By Barbara Correa
The prospect of online courses opening up a new world of learning to people with few educational opportunities has long been a goal for educators interested in technology and open source advocates interested in equity. Online courses have been around for awhile, but up to now they have been limited in size and run by prestigious universities with tuition to match, or by less distinguished schools or organizations. Now, for the first time, UC Berkeley and Harvard-caliber institutions are delivering high-quality online lectures and quizzes taught by professors from those schools. They are doing it massively, and they are doing it for free using a format with a clunky acronym: massive open online courses (MOOCs).
Incubated at top-level universities such as Stanford and M.I.T., the new MOOC model is spreading ripples of excitement and fear through the academy. Whatever people might think about MOOCs, both pro and con, one claim made about them is indisputable: they are turning traditional higher education on its head. Every institution of higher learning, from large and public to small and private, from community college to for-profit, vocational or some other hybrid of all of the above, is closely monitoring the MOOC movement. Professors, administrators and students are all keeping an eye on MOOCs’ potential to disrupt what has been the dominant path of secondary education.
Silicon Valley entrepreneurs and investors alike see the MOOCs as a potential new business opportunity and one that is aggressively challenging the established model of the university and its bottom line. As Philip DiSalvio, dean of University College at the University of Massachusetts, points out in a New England Journal of Higher Education blog post, "If
2colleges were businesses, they would be ripe for hostile takeovers, complete with serious cost- cutting and painful reorganizations."1
Exactly how MOOCs can do that, and whether they can make a profit in the process, is the main topic of this paper. It begins with a brief background and definition of MOOCs and how they compare to other kinds of Silicon Valley startups. The paper then analyzes three different business models or profit centers that MOOCs could operate under in the future. Finally, it examines the meaning of MOOCs for the transformation of higher education and for a post- industrial economy.
What the Heck is a MOOC?
The format of the massive open online course is usually credited to two Canadian thinkers who created the online course “Connectivism and Connective Knowledge.’’ George Siemens, an educator and theorist focused on learning in digital environments, coined the term connectivism, a theory of learning specifically relevant to the digital medium and social networking technologies. Stephen Downes, a National Research Council of Canada researcher, developed some of Canada's first online courses at Assiniboine Community College in Manitoba.
Connectivism is the theory that knowledge is distributed across a network of connections, and that learning consists of the ability to construct and traverse those networks.3 Downes and Siemens’ course is a “connectivist’’ class about connectivism (other topics covered include Net Pedagogy, Personal Learning Environments & Networks, and many others). Participants use Twitter, RSS feeds, blogs and virtual world interfaces such as Second Life to explore learning models of the future. The Connectivism MOOC debuted in 2008 for University of Manitoba students. Today, it is offered for college credit to tuition-paying students at Manitoba, but the ourse is also open to anyone with an Internet connection. Implicit in the spirit behind the MOOC movement is that education be Open to all, and being Online facilitates that openness, just as it allows courses to be delivered to Massive numbers of people simultaneously. The remaining letter, C for Course, speaks for itself.
Since Siemens and Downes facilitated their first MOOC, the medium has come a long way, mostly through trial and error and the involvement of visionary computer science professors. In the summer of 2011, Stanford decided to open up courses on artificial intelligence, machine learning and databases to any student with an interest. In order to make available the classes in an online format, computer science professor Daphne Koller and a group of students built a platform from scratch. That fall, each of the three courses counted over 100,000 enrollees.
Koller’s eureka moment had come four years previously when a group of faculty was asked to find a way to inject more meaningful interaction between faculty and students. Koller imagined replacing the standard one-hour lecture with multiple eight-to-ten minute units that would individually engage each student by pausing to ask questions which students would need to type in (Salman Khan revolutionized the short video format for learning with Khan Academy). She describes the shift as a movement away from filling minds with content toward active learning and interacting in the classroom.
This brainstorm led to the founding, with another Stanford professor, Andrew Ng, of Coursera, a for-profit company now offering more than 200 courses in math, science, engineering and more to hundreds of thousands of students. About the same time, another group of Stanford-affiliated professors and roboticists opened their first course online, Introduction to Artificial Intelligence. Within weeks, 160,000 from 109 countries had signed up. The result is Udacity, also a for-profit firm but more focused on science and technology with about 20 classes currently on offer. The third university-inspired MOOC that is getting the most attention in the press is edX, a not-for-profit enterprise started by Harvard and Massachusetts Institute of Technology.
The amount of attention these three university-launched MOOCs have received over the past year or so cannot be overstated. Each time they announce a new university partner in their ventures, the “big three’’ MOOCs seem to raise a new wave of publicity. However, it is important to note that while Coursera, Udacity and edX have received the most visible attention, there are many MOOCs that exist outside these high-profile platforms. In fact, the original pioneers of MOOCs have worked to differentiate their connectivist-inspired MOOC designs from the commercial and well-funded MOOCs that offer a more traditional course structure, using videos and online quizzes for instruction, often within a set timeline.6 It is this more traditional model that is being embraced by universities in the U.S. at a rapid clip. This paper is focusing primarily on the business potential of this more commercial MOOC model, particularly as used by Coursera and Udacity.
Coursera has agreements with over 30 university partners including Columbia, California Institute of Technology and UC Irvine, in which professors from those schools teach courses on the site. Udacity courses are taught by university professors and entrepreneurs from Silicon Valley, such as the founder of Reddit, a Web feed that tracks online trends. edX classes are taught by faculty from Harvard, M.I.T. and other partner schools including UC Berkeley.
How Do MOOCs Compare to Other Silicon Valley Startups?
Like other famous Silicon Valley start-ups such as Google, Facebook and Twitter, the for- profit MOOCs have focused on growing membership as a first step and leaving potential monetization strategies for later. However, with membership growing faster than expected at MOOC sites, business models have moved to the fore more quickly than they may have for tech startups of the past.
Compared to the monetization of Facebook, Twitter, et al., it doesn’t require quite the creative reach to see how Coursera and Udacity could become profitable. Since the MOOCs provide a tangible service that many people need and want, they have several clear-cut avenues to profitability that do not rely on advertising.
Last summer, under a Freedom of Information Act request, the Chronicle of Higher Education obtained the first contract Coursera signed with an institution of higher learning, the University of Michigan.7 In the contract, Coursera lists eight possible modes of monetization, including banner advertisements and corporate sponsorship (the article points out that the university partners have the right to reject monetization plans on a course by course basis).
Other ideas for turning a profit outlined in the contract were selling course content to community colleges, selling course content to universities who in turn charge tuition to their own students taking the courses on campus, and offering secure assessments, such as proctored testing. At least two of these strategies is already happening (licensing out course content and selling assessments). The methods of monetization that the companies have been touting most aggressively, however, revolve around job placement services, credentialing and data.
Job placement
Recommending and connecting Courserians (members who are enrolled in classes) to potential employers for a fee is a business model already in place at the for-profit MOOCs. Coursera announced in December an employer matching service, Coursera Career Services, which will begin with software engineering students. Through the program, participating employers receive a list of qualified students who have opted to participate. Coursera acts as the matchmaker between an interested employer and a student and makes introductions via email.8 If the communication results in a hire, the hiring company pays Coursera a flat fee, and the college offering the class also receives a percentage.
Udacity’s job placement service is more hands-on.
Like Coursera students, Udacity alumni opt in to allow their curriculum vitae to be viewed by companies seeking new hires. Other information is also shared with potential employers, such as the level of distinction a particular students reached in each course. “Collecting karma’’ by helping others on Udacity’s Web forums also can boost job candidates’ eligibility for software developer positions being filled by the 20 or so high-tech firms currently working with Udacity.9 Finders’ fees for Silicon Valley headhunters typically total 20 percent of a software engineer’s starting salary, which would translate into $15,000 per match.
As a business model, job placement services have the advantage of involving individual job seekers and the MOOC companies directly, while the universities are only tangentially involved. Placing universities at the center of money making strategies brings up some thorny issues, as seen in the quest for college credit for MOOC classes.
Credentials
At the moment, the biggest obstacle to MOOCs achieving relevance in higher education long term is their lack of credentialing. As wonderful as it is that 100,000 students across the globe, from Ann Arbor to Jakarta to Vienna, can take a high quality introduction to finance course for nothing, it is unclear how or whether that translates into something tangible, such as a job or college credit. That uncertainty is reflected in Coursera’s completion rate, which is about 12 percent.11
While many universities offering courses through a MOOC will provide a certificate of completion, the barriers to MOOCs offering college credit are formidable. Universities see the value of college credit as providing personal contact between instructors and students. That level of interaction isn’t possible when there are thousands of students in a course. The personal interaction, after all, is the foundation of the traditional fee-for-credit system that has been the university business model for generations. And so far, there doesn’t seem to be any simple way to replace it. But universities are trying.
The issue of credit-earning MOOCs is a topical one and is being examined on several fronts. In November, the Bill & Melinda Gates Foundation awarded $3 million in grants to study
8
MOOCs. Most of the projects being funded seek to answer how universities can maintain the quality of their courses while making them more affordable.12 Another chunk of the Gates funding will go to the American Council on Education to assess whether some MOOCs could qualify as credit-bearing courses. As part of the study, faculty teams will evaluate online courses. Students seeking to earn credit for one of them would take a proctored exam on the material offered in it and those who performed well would be given the option to purchase a transcript of the course to submit to a university.13
Some colleges are not willing to wait for the model to be tested. In October, Antioch University, a private system with five campuses in the U.S., announced a deal to offer its undergraduate students college credit for two Coursera classes developed by the University of Pennsylvania. The classes, Modern and Contemporary American Poetry and Greek and Roman Mythology, will be offered to students attending the Los Angeles campus, and will cost less than half what a typical course costs at the school.14 As part of the agreement, Antioch will pay Coursera a fee, essentially licensing the course content, one of the original monetization strategies envisioned by Coursera. Whether other universities will follow suit remains to be seen. However, paying for the use of an online platform like Coursera or Udacity would seem to cost a university in reputation. It’s one thing to participate in the next big thing by designing an innovative new curriculum or teaching in a futuristic format. But to pay a third party for content or exposure really marks a brave new world. Outsourcing Testing
There are other ways for MOOCs to make money on credentialing and student assessment that do not involve the universities directly. One method to provide MOOC takers
9
with a layer of credibility is to have them tested by a third party. This model is also directed at the cheating issue, which has emerged as a serious problem for MOOCs. It turns out that people working by themselves without any human monitoring tend to cheat at much higher rates than students in physical classrooms.15
Under this model, the student is the one who pays — about $85 a pop — for a proctored exam that will earn them tangible proof of performance. In many cases, however, it will be up to the student to take their certificate and present it to a university for credit or to a potential employer. All of the large MOOC providers have waded into this style of credentialing. Udacity announced it was partnering with Pearson Education Inc., which runs testing centers in 175 countries.16 At least one school, Colorado State University’s Global Campus, said it would give academic credit for a course on building a search engine if students taking the proctored exam. At $85 per student, third party testing may be a viable monetization strategy for MOOCs with classes of hundreds of thousands. And it offers a level of oversight of student performance.
Data
The third major area of potential for monetizing MOOCs is data. This is probably the most controversial and least thought-out options for making money from MOOCs. It is also possibly the most intriguing. As Coursera co-founder Daphne Koller has said, the massive scale of participation in the popular courses puts the MOOC firms in a unique position. “You can collect every click, every homework submission, every forum post from tens of thousands of students,’’ Koller said at a TED talk in June. “So you can turn the study of human learning from a hypothesis-driven mode to a data-driven mode. You can use these data to understand fundamental questions such as what are good learning strategies versus ones that are not.’’EdX, the nonprofit MOOC initiative from Harvard and M.I.T., has listed student learning data collection as one of its primary goals for the purpose of sharing freely with education researchers.
Koller says her company would most likely use such metrics to personalize feedback in course material and improve instruction and learning. But the potential for selling data captured from tens of thousands of students is unimaginable and would probably be very tempting. In a recent briefing on MOOCs, Educause, the nonprofit association devoted to information technology in higher education, outlined some possible business streams for MOOCs that would probably fall into the highly controversial category:
• Data mining: Sell student information to potential employers or advertisers.
• Cross- or up-sell: Course materials (e.g., videos) are freely available, but ancillary services
like assignment grading, access to the social networks, and discussions are fee-based.
• Advertising model: Courses have named sponsors.
What does this mean for the transformation of higher education?
Universities are under a great deal of pressure to to cut costs and make education more affordable while maintaining their reputations for high quality. Funding shortages due to the massification of higher education have shifted the responsibility for generating larger percentages of revenue onto university departments. This trend has been exacerbated by a widespread political inclination toward greater privatization of services previously provided by the state, including in higher education.18
At the same time, in the age of austerity, public voices questioning the return on investment of a college education are growing louder. Outstanding student loan debt stands at $956 billion.
The unemployment rate among recent college graduates is 9.5% and 39% of that group is mal-employed, or underemployed.20 And yet, universities are raising tuition and fees above the rate of inflation just to keep afloat. Members of the National Association of Independent Colleges and Universities raised fees an average of 3.9% for 2012-13, almost double the 2% increase in the U.S. Consumer Price Index.21 Tuition and fees at public institutions have risen even more — up close to 5%. Companion surveys from The Chronicle of Higher Education and the Pew Research Center in 2011 found that 80% of the general population think that a college education is not worth the price.
Given the enormous pressures of an economy requiring advanced learning, fiscal austerity at the state level and the drive toward privatization and competition in higher education, it is no surprise that innovations like MOOCs are appearing on the short list of potential solutions.
MOOCs, Vocational Education and the Economy
One natural outgrowth of the MOOC movement is the return to a more vocational model of education. The dismantling of the vocational component in U.S. public education and its impact has been well documented in the skills gap between employer needs and student skills.
There is a growing body of literature questioning the structuring of education for the knowledge-based economy that is the product of globalization in the wealthier countries. This move toward knowledge-economy training at the expense of job-ready vocational training is increasingly being seen as a false dichotomy.22
The MOOCs reflect the synthesis of knowledge-economy skills with employer-based competency training (at a price that is unbeatable). With their emphasis on technology and science-related course material, edX and Udacity offerings certainly have the potential to help fill the gap in this sort of training.
Until the credentialing issue is resolved, MOOCs are left open to criticism that they only reinforce existing gaps in knowledge and access to legitimate employment. Education via MOOC could create a bifurcated system in which less-prestigious universities offer MOOC courses for credit at a lower fee, said Larry Cooperman, director of OpenCourseWare at UC Irvine, which recently added six courses to Coursera’s offerings, including public health, pre- calculus and personal finance planning.
That outcome is exactly what happened with the deal between Antioch University and Coursera. “The conversation in people’s houses is, where are we going to send Johnny to school next year, and the idea that you would send them to Coursera or Udacity doesn’t exist right now,’’ said Cooperman. At the moment, MOOC courses for credit is for a certain kind of student. “We find it problematic that instead of democratizing the field we give greater advantages to the already advantaged.’’
The idea of vocational education as somehow inferior is an old one. But new models that may eventually translate into both profit and prestige are cropping up that challenge those assumptions. The Mozilla Foundation, creator of the Firefox Internet browser, has developed source code for Open Badges, which are digital credentials. Badge issuers so far include universities, organizations and government institutions such as Carnegie Mellon, Girl Scouts of America and NASA. Badge issuers use the Mozilla source code to create the parameters of their badges. Users complete the requirements to earn the badges and then compile them into badge backpacks that can be posted on social networking sites or packaged and linked to resumes.
The badges themselves are also embedded with meta data, so that a human resources director clicking on a badge earned from a university would be able to see not only detailed information about the person who earned the badge but also the test scores upon which the badge was issued. This type of high-tech credentialing is already making its way into some MOOC courses and represents both an avenue from education to employment and a monetization strategy: a Purdue University subsidiary that offers online courses in nanotechnology is issuing digital badges for $30 to course completers.
Conclusion
MOOCs are clearly a groundbreaking innovation in the rapidly shifting landscape of education, employment, and technology’s ability to bridge that gap. Their evolution from online education to massive and open to all represents a huge leap forward, but one that is filled with uncertainty.
The three areas of potential monetization for the commercial MOOCs, job placement, credentialing and data, are all in some form of being adopted or are already generating some revenue for MOOCs.
Despite their clear motivations as profit-seeking firms, the commercial MOOCs are all rooted in the academy. The people behind them are well aware that the reputations and longevity of their businesses depend on being part of the solution to reforming education to make it more affordable and more relevant in the future economy.