Online Learning: Disruption or Niche Product?

Michael Horn recently delivered a webinar through NBOA, titled, “Disrupting Class: Five Years Later.” In the book Disrupting Class, Horn and Clay Christensen applied Christensen’s theory of disruptive innovation to education. They concluded that online learning possessed the qualities of innovations from other industries that had started small, found a niche, developed further, and then displaced traditional practice. They boldly forecasted that 50% of all high school courses in the U.S. would take place online by 2019.

Going into this webinar, one might have expected Horn to backtrack or modify this prediction. We are now halfway to 2019, and online learning still appears to be a niche activity. In 2009, 1.0% of grades 9-12 enrollments took place online [Horn and Christensen]. In 2013, an estimated 5% of high school students took at least one online course (“Keeping Pace with K-12 Online and Blended Learning”), as compared with 13% of postsecondary students (“Changing Course:
Ten Years of Tracking Online Education in the United States“).

In the webinar, Horn defended the 2009 prediction, stating that the U.S. was still on track to reach 50% of course enrollments online by 2019. He provided one important qualification, that the prediction now refers to online and blended learning. In this way, Horn has significantly shifted tracks from the 2009 argument. Horn also did not substantiate the claim with recent numbers, instead suggesting that hard numbers were hard to find.

The argument for disruptive innovation rests on several specific assumptions. One is that disruptive innovation follows an exponential S-curve, with slow growth at the start of adoption, then very rapid growth during majority adoption, and then slowing growth with the last adopters. Is online learning still growing exponentially, even in its early days? Even if so, early exponential growth does not by itself lead to full adoption, or else all schools might be Waldorf. Early enthusiasm for an idea can evaporate later.

The theory also requires several conditions, as identified in Disrupting Class. The innovation is much simpler and basic than the current products and services that it might displace. It meets the needs of nonconsumers, customers who would like to access education but cannot. The company or product is significantly separated from the main provider in that space, so that it can develop without being co-opted. The product or service meets an underlying consumer need better than the existing products or services.

Horn believes that online education, or more specifically blended learning, still meets these conditions. Online education continues to stand as an alternate model, in contrast to face-to-face education. Blended learning is currently less sophisticated than established, traditional schools. It is, according Horn’s examples, lower cost, due to facilities and staff savings. Horn also referred to his 2013 study of different forms of blended learning, demonstrating that a number of different models exist, though most still represent alternative forms of school.

Horn spent a considerable portion of the presentation describing the advantages of blended learning, such as personalization, individualized learning pathways, and student ownership of the learning process. In this way, blended learning might meet core student needs better than traditional schooling. Horn also believes that considerable nonconsumer populations have adopted online learning, which has helped the innovation get its start. According to disruptive innovation theory, this may provide the conditions needed for online learning to mature and compete with traditional schools.

Disruptive innovation theory has come under substantial fire for being neither scientific nor accurate, for example in a widely shared June 2014 New Yorker article. In the article, Jill Lepore excoriated Christensen for handpicking case studies, making circular arguments, and feeding off panic.

Disruptive innovation is a theory about why businesses fail. It’s not more than that. It doesn’t explain change. It’s not a law of nature. It’s an artifact of history, an idea, forged in time; it’s the manufacture of a moment of upsetting and edgy uncertainty. Transfixed by change, it’s blind to continuity. It makes a very poor prophet.

In the webinar, Horn summarily dismissed these criticisms, and not even by name. He simply remarked that “a lot of people don’t understand” disruption theory and then proceeded into his prepared talk.

Private school leaders may have read Disrupting Class and rushed to plan their own online schools. Horn cited Global Online Academy and Online School for Girls, and I would add the Bay Area BlendEd Consortium to the list of prime examples. However, one might be surprised to hear that Horn does not believe that disruptive innovation theory applies to elite private schools, nor does he think that online and blended learning will displace them. Private schools are defined by a selective admissions process and a high-end product, which is likely to stay ahead of online education in perceived quality. In addition, elite private schools have the resources to co-opt online learning if they choose to do so, and as perhaps GOA, OSG, and BlendEd already have.

Horn’s new book, Blended: Using Disruptive Innovation to Improve Schools is due for publication in November. Presented as a practical guide, the book suggests that “blended learning is one of the hottest trends in education right now, and educators are clamoring for ‘how-to’ guidance.” Given this description, I am not confident that Horn will further explore the theoretical and empirical basis for his arguments. However, as secondary schools have relied more on practical experiences and community sentiment (as expressed by enrollment) to craft program, perhaps this book will further inform private school leaders’ strategies for online and blended learning. If you cannot wait until November, you may want to read two of Horn’s prior articles on the topic, “The Rise of K-12 Blended Learning” and “Is K–12 blended learning disruptive? An introduction to the theory of hybrids.”



  1. Thanks for your extensive coverage of the webinar for NBOA as well as your pointed questions! In case it’s useful to you or your readers, thought I’d give a bit more context here–and then I’m sure we can chat more at the NBOA event in February. I won’t address all your points here I’m sure.

    1) On the prediction of the growth of online learning/the S-curve — We had always assumed that the majority of online-learning enrollments would occur in blended-learning environments so this is not a change. A few helpful points. First, in Disrupting Class, as specified in Ch. 4, the data on which the S-curve was based counted both blended and virtual enrollments. That said, we did not have a full-depth of understanding of the blended-learning phenomenon so we did not use that terminology as much as now do. Second, as evidence of this, all of the vignettes in the book occur in schools–students using online learning in brick-and-mortar schools with teachers. Third, an important distinction with Keeping Pace is that when it uses the word online learning, it is generally only referring to virtual learning instances of online learning, whereas from our perspective, online learning can occur in both virtual and blended settings. Students learn online all the time in both. As a result though, the figures that Keeping Pace uses are only calculating instances of virtual online learning–at a distance in other words. Fourth, before Disrupting Class was published, we also did an S-curve calculation on the growth of home-schooling/full-time virtual schooling, and an S-curve–or substitution curve–did *not* materialize. The growth flattened out at around 10% of the K-12 schooling population, which makes some sense given that at least 90% of students likely need the custodial care or friendship opportunities, etc. that schools also importantly provide. So we’ve always both instinctively and empirically felt that the bulk of online learning enrollments would occur in blended-learning settings.

    2) I can tell you that a very conservative calculation on the numbers of students doing at least some online learning (blended or virtually) as of the previous school year was 7 million. This year I suspect it’ll come out closer to 9 million, but we’ll see. I obtained these rough figures just from conversations with some of the larger vendors in the space, and it doesn’t count any of the informal instances of flipped classrooms, etc. taking place, so the number likely significantly understates the activity going on. Doesn’t seem so niche.

    3) Lastly, on the Lepore point–I didn’t address it because no one asked nor was I in fact addressing her article indirectly in my comments. Her article was marked with lots of inaccuracies and misunderstandings (including the quote used here–disruption is not a theory about the failure of business, for example), but the beginning of her article I think was quite accurate in pointing out that the words disruption, disruptive, and disruptive innovation/technology are way way overused and misapplied. As a result, when people do that because they misunderstand what the theory does and does not say, they strip down its meaning significantly. So to the extent I said that, I was actually agreeing with the beginning of her piece. There have been retorts to her article online, but I’ll just link to one that I think provides the best empirical evidence to counter her arguments, which I hope is of use:

    Thanks again!

  2. Richard says:


    Many thanks for your extensive reply. I appreciate and am impressed that you would take the time to write this thoughtful comment in this space. Unfortunately, I will not be present at NBOA, as I am the academic dean at our school and primarily responsible for school curriculum and faculty professional development.

    When you estimate the growth of blended learning, where do you draw the line on what counts? Such a broad continuum exists from Carpe Diem to traditional schools with a LMS requirement for each class. Did the vendors with whom you spoke report aggregate sites or only those that met some measure of student online activity? Is a class with significant online activity truly blended if it meets in person four times a week? If the context of the discussion is the disruption of education by online learning, I would personally expect that only a tiny fraction of schools using LMS’s to count as blended.

    U Prep is one such school, very recognizable in structure but also an active experimenter with contemporary learning methods and willing to question fundamental structures in the school. We do have a LMS requirement for all of our classes, and a culture of online participation has developed in these spaces. In addition, two teachers are experimenting with “flexible Fridays,” in which students work independently, class attendance is optional, and the teacher is available to provide feedback or assistance if desired. Even though learning is taking place online within these classes, I would be hard pressed to call this blended learning.

    I am getting at the distinction between evolution and revolution. If online learning is truly disruptive, as opposed to sustaining, then I would expect to see fundamental changes in how students learn. To this point, I see this only very occasionally. I look forward to experiencing the revolution if/when it happens!

    With appreciation,


  3. Richard says:

    John Chubb takes a look at online and blended learning in today’s NAIS bulletin.