OER and the Future of Knewton

Jose Ferriera, the CEO of Knewton, recently published a piece on edSurge arguing that scaling OER cannot “break the textbook industry” because, according to him, it has low production values, no instructional design, and is not enterprise grade. Unsurprisingly, David Wiley disagrees. I also disagree, but for somewhat different reasons than David’s.

When talking about Open Educational Resources or, for that matter, open source software, it is important to distinguish between license and sustainability model, as well as distinguishing between current sustainability models and possible sustainability models. It all starts with a license. Specifically, it starts with a copyright license. Whether we are talking about Creative Commons or GPL, an open license grants copyright permission to anyone who wants it, provided that the people who want to reuse the content are willing to abide by the terms of the license. By granting blanket permission, the copyright owner of the resource chooses to give up certain (theoretical) revenue earning potential. If the resource is available for free, then why would you pay for it?

This raises a question for any resource that needs to be maintained and improved over time about how it will be supported. In the early days of open source, projects were typically supported through individual volunteers or small collections of volunteers, which limited the kinds and size of open source software projects that could be created. This is also largely the state of OER today. Much of it is built by volunteers. Sometimes it is grant funded, but there typically is not grant money to maintain and update it. Under these circumstances, if the project is of the type that can be adequately well maintained through committed volunteer efforts, then it can survive and potentially thrive. If not, then it will languish and potentially die.

But open resources don’t have to be supported through volunteerism. It is possible to build revenue models that can pay for their upkeep. For example, it is possible to charge for uses of materials other than those permitted by the open license. Khan Academy releases their videos under a Creative Commons Noncommercial Share-Alike (CC NC-SA) license. Everyday students and teachers can use it for free under normal classroom circumstances. But if a textbook publisher wants to bundle that content with copyrighted material and sell it for a fee, the license does not give them permission to do so. Khan Academy can (and, as far as I know, does) charge for commercial reuse of the content.

Another possibility is to sell services related to the content. In open source software, this is typically in the form of support and maintenance services. For education content, it might be access to testing or analytics software, or curriculum planning and implementation services. This is a non-exhaustive list. The point is that it is possible to generate revenue from open content. And revenue can pay for resources to support high production values, instructional design, and enterprise scaling, particularly when paired with grant funding and volunteer efforts. These other options don’t necessarily generate as much revenue as traditional copyright-based licensing, but that’s often a moot point. Business models based on open licenses generally get traction when the market for licensed product is beginning to commodify, meaning that companies are beginning to lose their ability to charge high prices for their copyrighted materials anyway.

That’s the revenue side. It’s also important to consider the cost side. On the one hand, the degree to which educational content needs high production values and “enterprise scaling” is arguable. Going back to Khan Academy for a moment, Sal Khan popularized the understanding that one need not have an expensive three-camera professional studio production to create educational videos that have reach and impact. That’s just one of the better known of many examples of OER that is considered high-quality even though it doesn’t have what publishing professionals traditionally have thought of as “high production values.” On the other hand, it is important to recognize that a big portion of textbook revenues go into sales and marketing, and for good reason. Despite multiple efforts by multiple parties to create portals through which faculty and students can find good educational resources, the adoption process in higher education remains badly broken. So far with a few exceptions, the only good way to get widespread adoption of curricular materials still seems to be to hire an army of sales reps to go knock on faculty doors. It is unclear when or how this will change.

This brings us to the hard truth of why the question of whether OER can “win” is harder than it seems. Neither the OER advocates nor the textbook publishers have a working economic model right now. The textbook publishers were very successful for many years but have grown unsustainable cost structures which they can no longer prop up through appeals to high production values and enterprise support. But the OER advocates have not yet cracked the sales and marketing nut or proven out revenue models that enable them to do what is necessary to drive adoption at scale. If everybody is losing, then nobody is winning. At least at the moment.

This is where Knewton enters the picture. As you read Jose’s perspective, it is important to keep in mind that his company has a dog in this fight. (To be fair at the risk of stating the obvious, so does David’s.) While Knewton is making noises about releasing a product that will enable end users to create adaptive content with any materials (including, presumably, OER), their current revenues come from textbook publishers and other educational content companies. Further, adaptive capabilities such as the ones Knewton offers add to the cost of an educational content product, both directly through the fees that the company charges and indirectly through the additional effort required to design, produce, and maintain adaptive products. To me, the most compelling argument David makes in favor of OER “winning” is that it is much easier to lower the price of educational materials than it is to increase their efficacy. So if you’re measuring the value of the product by standard deviations per dollar, then smart thing is to aim for the denominator (while hopefully not totally ignoring the numerator). The weak link in this argument is that it works best in a relatively rational and low-friction market that limits the need for non-product-development-related expenses such as sales and marketing. In other words, it works best in the antithesis of the conditions that exist today. Knewton, on the other hand, needs there to be enough revenue for curricular materials to pay for the direct and indirect costs of their platform. This is not necessarily a bad thing for education if Knewton-enhanced products can actually raise the numerator as much as or more than OER advocates can lower the denominator. But their perspective—both in terms of how they think about the question of value in curricular materials and in terms of how they need to build a business capable of paying back $105 million in venture capital investment—tilts toward higher costs that one hopes would result in commensurately higher value.

All of this analysis assumes that in David’s ratio of standard deviations per dollar, all that matters is the ratio itself, independently of the individual numbers that make it up. But that cannot be uniformly true. Some students cannot afford educational resources above a certain price no matter how effective they are. (I would love to lower my carbon footprint by buying a Tesla. Alas….) In other cases, getting the most effective educational resources possible is most important and the extra money is not a big issue. This comes down to not only how much the students themselves can afford to pay but also how education is funded and subsidized in general. So there are complex issues in play here regarding “value.” But on the first-order question of whether OER can “break the textbook industry,” my answer is, “it depends.”