The Knewton Blog

Our monthly newsletter features edtech and product updates, with a healthy dose of fun Knerd news.

The Unbundling of Higher Education

Posted in CEO Jose Ferreira on February 26, 2014 by

Television used to be four networks: ABC, CBS, NBC, and PBS. With cable came more channels — devoted to foreign language programming, cooking, sports, movies, history, and so on. Then the silos became narrower. History programming fragmented from the History Channel into the Biography Channel, the Military History Channel, and the Smithsonian Channel. On-demand services and the DVR took it further still, allowing people to choose exactly which show to watch, at exactly what time.

This process is the “unbundling” of broadcast television, where “unit size” has been reduced from networks to niche networks to individual shows. The DVR doesn’t yet let us break the show into even smaller units — e.g., I can’t automatically record just the weather segment of the Today show, or anything that ever appears on any channel about the Pittsburgh Steelers — but this will likely happen too. Cable companies are fighting to keep unit size bundled at the level of their packages, but consumer choice will inevitably force them to unbundle.

Unbundling has played out in almost every media industry. We now purchase individual songs instead of albums. A few dozen major national magazines fragmented into thousands of specialty magazines serving every type of interest, and then fragmented again into millions of blogs.

Air travel was once a luxury that people dressed up for. Travelers’ buying decisions eventually forced carriers to unbundle their services as people sought the cheapest tickets. Airlines now charge for anything beyond transporting your body: entertainment, food, bags, extra legroom, early boarding. Sure, people gripe about paying $10 for an in-flight snackbox — but they speak louder with their purchasing preference for the least expensive flights. Airlines have been compelled to shrink the unit size of air travel to fit consumer demand.

Breakage and buffets

Bundling works to increase margin by tricking people into thinking that there is more value in a product or service than there actually is. There are two reasons bundling can do this:

  • The complexity of the bundle reduces the product’s transparency, impeding consumers’ ability to do cost-benefit analysis.1
  • The bundle includes a lot of stuff you could use but don’t, i.e., “breakage.”

When a product’s outcomes are transparent, complex product bundles don’t last long. They are inevitably broken down over time into smaller unit sizes, as consumers purchase only what they need. It is only when outcomes are non-transparent — as in, for instance, education (up until now) — that complex product bundles can survive and price can keep increasing.

Breakage is a free lunch for suppliers; they are charging for services they don’t deliver. A few businesses are predicated entirely on breakage — e.g., gift cards, which count on a significant number of people never redeeming them. But not all business models with high breakage are doing something dubious. There may simply be no obviously better pricing model. Traditional health clubs allow you to use the facility whenever it’s open, since charging by the hour would be extremely unwieldy — so much so that bundling creates value by simplifying logistics and saving time.

In some cases, breakage improves the product itself (not just its delivery). Consider that worldwide hotel staple: the all you can eat breakfast buffet. There are core items that diners consume a lot of — cereal, pancakes, eggs. Then there are fringe items that are less heavily consumed — smoked mackerel, blackberries, cottage cheese. If most people partake of a small quantity of fringe items — or if nearly everyone really wants a different one — then fringe items enrich the experience for everyone. All users receive relatively equal value.

For both health clubs and all you can eat buffets, the bundled product is a market equilibrium unit size. Many people buy breakfast à la carte, but enough buy the buffet that it is ubiquitous. Gym membership is an even more pervasive unit size within its industry. The only unbundling health clubs suffer is boutique fitness studios that offer yoga, spinning, etc.

But what about when some users consume far more than others? Or when a majority never uses any fringe items? Then value is created for all users but is unequally distributed. Everyone uses some of the bundle, but many users pay for things they don’t want and don’t use, increasing profit margin and/or subsidizing other users. This brings us to the university model, which is the perfect storm of bundling. It is both health club and all-you-can-eat buffet. And its product unit size is gargantuan, with extremely complex, non-transparent outcomes.

Will universities ever be unbundled?

Universities bundle services like mad: courses, academic research, credentials, housing, food, athletic programs, job placement, extracurriculars, the arts, study abroad, social life, etc. As my old Kaplan Inc. boss and mentor Andy Rosen described in his excellent book Change.edu, universities have for decades been competing by adding even more services. Originally, the university bundle included courses, food, and board. Over time they’ve added more services, at first academic (extracurriculars, better libraries) and now luxury (rock-climbing walls, European-style bistros).

Higher education is still trending towards increasing bundle size. The more they bundle, the more they can raise prices. So when, if ever, can we expect higher ed to start trending in the other direction, towards unbundling?

My guess is in about five years, plus or minus. The forces that will make it inevitable are picking up steam right now.

Let’s pretend that we wish to create a new startup non-profit: an unbundled, purely academic university. We’d offer high-quality courses and none of the other stuff. (These courses wouldn’t be MOOCs. MOOCs represent the unbundling of the course itself; they strip away the credits, and most of the high-quality materials and support services, to offer just lectures.) We’d pay professors $200,000 (including benefits, payroll taxes, etc.). At those rates we could attract some of the very best professors in the world. Let’s say each professor teaches 250 students/year and does nothing else (e.g., no research). Teaching two or three sections per semester would leave ample time for prep and office hours. Add in materials and tech fees around $100 and we could offer these courses for $900 per student per course, excluding marketing costs and considering only the cost of product delivery. These courses would be academically equivalent (incredible professor, great materials, office hours2) to any “regular” university course, but delivered online at around 20% of the all-in cost of buying such a course bundled with food, lodging, athletic facilities, Jacuzzis, and rock walls at an elite university.

Universities are creating courses like this as we speak. They aren’t, generally, as high quality as the one we just postulated, but they’re getting better all the time. Soon, data will make learning outcomes for these courses highly transparent. As it becomes easier to evaluate efficacy, and as credit acceptance policies become more flexible, these high-quality à la carte courses will become commonplace.

Once that happens, the unit size of higher ed will begin to change from a traditional four-year on-campus bundled experience to more of a course-based experience. Some students will take on-campus courses from their home institution with high-quality online courses from elsewhere. Others will choose online-only degrees. If the per-course fee of the bundled campus experience too greatly exceeds that of the à la carte experience, more and more people will choose the latter. Universities may eventually come to look like a bit like airlines, where a minority of customers pays very high prices for first class. Everyone else will average down the price of university by taking some or all of their courses online, driving institutions to offer more such courses and focus on these types of students.

Even the course itself is an arbitrary unit size. So is the semester. Using “time spent” as the measure of how much a student has learned is absurd. Personalized course materials made possible by real data — based on “normed” assessment items measuring student proficiencies on every concept — will soon allow students to prove mastery in whatever amount of time is best for them.

So unbundling is the process of breaking down a recently stable product unit size into component parts, forcing margin reduction and lower prices for consumers. Will TV’s fundamental unit size be the show, or something even smaller (e.g., any segment that discusses the weather or the Steelers)? Will the basic unit of education be the semester-long course, or the modules that make up that course? We’ll have to see. My guess is semesters and courses aren’t going away. But they are going to become a lot more fluid, transferable, and transparent. They will be purchasable à la carte, and price per unit will have to come down. There will always be a market for the elite bundled college experience. But new unbundled or partially bundled alternatives will emerge, and eventually they will dominate the industry.


  1. University Ventures wrote an excellent blog post on the unbundling of education and the economics therein that is well worth a read. 

  2.  Yes, these classes and office hours would be online. But, at this point in the evolution of the Internet, it is difficult to argue that people can’t form strong social and emotional connections online.