tech company to fund Coursera MOOCs that will train a pool of low-cost African staff for future outsourcing initiatives.

    But is that so bad, if students are getting an IT education out of the deal?



    Trucano, M. (April 4, 2013). MOOCs in Africa. EduTech: A World Bank Blog on ICT use in Education. Retrieved from

    The World Bank (2013). New Economy Skills for Africa Program-Information and Communication Technologies (NESAP-ICT). Retrieved from,,contentMDK:22335863~menuPK:617610~pagePK:148956~piPK:216618~theSitePK:282386,00.html

    Friday, November 15, 2013 at 5:05 pm | Permalink
  5. ashok baktha wrote:

    Failure? I think that is a bit too harsh. Rather, I would tend to agree with Dave Llorens’s perspective: Udacity: Shifting Models Means Never Having to Say You’re Sorry | All MOOCs, All The Time

    Saturday, November 16, 2013 at 2:19 am | Permalink
  6. It’s time to move away from commercial manipulation of the MOOC and support grass roots innovation and collaboration!

    Saturday, November 16, 2013 at 12:23 pm | Permalink
  7. Indeed a disservice.
    In a wider horizon, it means that though “massive” while “open” ist incompatible with the “revenue” model. Open education, when bounded by a (traditional) hierarchical structure that has business targets, is bound to fail. We need to learn from it, in order to avoid damaging openness.

    Saturday, November 16, 2013 at 6:49 pm | Permalink
  8. Keith Brennan wrote:

    Thanks, Ashok, for the link. It’s a thoughtful piece on the meaning of the Udacity direction change, and Open Online Learning generally.

    I’d add that lots of hype cycles, theories, pedagogies, interventions and disruptions are characterised by exactly the same hubris, disregard for evidence, and foundationally flawed methods as the ones Thrun based his revolution on.

    And lots of those theories have adherents who have drunk the Kool aid in exactly the way.

    Thrun’s failure may not be the failure of any other sector, section of cheerleading section other than his own, but other technology in education revolutionaries can learn from his mistakes.

    The key difference between Thrun and others may not be in how wrong he was. It may not be in how influentially wrong he came to be. It may be in the fact that he has publically held his hands up and said it, where other ideologues, theorists, and disruptors won’t.

    The other major issue with Thrun’s turnaround (apart from damaging educational practices that work better than the disruptions he abandoned, in the eyes of the public and policymakers, damaging more careful and committed online educators, damaging the people who failed to benefit from the educational experiences he put them through). Choosing to drop his focus from the consituency that he so emotively evoked in his Ted talk. The underserved, alienated from education, hungry for opportunity mass of people with no access.

    Sunday, November 17, 2013 at 12:33 pm | Permalink
  9. Curious, what does a “successful” or “failed” MOOC look like?

    Sunday, November 17, 2013 at 1:16 pm | Permalink
  10. Esther Jackson wrote:

    Maybe with some refinement, other approaches could be applied to Udacity for MOOC to have future success as we know technology is continuously changing. Online learning is in a constant mode of change which opens up the world of education with numerous possibilities for learning mechanisms to be implemented.

    Monday, November 18, 2013 at 3:16 am | Permalink
  11. Keith Brennan wrote:

    Hi Benjamin.

    I’d guess in Udacity’s context, in San Jose, it looks like a course which significantly underperformed in comparison to a face to face course of exactly the same type, with exactly the same purpose, to pretty much the same students (when you lift the high school students out of the data).

    Given that they had learning outcomes, and previous test scores to measure against, failure wasn’t that difficult to quantify.

    Additionally, the fact that the students who most needed the course, and who this type of course was squarely aimed at did hugely worse than they did in face to face versions, and the picture is pretty clear.

    In other contexts, lack of participation from students who wanted to participate, failure to achieve personal or desired course learning goals, lack of clarity about course resources, structure, and supports that leads to iability to navigate/find them, unstated and significant barriers to entry, courses where the technology breaks, or is wholly innapropriate, courses whose design/resources and stated aims are clearly not aligned.

    In some MOOC tyoes it’s more difficult to quantify, in some less, but it’s not impossible.

    Monday, November 18, 2013 at 5:23 am | Permalink
  12. It was very clear from the beginning that at some point of time MOOC providers have to think of an economical model so as to sustain the huge cost involved. I believe edX has been doing it already ( )which is such a sensible decision.

    Tuesday, November 19, 2013 at 2:52 am | Permalink
  13. Mythix wrote:

    So why do we have an education system if not to prepare students for what comes next in their lives? It would be wonderful if one did not have to have a job, did not need to be prepared for anything. I grant you that following ones passion is a wonderful thing if you know what that is. What if you have no clue, how do you prepare students for the rest of their lives?

    Tuesday, November 19, 2013 at 9:40 am | Permalink
  14. Tracy Marks wrote:

    Having participated actively in eight Moocs in the past year (five at Coursera, three at Canvas) and earned full completion credentials at four, my experience is that the “most successful” Moocs not only have dynamic, knowledgeable professors, they also have very active discussion forums with students creating dozens of topics related to the course and getting very involved in dialogue. The instructor’s participation in the discussion forum also encourages greater involvement in the course.

    This does not happen in a join-anytime course – I’ve noticed that some of the Udacity courses indicate that only a dozen or so people have taken them in a year, probably at different times, which would mean that one is basically doing the work entirely on one’s own and if lucky, with occasional communication with an instructor.

    Some of the Moocs I’ve completed with active discussion (Modern Poetry, The Fiction of Relationship) resulted in a lot of connection between students – in both cases, with a large group of students continuing the study on their own for months after the course ended without the instructor. This is not going to happen in a join-anytime course which is mostly self-study.

    Maybe only 10% of the students will be actively involved in a Coursera-style discussion forum, but those students are likely to be both emotionally and intellectually engaged and certainly more likely to complete the course.

    Also, when assignments have specific due dates, and in many cases involve or benefit from interaction or sharing with other group members (e.g. “post your essay here – feedback welcome”), students are more likely to participate.

    Tuesday, November 19, 2013 at 11:31 pm | Permalink
  15. I enjoyed your response to the Thrun profile. Please read mine:

    Thursday, November 21, 2013 at 8:49 pm | Permalink