ArsDigita Archives

Collaborative E-Learning: The Right Approach

by Kathleen Gilroy, CEO, The Otter Group (

Submitted on: 2001-03-28
Last updated: 2001-03-28

ArsDigita : ArsDigita Systems Journal : One article


We are at a critical moment in the evolution of e-learning. After many years of development, e-learning has become an important business process for corporations, which are now exploring how to better educate and manage their employees who rely on fresh knowledge to perform. E-learning is also at the top of the agenda of public and private universities, which are looking for ways to extend their influence and reach new types of customers. And e-learning has attracted the attention of the investment community as companies have emerged to capture market opportunities in technology, content, and services.

Despite all the recent activity, models for how people teach and learn online are still immature. A new approach could take care of the long-standing e-learning bugaboos: low enrollments and high attrition rates stemming from user dissatisfaction. The cause of this problem is the separation of people in time and space; but it can be overcome by building environments where people talk to one another, build relationships, and teach one another.

While there is no simple answer, there is one key idea that has been overlooked in the design and implementation of many of the e-learning programs on the market today: Learning is fundamentally both social and experiential. It is the context of the learning--all of the elements that comprise the experience around the content--that is most important.

This paper will lay out the OTTER Group's model for how best to teach and learn online. It will look at many of the elements that must be managed to create e-learning programs where real knowledge is gained, where communities of learning are created, and where high levels of student satisfaction are generated.


The emphasis of most e-learning programs to date has been on the accumulation, organization, and delivery of content. This is manifested in all aspects of how the new sector has been organized: in the business and operating models of the service and technology providers; in the design and organization of the content and learning management systems that are now widely used; and in the investments venture capitalists, publishers, universities, and corporations have made.

With all of the information now available on the Web, it is possible to find really good content on just about any subject. And once information becomes digital, it wants to become free--so much so that MIT is now considering making the underlying content of all its 2,000 courses available on the Internet without charge. In a recent article in The Chronicle of Higher Education, MIT's president, Charles Vest, was quoted as saying: "I think we're in a kind of brief shining moment in general in that the World Wide Web is making information available to the world for free. I would like to think that, for at least a brief period of time, we could be a leading source of higher education on the Web." According to MIT Civil Engineering Professor Steven Lerman, MIT can make this content available free of charge because, "the syllabus and lecture notes are not an education. The education is what you do with the materials."

MIT is visionary in realizing that it can give away its syllabi and lecture notes and retain what is of real value: the navigation through these materials by superb teachers (the pedagogical process); the social aspect of learning with incredibly smart peers (the learning community); and the knitting together of content, pedagogy, and community into a unique learning experience, which is what people are buying when they step on to the MIT campus.

Former OTTER students, even when working with the best professors at Harvard and MIT, have told us that they value their interaction with one another as much as they value the content being delivered by their professors. Thus, the ideal class is organized around what we call the 50/50 rule. At least 50 percent of the time students spend in the virtual classroom is spent interacting with and learning about other students. When the social aspect of the classroom is missing, student dissatisfaction rises dramatically, as does the attrition rate.

When we think about our own learning experiences, we remember not only what we learned, but how and where we learned. A very large part of the value we derive from our educational experiences comes socially and informally--from the context. It comes from the relationships we build around the substance of what we are learning. In college, it's the late-night study sessions, the interesting conversations around the lunch table, an informal chat during a professor's office hours, the insights of our classmates in a case study discussion. At conferences, seminars and off-site meetings it's the networking. The rich context that we gain from informal, peer-to-peer conversation is often what helps us make the content more memorable and useful.

In most e-learning programs offered today, the burden for learning is placed wholly on the shoulders of the learner. When a learner goes to a course web site, she enters a grid that does not vary from course to course, consisting of a menu of activities: announcements, documents, assignments, external links, communications, and tools. The course is served up as content that is devoid of any context. She is expected to navigate this material on her own, without much support. She is offered email links to faculty and other students, but not much more.

E-learning should be first and foremost about creating a social space that must be managed for the teaching and learning needs of the particular group of people inhabiting that space. This requires a platform that can be easily modified to take into consideration the needs of the particular learners in the course. In an optimal arrangement, a student will know a great deal about his fellow students and faculty before he begins working through the material. He will be prompted with questions that have been very carefully designed to encourage him to link the material he is learning to his own knowledge and experience, as well as stimulate him to interact with other students and the faculty via email and chat. This model will use the database underlying the course to link people and information in new ways that will help him understand the community of learners he has joined, as well as affect his relationship with the material itself.

With the right enabling technologies, the learner can take advantage of the context in interesting new ways: if she thinks that someone has posted something particularly insightful, she can choose to automatically filter out every posting that individual has contributed to the course. She can rate her fellow students' postings and emails and have the system mine the data for the most highly rated information. Even in an online discussion with thousands of comments posted by hundreds of students, the most valued information will automatically be recognized by the professors and read by all of the learners. Such a system also allows for something that is often overlooked in the e-classroom: recognizing and acknowledging the most valuable contributors.


When the focus is no longer content but rather the management of the learning experience, then the pedagogical process becomes the most important factor in the design and support of that experience. And that process is fundamentally idiosyncratic. It's also what makes learning pleasurable and beautiful. To experience a wonderful teacher's pedagogy is to be inside her mind.

To be effective, each course must be customized both to the pedagogical process of the teacher or subject-matter expert and to the individual needs of the learner. Currently, customization of online learning programs often amounts to changing color schemes or turning features of the platform on or off.

But customization at a deeper level can mean changing the fundamental organizing principles of the course space. In a pilot course the OTTER Group is currently running, we wanted an organization that reflected the functional tasks of a global securities trader with very little time. The design was for a small group of people (30) focused on very specific tasks. Many complex ideas were presented in a linear fashion, and the students needed to be guided by specific questions. This tight focus demanded a course environment reflecting that task-oriented need. To guide students, we made class discussion the most prominent object on the screen, followed by the group project area. Document exchange--relatively unimportant for this course-was placed in the least prominent spot. We employed simple page designs, making them task-oriented rather than information-oriented by emphasizing both the names of the task areas within the course space and deliberately organizing the functions that were accessible in each of those areas. Rather than have a "home" page, we made "discussion" the root page of the class. We also created a directory page called "Who's Who" because we wanted to encourage students to get to know each other.

The importance of these types of contextual changes is best illustrated in one of my favorite books of 2000: The Tipping Point, by Malcolm Gladwell. In his chapter on "The Stickiness Factor" he explains how the producers of Sesame Street made television sticky: "They discovered that by making small but critical adjustments in how they presented ideas to preschoolers, they could overcome television's weakness as a teaching tool and make what they had to say memorable. Sesame Street succeeded because it learned how to make television sticky."

Gladwell analyzes two Sesame Street segments to see how the context and formatting affect learning and retention. One involves the spelling of the word HUG by a female Muppet. In this segment the letters HUG are the central feature on the screen. In a second segment involving the spelling the word CAT, Oscar is the central feature. Researchers at Harvard's School of Education tracked pre-schoolers' eye movements and found that the students focused on the letters in the HUG segment but on Oscar in the CAT segment. The kids weren't watching the letters because Oscar was so interesting: "Oscar was sticky. The lesson wasn't." This seemingly small adjustment in the context of how the information was presented had enormous implications for how much was actually learned.

In another example of how to create stickiness, Gladwell cites an experiment done at Yale in the 1960s on increasing the likelihood that students would get tetanus shots. It turns out that variations in the type of information given the students about the dangers of tetanus had no impact whatsoever on the likelihood of their getting vaccinated. Only one thing dramatically raised the rates of vaccination from 3 to 28 percent: "including a map of the campus, with the university health building circled and the times that shots were available clearly listed."

Gladwell's analysis of the importance of contextual and formatting innovations should be mandatory reading for anyone thinking about using the Internet as a teaching tool. These small but critical adjustments in context and format are just as important to the learning process of global traders as they are to pre-schoolers.


An enormous amount of lip service has been paid to the creation of learning communities in the e-learning arena. Most of the services promoted as communities are really just transactional: fronts for the sale of content. The truth is that a community is very difficult, if not impossible, to create from scratch--at least without enormous resources (time and/or money) matched with deep social and structural understanding. It is much easier to recruit and organize an existing community into a learning community than to start fresh.

Sociologist Amitai Etzioni, in a recent article in Contemporary Sociology, defines communities as having the following attributes:

  • A web of affect-laden relationships among a group of individuals, relationships that often crisscross and reinforce one another (rather than merely one-on-one or chainlike individual relationships).
  • A measure of commitment to a set of shared values, norms, and meanings and a shared history and identification within a particular culture.
Etzioni expands this definition to say that communities need not be territorial: "many ethnic, professional, gay, and other communities are geographically dispersed." Also, the bonds that are found within communities go beyond the instrumental into the realm of affection. And they are at some level exclusive.

It is important to keep these definitions in mind when developing learning communities. In the OTTER Group's work with universities, for example, we have been most successful when we have drawn upon alumni as the starting place for marketing and managing e-learning programs. An alumni group fulfills all of the criteria Etzioni lays out: an exclusive group whose relationships crisscross and reinforce one another, by virtue of common cultural experience, with a set of shared values, norms, and meanings, linked together with bonds of affection.

In the early pre-web 1990s, we developed Harvard Business School's first distance learning program, a seminar via satellite on service management, marketed to HBS's 50,000 alumni. This proved to be an extraordinarily successful program, in part because it tapped into the affiliations and knowledge that were already in place within this tightly-knit community. First and foremost the learning experience was very simple to market: we conducted one direct postal mailing to the HBS alumni list of 50,000. Our extremely high conversion rate of 7,000 participants has only been replicated when we were marketing to communities of similarly high degrees of affiliation: other alumni groups and the extremely tightly bound American Library Association.

For HBS service management program, only 1 in 6 of the attendees were actual HBS alumni. The rest were "guests"--colleagues, employees, and friends that the alumni had referred of their own accord. The service management learning experience was extremely well received by this community of participants, with an overall satisfaction rating of 9 out of 10. Over 90 % of the 7,000 participants said that they would buy similar e-learning programs from HBS. We would not expect to find such loyalty among students derived from a random sample.

In a current pilot program on the science of persuasion as a social skill, the learning community begins with the marketing of the course. We are sending emails to the students who have taken the course and asking them to go to our web site where they can add comments about how they have successfully used the material in the course in their personal and professional lives. Specifically, we are asking them to debunk the three major misconceptions people have about the mastery of persuasion: that it is innate and cannot be learned; that it is self-evident and therefore trivial; and that its practice is unethical.

This inquiry is intended to reactivate and engage the community of course alumni who will themselves refer us to interested buyers within their organizations. Further, the responses of former students will build a set of multiple points of view about the course to be read by prospective students. By reading testimonials that are focused on the alumni's successful experiences in applying the course knowledge, prospective students will have a chance to see the existing community around the course and imagine themselves fitting into it. The same process applies to working with corporations: draw from a pool of people who are already affiliated with one another; make selection for the course exclusive; and tap into the connecting and selling energy of managers and sponsors.

To make all of this work, a new category of community manager is needed: the Learning Director. For university programs, Learning Directors are drawn from the alumni of the school or course. In corporations, they are drawn from the ranks of key managers and/or the training and education departments. They are trained to act in the role of what Howard Gardner calls pedagogistas. Learning Directors have some knowledge of the content, but their expertise is really in the context: they understand the personal and organizational issues around engaging the students. Learning Directors make sure the students feel connected to the professor, material, and one another. They highlight student comments that are very insightful or relevant, and they prod somebody who has not spoken up in three weeks.

In our course on financial technology for global traders, the Learning Director understands the course materials; he is also a bank insider who knows how the course is going to be received by the learners. He knows what will interest them, and how they can apply what they are learning to their business practices. He acts as a mediator between the professor and the students. He selects the most relevant of the professor's questions to highlight for student discussions. He also selects the best comments and ideas to flow back to the professor.

Learning directors also serve as the beta testers for our pilot programs, allowing us to create programs that are highly interactive at large scale. We like to think of them as the Avon Ladies of the knowledge economy.


One of the great advantages of the web is that it can organize information so that it is personalized to an individual's needs. But today's e-learning programs are often organized around the needs of the content providers, not around those of the individual learners: students are served up homogenized, standardized content "course cartridges" and "e-packs." In the physical world, people organize their own notebooks, choose their study techniques and even pick where they will sit in the classroom based on their own needs. Personalization is an area where the power of databases can rival offerings in the physical classroom. Information can be organized in such a way that learners are given only what they need when they need it. They also can be given total control over their learning environments. Meaningful user-controlled personalization is something that needs to be incorporated into e-learning design from the beginning, rather than as a frill or afterthought.

The Web offers the ability to create deep profiles of students and use that information to create personal, unique learning experiences. Profiling is more than just finding out what skill and information gaps the students have. It is about understanding the learner's context as a whole human being and shaping the content and course experience accordingly.

Sophisticated polling methods can be used to build a set of independent variables about students that can later be cross-referenced with questions that deepen their understanding of one another and the material they are studying. In the OTTER Group's course on persuasion, we have students take a 360-degree evaluation over the Internet. They assess themselves along several key metrics and then they have ten people (bosses, spouses, subordinates, clients) assess them along those same metrics. This assessment is handled anonymously via email and then collated in a central database on the web. Once these profiles are built, they can be used as reference points in teaching case materials.

For instance, because it turns out that the buyers of SUVs and Minivans are very different psychologically, one of the questions we asked students in the persuasion course is which they'd prefer to buy. According to an article in The New York Times, "Sport utility buyers tend to be more restless, more sybaritic, less social people who are 'self-oriented,' to use the automakers' words, and who have strong conscious or subconscious fears of crime. Minivan buyers tend to be more self-confident and more 'other-oriented'--more involved with family, friends and their communities." The SUV vs. Minivan distinction proved to be an interesting variable in how groups of students made decisions about retaining or firing struggling managers in our case study discussions. In a case in the course, a manager ineptly handled a politically charged situation with two warring bosses. After reading the case, students were asked to decide whether the manager should have been fired or retained. The class was equally divided. Interestingly, SUV drivers-i.e., those who were more self-oriented--were more likely to want her fired than minivan drivers. This exercise ended by asking students what they thought her job title is today, ten years later. Most students placed her as either VP of Manufacturing or unemployed. (She is currently the CEO of Handspring.) Cross-referencing unusual and compelling personal data not only helped this particular "learning community" understand itself better, but it also changed the students' relationship with the material itself by giving them insight into their own decision-making processes.


The technology platform that the OTTER Group prefers for our e-learning programs is both free and open source. It was originally developed by an MIT computer scientist, Philip Greenspun, to support collaboration and knowledge-sharing among a community of amateur photographers. It is worth studying the operation of this true learning community (, as there are many lessons to be learned there for teaching and learning online. Greenspun first created to share what he knew about photography. He started by writing a 30-page story about a trip to Berlin and Prague, illustrated with about 60 photos. He then invited people to contribute their experiences. By simply including an "add-a-comment" link at the bottom of every article, he helped users build a great repository of photographic knowledge.

The company that has built our platform of choice, ArsDigita, was founded by Greenspun and evolved out of The toolkit upon which the platform rests, the ArsDigita Community System, places social interaction and collaboration--the management of the learning experience--at the center not only of its design, but also of its operating philosophy. One of the great strengths of the open source software movement and the reason that it yields superior programs--programs that reflect the specific needs of the communities of people and individuals who are using the software-is that you have access to a large community of developers with many ideas. With closed source software, you only have as many ideas as the group of people who created the program generates or as their lawyers will allow.

For example, the next version of the ArsDigita software will contain chat protocol that is written with academic exchange and collaboration in mind: chat postings can be threaded as they are posted so that inquiry and answers between students and professors can be well documented. Automatic transcripts of chats are generated that can be organized by time and by thread. Because the software is open source, this kind of enhancement can be readily added to the functionality of the platform by any of the many people working on the system, and this best practice of archiving threaded chat can then be used freely by others.

As more users adopt the software and adapt it for their specific needs, the best practices in e-learning will be shared by getting built into the software itself. In the past year, the ArsDigita Community System has been developed as a content and learning management system, now called ACES (the ArsDigita Community Education System). You can read the case study about its use at MIT's Sloan School of Management, as well as download the software, at:

We believe open source to be the best technology strategy for the development of shared knowledge and learning. In a recently published book on the open source software movement, Rebel Code: Inside Linux and the Open Source Revolution, author Glyn Moody talks about the values that have driven the development of the Internet: "as the Internet moves closer to the heart of the modern world, it inevitably carries with it the free programs that drive it, and seeds the values that led to their creation. Its basic code of openness, sharing and cooperation is starting to spread outside the confines of one or two high-tech industries."

We believe that the code of openness, sharing, and cooperation that is at the heart of the e-learning process is also at the heart of successful academic institutions and corporations. At OTTER, openness is a pivotal part of our company's technology and operating strategies and value system. Sharing knowledge and ideas is one of the great joys of being human. We take advantage of that phenomenon in our course design. We pay careful attention to it in the technology platform we choose. We have the most meaningful jobs in the world: we connect people-including some of the world's great teachers-across time and space to share ideas, knowledge, and wisdom so that they can teach and learn from one another. And we learn so much in the process.

Cambridge, Massachusetts
March 28, 2001


Michael Feldstein made many significant contributions and helped write sections of the paper. Sarah Milstein's editing gave it clarity and focus. Philip Greenspun's book and seminar also greatly helped refine my thinking about the topics covered.