BookPDF Available

Thwarted Innovation: What Happened to E-Learning and Why

Thwarted Innovation
What Happened to e-learning and Why
by Robert Zemsky and William F. Massy
A Final Report for
The Weatherstation Project
of The Learning Alliance at
the University of Pennsylvania
in cooperation with the Thomson Corporation
copyright © 2004 by The Learning Alliance at the University of Pennsyvlania
A Learning Alliance Report
Thwarted Innovation •••••••••••••••Page ifc
Table of Contents
ii Acknowledgements
iii Report Summary
4Where’s the Data?
5What’s the Concept?
7Chapter 1: The Dynamics of Innovation
7Utterback and the Emergence of a Dominant Design
9 Innovation’s S-Curve
10 e-learning’s Adoption Cycles
12 Framing Questions
13 Chapter 2: Plausible and Implausible Measurement
13 From Example to Projection
14 Surveying the Terrain
17 The Measurement Challenge
18 Chapter 3: A New Measurement Strategy
18 Campus
and the Interview Process
20 Two False Starts
20 The
22 Chapter 4: First Findings
22 Tracking e-learning’s Four Adoption Cycles
26 Attitudes and Expectations
28 An Interpretative Frame
29 Shifting Institutional Priorities
30 Faculty vs. Administrative Volatility
31 Making Sense of a Mosaic
33 Chapter 5: The Corporate Market for e-Learning
33 The Shape of the Provider Market
37 Tracking the Corporate Market
40 Googling the Market
42 Do We Have a Market Tracker?
44 Chapter 6: e-learning’s Troubling Assumptions
44 Assumption 1: “If we build it, they will come.
48 Assumption 2: “The kids will take to e-learning like ducks to water.
52 Assumption 3: “E-learning will force a change in how we teach.
54 A Fourth Assumption
56 Conclusion: What’s Next?
57 Contextual Changes
58 Technological Changes
59 Market Conditions
59 Three Practical Steps to Start the Process
60 Not the End of the Story
61 Appendices
he Weatherstation Project
was first
broached one night at a dinner of the
executive committee of the National
Center for Postsecondary Improvement (NCPI).
Eventually the conversation turned to
discussions of what each of us was planning to
do once work for NCPI had been completed. Our
suggestion that we were thinking about studying
the market for e-learning was met immediately
with guffaws—“The ink will be hardly dry on
your report when it will be out of date!” Not the
case, we responded: We wouldn’t be writing a
report but establishing
to track
the changing climate for e-learning both on
college campuses and across corporate America.
In the end, we did both—and in each case, our
debts to others are substantial.
First and foremost we thank the leaders on
the six campuses who allowed us to establish
that made this project possible.
More often than not, it was they who explained
e-learning to us rather than vice versa. In many
ways, this report is theirs: Vivian Sinou, Dean,
Distance and Mediated Learning, Foothill College
(CA); David Smallen, Vice President for
Information Technology, Hamilton College (NY);
David A. Gift, Vice Provost for Libraries,
Computing and Technologies, Michigan State
University; Dean L. Hubbard, President,
Northwest Missouri State University; James
O’Donnell, then Vice Provost, Information
Systems and Computing at the University of
Pennsylvania and Peter Conn, Penn’s Deputy
Provost who took over for Jim when he left to
become Georgetown’s Provost; and Judy C.
Ashcroft, Associate Vice President and Director,
Thwarted Innovation •••••••••••••••Page ii
Division of Instructional Innovation and
Assessment, the University of Texas at Austin.
The Project was managed by Pamela Erney—
with style, patience, and persistence. The website
that tracked the changing attitudes of the faculty
and staff on the six campuses was designed,
executed, and managed by Barbara Gelhard. The
initial design of the tracking instruments was
greatly shaped by our colleague Susan Shaman.
Masako Kurosawa of Japan’s National Graduate
Institute for Policy Studies helped us think
through the labor market implications of e-
learning. Our editor Marc Iannozzi, as he has so
often done in the past, made sure we actually
said what we meant to say.
The tracking of e-learning across the Web and
the management of the
done by three accomplished graduate students at
the University of Pennsylvania: Jesse Lytle, now
of Mt. Holyoke College, Aimee Tabor, and Liza
Herzog. They, quite literally, made the project
work by making the work fun.
Our final debt is to the Thomson Corporation.
The officer with whom we worked most closely
was Michael Brannick, President of Prometric.
As Michael would be the first to say, our results
proved “quite a bit different from what. . . [he]
envisioned” when we launched the project three
years ago. Despite his misgivings, he helped at
every turn, sharing his experiences as well as
his contacts with us.
Robert Zemsky
William F. Massy
June 2004
Report Summary
hwarted Innovation
is a major new study
from the University of Pennsylvania in
collaboration with the Thomson
Corporation, which answers the question: “Why
did the boom in e-learning go bust?” Researchers
Robert Zemsky and William F. Massy used e-
learning Weatherstations at campuses across
the country to decipher precisely what happened
and why. In the end, they trumped three of e-
learning’s most troubling assumptions.
If we build it they will come—not so; despite
massive investments in both hardware and
software, there has yet to emerge a viable
market for e-learning products. Only course
management systems (principally BlackBoard
and WebCT)—and PowerPoint lectures (the
electronic equivalent of clip-art) have been
widely employed. At the institutions partici-
pating in the study, more than 80 percent of
their enrollments in “online” courses came
from students already on their campuses.
•The kids will take to e-learning like ducks to
water—not quite; students do want to be con-
nected, but principally to one another; they want
to be entertained, principally by games, music,
and movies; and they want to present them-
selves and their work. E-learning at its best is
seen as a convenience and at its worst as a
distraction—what one student called “The fairy
tale of e-learning.
E-learning will force a change in the way we
teach—not by a long shot; only higher
education’s bureaucratic processes have
proved more immutable to fundamental
change. Even when they use e-learning prod-
ucts and devices, most faculty still teach as
they were taught—that is, they stand in the
front of a classroom providing lectures in-
tended to supply the basic knowledge the
students need. Hence, we see the success of
course management systems and PowerPoint—
software packages that focus on the distribu-
tion of materials rather than on teaching itself.
What is
Thwarted Innovation
’s conclusion? E-
learning will become pervasive only when
faculty change how they teach—not before.
Thwarted Innovation
refocuses the debate over
the success or failure of e-learning because it has:
Tracked the changing attitudes about and
perceptions of e-learning by faculty and
technical staff over 18 months across a wide
sample of colleges and universities each with
substantial investments in e-learning.
•Mapped the changing supply of e-learning
providers and products.
Thwarted Innovation
makes sense of these
data by supplying a strategic story that explains
what happened to e-learning and why. As
Zemsky and Massy point out in their report: “
In retrospect, the rush to e-learning produced
more capacity than any rational analysis would
have said was needed. In a fundamental way,
the boom-bust cycle in e-learning stemmed from
an attempt to compress the process of
innovation itself. The entrepreneurs’
enthusiasm produced too many new ventures
pushing too many untested products—products
that, in their initial form, turned out not to
deliver as much value as promised. . . .The hard
fact is that e-learning took off before people
really knew how to use it.
Thwarted Innovation •••••••••••••••Page iii
Thwarted Innovation •••••••••••••••Page iv
Thwarted Innovation •••••••••••••••Page 1
Three innovations have dominated the educational arena over the last two
decades. The first is the development of high-stakes testing, in which
educational providers are held accountable for the performance of their
students on a host of national, standardized—and, in the case of the Trends in
International Mathematics and Science Study (TIMSS), internationally normed—
exams. The second innovation is the development of national, and occasionally
international, ranking systems designed and marketed to inform the public about
which institutions, firms, and programs represent the best providers of education.
The third major educational innovation—and
the only one of the three that actually focuses
on educational content—derives from the
linking of rapidly maturing information
technologies to a renewed interest in how,
when, and why people learn. Dubbed “e-
learning” and often linked to the dot-com
boom and the promise of e-commerce, e-
learning offered a truly student-centered
approach to education. It boasted the
potential of being design-rich, being capable
of delivery anywhere and at any time, and
being fully customizable to take full advantage of each individual student’s
personal learning style.
E-learning was also the innovation that garnered the most venture capital, the
most press, and, not surprisingly, the most grandiose promises. Among the claims
made to support e-learning investments, three are worthy of specific note. First
and probably foremost, the marriage of new electronic technologies and newly
accepted theories of learning promised to yield a revolution in pedagogy itself.
Thwarted Innovation •••••••••••••••Page 2
Learning would be customized, self-paced, and
problem-based. Course instructors would be
replaced by designers and facilitators—the “sage
on the stage” would become the “guide on the
side.” Students would have the ability to model
outcomes, conduct experiments based on well-
documented laboratory simulations, rapidly
exchange ideas with both fellow students and
teaching faculty, and, where appropriate, join
global learning communities not unlike the
contract bridge communities that have made
tournament bridge on the Internet an exercise in
international competition. Feedback on student
papers would be instantaneous—or nearly so.
Course materials would be rapidly distributed at
substantially lower costs than the antiquated,
bookstore-supplied text books and bulk packs.
Nor would the pedagogical revolution be
limited to either K-12 or higher education.
Corporate learning programs would be
transformed as well. Entirely new batteries of
skills-based learning sequences—covering
everything from introductory accounting to
advanced router maintenance and repair—would
be developed, along with accompanying
assessment and certification mechanisms. Just-
in-time learning would become the norm, with
employee-learners becoming more responsible for
amassing their own portfolio of skills. The
possibility even emerged that the boom-and-bust
cycle of corporate training that had traditionally
tracked the peaks and valleys of the business
cycle would have less impact on whether, how,
and why employees acquired new skills.
E-learning’s second promise derived from its
ability to be delivered any time and anywhere a
computer and connection to the Internet could be
found. Already, analysts were projecting a surge
in the demand for adult education, as more
people sought to start and finish baccalaureate
and post-baccalaureate programs, as well as to
acquire the new kinds of skills on which an
information economy depended. E-learning and
distance education would become synonymous
terms, as state agencies and private providers
brought new programs to market. Lifelong
learning would become an electronic reality.
E-learning’s third and perhaps most radical
promise was that the market would provide the
financing necessary for the industry to live up
to its potential. Funding would come first in the
form of substantial venture capital to launch
the panoply of products already in the offing
and then in the form of market revenues to fuel
to the necessary expansion. Predictions of e-
learning’s bounty literally knew no limits. The
most quoted projections—those made in 2000 by
Michael Moe in the Merrill Lynch white paper,
The Knowledge Web
—boldly proclaimed:
Our estimates for the U.S. online market
opportunity for knowledge enterprises will grow
from $9.4 billion in 1999 to $53.3 billion in
2003, representing a CAGR [Compound Annual
Growth Rate] of 54 percent.
At an estimated $105 billion, the spending
power of college students is huge. Not
surprisingly, a growing percentage of their
spending is moving online. Currently, students
Thwarted Innovation •••••••••••••••Page 3
spend $1.5 billion online, an amount which is
expected to almost triple to $3.9 billion by 2002.
We estimate that the U.S. market for online
higher education alone will grow from $1.2
billion in 1999 to $7 billion in 2003.
With that level of market
anticipation at hand, the rush was
on. Columbia University launched
Fathom. New York University nearly matched
those efforts with NYUonline. Cardean
University became the model of for-profit/non-
profit collaboration, in which some of the best
known American and European universities
partnered with UNext to launch a high-cost/high-
prestige model of business education. Individual
states made similar investments, choosing to
focus on providing low-cost, ready access to the
educational assets already available on publicly
funded university campuses. California’s brief
fling with its own electronic university and the
better known Western Governors University were
probably the two most widely recognized
examples, although efforts in Massachusetts,
Maryland, and Missouri in the end demonstrated
greater staying power.
Not surprisingly, perhaps, the reality never
matched the promise—not by a long shot. There
has been no pedagogical revolution, although there
has been a noticeable shift in corporate training
spurred in part by the economic downturn that
once again reduced training budgets and training
staff. Fathom and NYUonline are gone; Cardean
and UNext are in the process of their third or
fourth makeovers. While there has been a
burgeoning of distance education, the big success
stories owe more to their past market triumphs—
as in the case of both the University of Maryland’s
University College and the University of Phoenix—
than to any particularly imaginative melding of
learning and technology.
E-learning’s altered fortunes have occasioned
considerable comment. More often, e-learning is
now the butt of bad jokes—as in, “Can you
imagine telling your children to go to their
rooms and study college for four years?” In
general the cynics have had a field day, pointing
out that e-learning was just one more fad,
exhibiting more hype than substance, whose
demise proved to be little more than an echo of
the dot-coms’ bursting bubble.
However, to dismiss e-learning as just another
fad or, worse, a bad joke is to miss the point.
Understanding what happened to e-learning and
why is critical if we are to understand how and
why technologies are likely to affect educational
processes both now and in the future. What made
e-learning such an attractive investment to both
those who contributed sweat equity and those who
contributed venture capital? While all innovations
overestimate their promises, why were the claims
made on e-learning’s behalf so extravagantly off-
the-mark? Did e-learning simply flame out upon
takeoff? Or is it possible that, once the hoopla
has died down, e-learning will follow the same
trajectory as other innovations that first begin
with experimenters and pioneers, then expand to
a group of early adopters before becoming
commonplace and taken for granted? Given that
e-learning will be judged by its capacity to win a
Thwarted Innovation •••••••••••••••Page 4
place in an increasingly competitive higher
education market, how should one gauge the
likely size of e-learning’s share of that market—
both now and prospectively?
Where’s the Data?
It is to those who have asked these and
similar questions that
Thwarted Innovation: What
Happened to e-learning and Why
is addressed.
What we sought in this study was a conceptual
understanding of this phenomenon’s process of
change and innovation, on the one hand, and a
practical way of estimating e-learning’s current
and future trajectory on the other.
We first wrote about e-learning and the
rapidly changing world of information
technology in
Using Information Technologies to
Enhance Academic Productivity
, a 1994
EDUCOM monograph that emerged from a
Wingspread-sponsored roundtable. More
recently, Massy returned to this subject in
Honoring the Trust: Quality and Cost
Containment in Higher Education
, while Zemsky
began exploring key measurement issues as
part of
The Weatherstation Project
. This major
effort, funded by the Thomson Corporation in
partnership with the University of
Pennsylvania, sought to develop tools for
gauging how fast and in what direction the
market for e-learning was growing.
The Weatherstation
was intended as
an antidote to those first descriptions of the
market for e-learning, which were often warped
by missing data and overly hopeful assumptions
about how quickly new products would come to
market and how receptive learners and
instructors were likely to be. What we knew
when launching this project in the summer of
2001 was that facts were lacking. There had
been no tracking of students, products, or
purchases. No one knew how many students or
workers were taking e-learning courses in any
given year, nor how much either businesses or
colleges and universities had spent in pursuit of
e-learning initiatives, nor what students or
employees themselves had spent. Even less was
known about the structure of the market for e-
learning. How was it segmented? Who
constituted the major niche players? Equally
unknown was whether e-learning’s promised
economic efficiencies were allowing colleges and
universities, in particular, to recoup their initial,
often substantial investments in either hardware
or software—or whether the promise of new
enrollments on the part of remote learners was
proving sufficient to justify continued
investments in web-based distance education.
The educational data were nowhere to be
found. No standard category in the surveys
comprising the federal government’s annual
Integrated Postsecondary Education Data
System (IPEDS) asks institutions to report the
number of course credits they award online or
the number of transfer credits they grant for
online courses. No agency counts how many
online courses are offered as part of an
institution’s regular curriculum at either the
undergraduate or graduate/professional level.
No survey asks institutions to report how much
they are spending on their e-learning initiatives.
Similarly, there are no national sales figures
for e-learning software. One of the reasons
Thwarted Innovation •••••••••••••••Page 5
Michael Moe’s projections proved to be so
transitory was that they were based on market
surrogates that overestimated the actual dollar
transactions involved in the e-learning market.
The Knowledge Web’s
1999 figure of $1.2
billion spent on e-learning is an estimate that
includes monies spent on communications,
market aids, technical support, and
maintenance, as well as software, professional
training, and content creation. And the 2003
projected estimate of $7 billion is largely based
on what Moe and his colleagues knew about the
projected growth in computers, connectivity, and
the utilization of the Internet.
What’s the Concept?
In part, at least, data are lacking because e-
learning is still a concept in search of consistent
definition. Currently, three broad domains
define e-learning’s principal market niches:
1. e-learning as Distance Education. Mention
e-learning, and most people still assume the
reference relates to distance education or
education delivered on the Web. In fact, the
most successful forms of e-learning
courses delivered on the Internet—courses
that teach a particular subject; courses that
are part of a degree program most often at
the graduate or professional level; and,
finally, courses that offer certification in a
vocational or technical skill. For the most
part, however, what the Web provides are
merely correspondence courses distributed
2. e-learning as Facilitated Transactions
Software. E-learning’s second big triumph
has been in the development and expansion
of course management systems—BlackBoard
and WebCT are the best known—that both
organize courses and present materials
online. Principally used within higher
education, course management systems at
many institutions link teachers with students,
students with each other, and students to
sources. Schedules and assignments are
posted on the Web. Reading materials are
available for download, replacing the
proverbial “bulk packs” of an earlier
innovation. An important, growing subset of
this market involves computerized
assessments—principally the grading of tests.
3. e-learning as Electronically Mediated
Learning. The third category of e-learning—
and the one that initially attracted the
greatest attention—centered on the learning
materials themselves rather than their
distribution. This category includes a host of
products, services, and applications;
computerized test preparation courses (test
prep) that prepare students to take the SAT,
GRE, or any of a half-dozen standardized
tests; complex, integrated learning packages
such as Maple or Mathematica that teach
elementary calculus; course objects that
simulate everything from chemical reactions
to social interactions to musical
compositions; and tools like Macromedia’s
Dreamweaver and Flash
that help students
build their own websites and multimedia
Thwarted Innovation •••••••••••••••Page 6
presentations. This component of e-learning
includes the interactive CD-ROMs as well as
the websites that publishers of college
textbooks are increasingly making an
integrated aspect of their products. Despite
their seemingly diffuse nature, what all
these products and resources have in
common is that they involve electronically
mediated learning in a digital format that is
interactive but not necessarily remote.
Over the two years during which
Weatherstation Project
operated, we
proceeded along parallel tracks. We
understood that we would be unlikely to develop
workable tools for measuring the market for e-
learning unless we had a conceptual framework
defining what it was we should seek. The way
to test the conceptual framework was to see if
we could produce meaningful measurements
and, just as important, if we could understand
what the data were telling us. Could we develop
a plausible and realistic story explaining what
was happening to e-learning and why?
We believe that we have succeeded on all
three counts: that is, in offering a conceptually
concise way of understanding e-learning
principally as a market-driven innovation; in
providing a concrete measurement strategy for
tracking the e-learning market; and in evolving
a plausible storyline melding construct and
Thwarted Innovation: What Happened to
e-learning and Why
takes up each development
in order—beginning by specifying a conceptual
construct; proceeding to the development of
measurement instruments and an analysis of
the initial data those instruments provided; and
concluding with a narrative laying out a means
by which to gauge e-learning’s future trajectory.
In a sense, each of these sections can be read
separately, though we would urge a
consideration of the work as a whole.
Thwarted Innovation •••••••••••••••Page 7
The story of e-learning is fundamentally about what students of the
subject call “radical technological innovation. An innovation is judged to
be radical when the invading technology has the potential to deliver
dramatically better performance or lower costs in what previously had been a
stable industry. The operative word is
. When the new technology first
emerges, it often appears to be clumsy and inferior to its established predecessor.
In the beginning, it is the new technology’s promise rather than its performance
that attracts initial adherents. A large part of that promise is the vision of an
altered future—one that is not only different, but also dramatically better.
In the case of e-learning, the convergence of personal computers and
ubiquitous connectivity sparked a utopian vision in which teachers taught and
students learned in fundamentally different ways. Just over the horizon was a
world of active learners with teachers who guided and facilitated rather than
proclaimed and judged. Learning would be both continuous and exciting, while the
products of such learning would tangibly reward both learner and teacher. When
e-learning was first introduced, now more than 30 years ago as Computer-
Assisted Instruction (CAI), it was readily acknowledged that exploration of the
new technology’s future capacities had only just begun. While to the faithful the
potential was clear and present, few pretended to know exactly how “going
digital” would actually alter the day-to-day practices of professors.
Utterback and the Emergence of a Dominant Design
In actuality, much more is known about the dynamics of innovation,
particularly about what happens when a new technology enters the marketplace.
For one thing, the introduction of a radical new technology creates fluidity in both
markets and product designs. New entrants to the field bring novel design
concepts and target new market segments. Established firms field additional
Chapter 1:
The Dynamics of Innovation
Thwarted Innovation •••••••••••••••Page 8
innovations as they struggle to defend their
territory. MIT’s James Utterback, a leading
authority on technology-based innovation, points
out that, in the early days of a radical innovation,
“Market and . . . industry are in a fluid stage of
development. Everyone—producers and
customers—is learning as they move along.” But
the fluidity is not sustained. Ultimately, as
Utterback notes, in the case of a successful
innovation, “Within this rich mixture of
experimentation and competition some center of
gravity eventually forms in the shape of a
dominant product design
. Once the dominant
design emerges, the basis of competition changes
radically, and firms are put to tests that very few
will pass.” What emerges from this competitive
process is an innovation in a newly standardized
format that readily attracts new users.
The early days of automobiles were
characterized by just such a cycle. The number
of automobile manufacturers peaked at 75 in
1923, but then dropped to 35 in the late 1920s
and to 14 in 1960, even as the market
expanded. Creation of the dominant designs we
know today required a period of trial and error
in engineering laboratories and in the
marketplace. What gelled after 1923 was a
standardized conception of an effective
automobile: for example, one fueled by gasoline,
not steam; a self-starter, with four- to six-
passenger seating; and a vehicle with the all-
steel enclosed body introduced by Dodge in that
year. The pace of innovation, and the number
of innovating firms, slackened once this
dominant design emerged. Subsequent
competition turned to product refinement, cost
reduction, styling, and brand positioning. The
slower pace of innovation boosted the premium
on capital and market dominance, which in turn
produced a further shakeout in the industry.
The triumph of the automobile as the
world’s primary form of transportation
teaches a second lesson as well. The
dominant design may take a long while to
emerge, and it may involve changes not directly
related to the precipitating technology. For
example, the automobile’s dominant design did
not consolidate until a paved road system came
into being and gasoline became widely available.
When cast in these terms, the parallel
between automobiles as the key element in an
innovative transportation system and computers
as potentially the key element in an equally
transformed postsecondary learning system is
both instructive and prophetic. On and off
college campuses, e-learning could not take off
until wide-bandwidth Internet access was readily
available, until smart classrooms were
constructed, and until all faculty and students
had access to computers—investments that
students, universities, and most corporations
have been making and are continuing to make.
Still missing, however, are many of the key
elements of a dominant design. The avenues are
in place; still lacking is a standardized design for
the vehicles that the system will employ.
Put another way, a radical innovation for a
complex process such as e-learning requires
more profound changes than simply creating an
Thwarted Innovation •••••••••••••••Page 9
infrastructure: one’s very conception of the
supplying or consuming entity may have to
change. When the innovation relieves one
constraint, other constraints may well lurk close
by. As these limits are overcome, as the
innovation marches toward its dominant design,
attracting the intellectual and financial capital
necessary to establish a supportive
infrastructure, the innovation itself becomes
transformed—pushed in fewer directions, under
the direct influence of fewer innovators, but all
the while becoming more practical and hence
attractive to a growing number of new users.
Innovation’s S-Curve
E-learning’s pattern of innovation, change,
and adoption follows the classic S-curve shown in
Figure 1a. The curve has been shown to apply to
innovations as diverse as doctors’ adoption of
new drugs, farmers’ adoption of hybrid corn, the
railroad industry’s adoption of diesel engines,
and the emergence of the flat factory as
manufacturing’s dominant architecture.
Adoption processes usually start slowly because
of the need for experimentation. They accelerate
once the dominant design emerges, and then
eventually reach saturation. The actors at the
various stages of adoption differ markedly. For
example, innovators and early adopters are
driven by different motivations and play
different roles than the majority of users.
Researchers usually categorize and characterize
the actors in the following way:
, who represent the first few
percent of the eventual user population, seek
out and experiment with new ideas—often
driven by an intrinsic interest. They are the
pioneers and, like other pioneers, must
endure many trials and tribulations. Their
role is to determine how to use the new
product or service and demonstrate its
potential value.
early adopters
, roughly the next 15
percent of users, are moved to adopt once
the innovators have proven the
concept. They usually are
tightly connected to others in
the field and often are viewed
as opinion leaders. Early
adopters seldom consider
themselves to be pioneers, but
rather as hard-headed decision-
makers who pursue the
innovation for extrinsic rather
than intrinsic reasons. But
because they participate in the
fluid stage of adoption, before
Figure 1a. The Stages of Technology Adoption
Percent of
Early adopter
arly majority
Late majority
Thwarted Innovation •••••••••••••••Page 10
the dominant design has become established,
they shoulder substantial risk. One of the
early adopters’ principal contributions to the
emergence of a dominant design is their
success at finding alternative ways to exploit
the innovation and to test their alterations
under normal conditions of use.
early majority,
roughly the next third of
the population of eventual users, enters after
the dominant design is established. They
display less leadership than the early adopters
but are open to new ideas and tend to be well-
respected by their peers. They want to stay
ahead of the curve, and in so doing they drive
the first big wave of market expansion.
late majority
, the next third of the
population of eventual users, are people who
adopt after half the population has already
done so. They are followers, either due to
their conservatism or because their
attention was focused elsewhere during the
earlier adoption stages. Late-majority users
drive the next wave of market expansion,
which is characterized by intense
competition as the innovation matures.
, the last 15 percent or so,
resist adopting the innovation despite its
now-obvious advantages and the risk of
becoming isolated. In the end, of course, the
diehards die or retire from the field.
Innovation stages are usually described in
terms of demand, but the ideas apply to the
supply side as well. Innovating firms and the
individuals who launch those firms conceive
ideas and realize them in practice. Early
adopters may be individuals, but more likely
they are a part of the firms that bring the
innovation to scale and test design alternatives
in the marketplace. This role turns out to be
critical for radical innovations such as e-
learning. “Majority firms” expand the market
and move it toward maturity, while “diehards”
hold on by their teeth in declining markets. The
same firm, or its precursors or descendents,
may play all five roles at different times.
Market saturation occurs when the ranks of
potential adopters have been depleted. Further
growth may be limited to increases in the user
population, or the stage may be set for a new
breakthrough and a new adoption cycle. The
breakthrough may introduce the innovation to
new market segments, or it may represent new
applications in current segments. Either way, it
superimposes a new S-curve on the earlier model.
e-learning’s Adoption Cycles
On occasion, new and nearly simultaneous
waves of related innovations occur. The
overlapping of innovations’ adoption cycles
produces a complex situation that is more
difficult to analyze and predict, even though the
underlying dynamics follow the traditional S-
curve. Today’s applications of technology to on-
and off-campus teaching and learning present
this kind of complexity, in large part because
they have undergone four distinct adoption
cycles, as depicted in Figure 1b.
Each cycle represents a different stage of
innovation that also requires a different level of
change in the existing instructional culture. In
Thwarted Innovation •••••••••••••••Page 11
theory, each ought to
build upon the previous
adoption cycle and
smooth the way for the
next. In fact, however,
the cycles sometimes
proceed along generally
parallel tracks and at
other times may work
against each another. The cycles include:
1. Enhancements to traditional course/
program configurations,
inject new
materials into teaching and learning
processes without changing the basic mode
of instruction. Examples include e-mail,
student access to information on the
Internet, and the use of multimedia and
simple simulations. The typical application
uses off-the-shelf software, such as
PowerPoint, to enhance classroom
presentations and homework assignments.
2. Course Management Systems
which enable
professors and students to interact more
effectively. They provide better
communication with and among students,
quick access to course materials, and
support for administering and grading
examinations. A special subset of these
activities come bundled together to enable
the creation of true online courses and
learning networks.
3. Imported course objects
which enable
professors to embed a richer variety of
materials into their courses than is possible
with traditional “do it yourself” learning
devices. Examples range from compressed
video presentations to complex interactive
simulations. Online entities are springing up
to collect, refine, distribute, and support
electronic learning objects, and a few
institutions are experimenting with
enterprise-level Learning Content
Management Systems.
4. New course/program configurations
result when
faculty and their institutions re-
engineer teaching and learning activities to
take full and optimal advantage of the new
technology. The new configurations focus on
active learning and combine face-to-face,
virtual, synchronous, and asynchronous
interaction in novel ways. They also require
professors and students to accept new
roles—with each other and with the
technology and support staff.
The four levels of e-learning innovation are
currently in different stages of their adoption
Enhancements to traditional course/
program configurations
are moving rapidly
through the early majority stage.
management tools
are just now moving into the
early majority stage—not so much in terms of
Figure 1b. e-learning’s Adoption Cycles
1. Enhancements to traditional course configurations
. New course management tools
3. Imported learning objects
. Ne w course configurations
Stage of innovation
im e
Thwarted Innovation •••••••••••••••Page 12
the number of individual faculty and trainers
using them, but rather in terms of the
proportion of students and trainees who are
enrolled in courses and programs that deploy
course management software. These first two
adoption cycles have largely built upon and
reinforced one another. Their momentum,
however, has not transferred to either the
importation of learning objects
or to the
development of new course/program
. Both remain at the innovation
stage still in search of the kind of acceptance
that attracts early adopters.
The adoption cycles of off-campus and
distance education have followed the same basic
track: good use of the presentation
enhancement tools represented by PowerPoint;
heavy reliance on the kind of course
infrastructure that a good course management
system provides; computerized assessments;
and threaded discussions. At best, it would
include the importation and use of elementary
learning objects; in reality, it has prompted
almost no development of new course/program
configuration beyond taking advantage of the
Web’s capacity to promote self-paced and just-in-
time learning.
Framing Questions
Of course, the fundamental question is:
“Why do the innovations associated with e-
learning appear to have stalled out?” A more
nuanced inquiry would further ask: “What
effect did the widespread and rapid introduction
of teaching enhancements and course
management software have on subsequent
adoption cycles? Did either or both
inadvertently constrain the development of
course objects or new course/program
configurations? What role, if any, did e-
learning’s association with online and distance
education play in the reluctance of more
traditional on-campus programs to move much
beyond the deployment of course management
systems and the use of presentation tools like
PowerPoint? To what extent is there a set of
dominant designs that promotes the spread of
learning? And, to the extent there are no
dominant designs, does their absence help
explain e-learning’s thwarted innovation?
Finally, to what extent does e-learning’s
adoption of the market model embedded in e-
commerce and exhibited by the dot-com bubble
help to explain what happened?
Thwarted Innovation •••••••••••••••Page 13
When we began
Weatherstation Project
we did not seek answers to
those questions. Indeed, in the winter of 2001, we were not
prescient enough to know that they were the questions that needed
asking. Rather, we set out to develop a set of tools that would chart e-learning’s
forward progress as a major educational innovation. If we were not as sanguine
as Michael Moe and his colleagues about the coming size of the market for e-
learning, we were nonetheless convinced there would in fact be a market. We
believed it would include much more than course enhancements and course
management systems, and that it would become a large and therefore significant
component of the financial structure of postsecondary education in the United
States and elsewhere.
From Example to Projection
What we did understand, however, was that the measurement strategies then
being employed to estimate market demand for e-learning and e-learning products
were leading respected institutions and corporations to project and then invest in
what they believed would shortly become a multi-billion dollar market. The first
and initially dominant measurement strategy involved the collection of evidence
from early successes—for the most part, stories innovators like to tell one
another and anybody else who will listen. The most important analysis that
began in this way was Moe’s
Knowledge Web,
which collected as many of these
examples as possible and then, using a compounded surrogate measure,
extrapolated e-learning’s anticipated rate of growth.
The compound measure Moe and his colleagues chose—the anticipated increase
in computing as reflected in the sale of computers, the growth of connectivity, and
the utilization of the Internet—was not bad. Had e-learning already proceeded
Chapter 2: Plausible and
Implausible Measurement
Thwarted Innovation •••••••••••••••Page 14
beyond the early adopter stage, it could be
expected to grow at roughly the same rate as
other innovations dependent on computing
technology. The problem was that, in 1999, e-
learning had relatively few innovators and
almost no users who fit the classic description
of early adopters. PowerPoint had yet to begin
its steady advance across the educational
landscape. Most course management systems
were still being prototyped, while course objects
were primarily curiosities more to behold than
to use. The big successes—Maple in calculus
and Studio Physics—were more often cited as
special exceptions rather than precursors or
harbingers of things to come. Given this
nascent development, Moe found himself
lumping together all of e-learning’s early
manifestations in order to establish a baseline
for future projections. The net result was the
most widely quoted projection of e-learning’s
future track, which involved multiplying an
estimate of the rate at which computer usage in
general would likely grow by an estimate of the
monies then being spent on communications,
market aids, technical support, software,
professional training, and content creation.
What did Moe miss? The answer lies
in the nature of the innovations he
was trying to track and the fact
that the adoption cycles were not only
overlapping but at times competing. The coming
increases in the use of course enhancements
and course management systems could not be
summed and then used as a baseline for
estimating the growth in the importation of
course objects and the development of new
course/program configurations. What Moe
collected and then multiplied were wisps of
wind—a not unexpected compilation of hopes,
nascent innovations, and the sales pitches with
which experimenters and inventors have always
festooned their initial achievements.
Surveying the Terrain
Five years later, there is substantially more
data available with which to gauge e-learning’s
progress, though the strategies employed to
estimate and project future growth largely fall
victim to the same kind of pitfalls that Moe
confronted. Today the dominant measurement
strategy is the one-time survey asking
university administrators and the heads of
corporate training departments about their
current use of e-learning, broadly defined. The
most recent, best funded, and in many ways
most interesting as well as revealing of these
efforts is
Sizing the Opportunity: the Quality
and Extent of Online Education in the United
States, 2003 and 2003.
Sponsored by the Sloan
Foundation and conducted by the Sloan Center
for Online Education co-located at Babson
College and the Franklin W. Olin College of
Sizing the Opportunity
asks and
affirmatively answers, “Will students,
institutions, and faculty embrace online
education as a delivery method?” Just as
Sizing the Opportunity
found that the
“quality of online education [will] match that of
face-to-face instruction.
Thwarted Innovation •••••••••••••••Page 15
It is not these findings that concern us, but
the survey’s means of moving from the data
supplied by their respondents to conclusions of
optimism and hope. All such surveys share two
dominant characteristics. First they are
snapshots that report frequencies at a single
point in time. At the same time, getting
institutions to complete these surveys is a
major problem that almost always results in low
response rates. In the case of
Sizing the
the overall response rate was 32.8
percent. The question always remains: when
two out of three of those surveyed—in this case,
degree-granting postsecondary education
institutions—do not return the survey form,
what does their non-response tell us about the
subject being studied?
Sizing the Opportunity
also testifies to some of
the other enduring problems with broad-based,
one-time surveys designed to study either an
educational market or the spread of an
innovation, or, in this case, both. By design, the
survey was to be completed by the institution’s
chief academic officer. In fact, chief academic
officers seldom fill out surveys—almost always
designating the task to someone who reports to
them, with a request to collect, scrub, and finally
submit the answers. What the chief academic
officer does decide is whether his or her
institution will actually participate in the effort.
Having a major sponsor like the Sloan Foundation
underwriting the cost of the survey almost always
increases an institution’s willingness to
participate but seldom above the roughly 33
percent achieved by
Sizing the Opportunity
What all such surveys produce, then, is a
biased sample—an acceptable outcome, if the
analysts use that bias to estimate how the non-
respondents probably differed from those who
did respond. One scenario reflected in the
Sizing the Opportunity’s
response rate—the one
we actually think is most likely—reads as
follows. Provosts, deans, and academic vice-
presidents at institutions that had made
substantial investments in online and other
forms of e-learning were more likely to have
their institutions participate in the survey.
Institutions with formal programs of online
education and the larger technical staffs that
come with such programs were similarly likely
to have significantly higher response rates.
There was also likely a market effect.
Generally, medallion institutions and major AAU
research universities are disinclined to
participate, while institutions that serve the
middle of the market and that are often the
most eager to expand their student markets are
more willing. I. Elaine Allen, a principal author
Sizing the Opportunity,
recognized these
market dynamics when she told
of Higher Education
that faculty at private
baccalaureate institutions were the most
reluctant to participate in online education
programs: “They are,” she noted, “a very
entrenched bunch of objectors.” The
story went on to quote Allen as observing:
“There may be two groups emerging,
with two very different strategies for
moving forward.
Ms. Allen said public and for-profit
institutions—most of which already offer
at least some online courses—would
Thwarted Innovation •••••••••••••••Page 16
probably focus their energies primarily
on expanding and refining their use of
the Internet. But many private
institutions that have not adopted online
learning may steer clear of the
technology, because their faculty
members distrust teaching outside the
lecture hall.
What may largely account for
Sizing the
affirmative answers to the
question regarding whether students, faculty,
and institutions were willing to embrace online
education is simply the natural inclination of
one segment of the population to respond, and
the equally natural though different inclination
of an even larger proportion of the population
not to respond. What we can say is that
Weatherstation Project
’s survey results, using a
different methodology and theoretical
framework, are sufficiently different from those
reported in
Sizing the Opportunity
There is also the problem of timing and
focus associated with any survey.
Sizing the
rightly sought to distinguish
between degrees or levels of “onlineness.
courses were those with no online
content—though it is not clear whether such
courses could have an in-class component that
used learning objects or simply PowerPoint
demonstrations not on the Web. A
course was one with face-to-face
instruction as the dominant mode, but in which
web-based materials and systems like
Blackboard and WebCT were used to distribute
assignments and collect student work. A
course was one that used both
modes of instruction and typically included both
online discussions and face-to-face meetings.
Sometimes such courses are referred to as
“bricks and clicks.” Finally there are true
courses in which at least 80 percent of
the content is delivered online, and typically
there are no face-to-face meetings.
Sizing the Opportunity
did not focus
attention on was that element of e-learning that
is independent of the Web—largely our two
categories of imported course objects and
refigured courses/programs. In our terms,
Sizing the Opportunity
deals only with our first
two adoption cycles—course enhancements and
course management software with a primary
emphasis on the utilization of the Web to
promote relatively simple online courses.
Because surveys like
Sizing the Opportunity
are temporal events, they risk being rendered out
of date before their results can ever be published.
By mid-2003 it was clear that those responsible
for promoting e-learning initiatives on college
campuses were becoming increasingly concerned
about higher education’s changing budget
circumstances. Colleges and universities in
general and public institutions in particular have
a bad habit of cutting programs and initiatives
with strange soundings names whenever there is
a substantial reduction in funding. It is an open
question whether participants in
Sizing the
’s survey would respond today as they
had done earlier in the year—or whether they
would even bother to participate in the survey at
all. What is missing more generally is a sense of
timing and change, though
Sizing the Opportunity
did ask respondents whether they thought online
education would be more or less superior to face-
to-face learning three years hence. We wish they
Thwarted Innovation •••••••••••••••Page 17
had also asked their respondents whether they
were as optimistic about online learning’s promise
today as they had been three years prior.
The Measurement Challenge
The Knowledge Web
Sizing the
are important achievements,
pioneering explorations of a landscape
dominated by mists and misconceptions. Moe
and his colleagues established the notion that
analysts should be watching and calibrating the
market for e-learning and not just the frequency
of the reported use of e-learning and e-learning-
like experiments. Allen and her collaborators
offered an important corrective at a time when
much of higher education was prepared to write
off e-learning as a thankfully passing fad.
What these first major explorations missed
was the increasingly segmented nature of the
market for e-learning—that there were at least
four adoption cycles unfolding simultaneously
and often in conflict with one another. In their
focus on singular metrics—in Moe’s case on e-
learning’s anticipated compound annual growth
rate, and in Allen’s on frequency counts—they
largely ignored the changing content and the
evolving nature of the innovations themselves.
Thwarted Innovation •••••••••••••••Page 18
In designing and launching
Weatherstation Project,
our intention was to
focus on the dynamics of innovation and then collect data that we and others
could use to chart how the market for e-learning was changing over time—and
by extrapolation how it was likely to evolve in the future. We were struck from the
outset by a dangerous irony that had emerged: the sense of disappointment in the
fall of 2001 that was beginning to pervade the market for e-learning was as
misplaced as the euphoria that once led the industry’s optimists to celebrate an
invincible revolution. The fact of the matter is that, in the fall of 2003, e-learning is
alive and well. Money is being spent, smart classrooms are being built everywhere,
and collegiate faculty and corporate trainers are successfully integrating
electronically mediated learning into literally thousands of courses focusing on both
traditional and non-traditional subjects. That said, it is also the case that e-learning
is evolving in ways few predicted and with economic consequences that even its most
ardent supporters are still struggling to understand.
There is a larger lesson in the uncertainty, even confusion, that surrounds
the market for e-learning today: namely, if educational institutions and
corporations are to be serious about e-learning and its market, they require a data
collection strategy and set of measuring instruments that can track not so much
current usage and sales as the dynamic rhythms of e-learning’s competing adoption
cycles. That, at least, was our promise at the launching of the University of
Pennsylvania/Thomson Corporation
Weatherstation Project
and the Interview Process
Based on the premise that it was important to understand the characteristics of
e-learning’s emerging markets,
Weatherstation Project
began as a partnership
between a major research university and one of the nation’s leading suppliers of e-
learning and traditional print materials to the education market. The project’s
Chapter 3.
A New Measurement Strategy
Thwarted Innovation •••••••••••••••Page 19
measuring and tracking strategies are reflected
in its name. Given the absence of standard
institutional data reflecting e-learning usage or
supplier-provided data on e-learning sales,
initially established 12
observation posts (the metaphorical
in the project’s title): six on
college campuses and six within for-profit
corporations. On the six campuses at which we
, our intent was to
create three panels on each participating
campus to be comprised of 15 faculty, 15
administrators, and 15 students who would
agree to report quarterly on their attitudes
toward, expectations of, and uses of e-learning.
The process began with an interview,
either in person or via the telephone,
that explained the nature of the project
and asked panel members a set of standardized
questions about their own use of e-learning;
their sense of e-learning’s likely rate of growth;
its principal benefits; the forms of support it
was receiving on campus; the products and
services actually being used; and any new
developments or hot prospects they had spotted.
After their initial interview, respondents were
sent an e-mail asking them to visit an enclosed
URL to see how the project team had coded their
answers to the interview’s questions. Using the
interactive features of the website, respondents
were able to change their answers to reflect
their current experiences with e-learning.
Each quarter thereafter, respondents were
sent follow-up e-mails, again with a customized
URL, asking them to check previous answers
and tell us, via the website, how their attitudes
and experiences had changed. We also
regularly reported to respondents the project’s
preliminary findings. We recognize that such
reporting ran the risk of influencing their
subsequent answers to the
quarterly probes, but
also considered the risk to be acceptable, given
that the reporting seemed to increase the
likelihood that panelists would continue to
respond to our e-mail probes.
In sum, then, the measurement strategy
embedded in our use of
resembled that of the Nielsen Ratings, which
track TV viewing through a sample of
households. In
Weatherstation Project,
sample of institutions reflected both the
experimental nature of the project—just six
—and our desire to
have as broad a mix of institutions as possible.
All of the participating institutions had
reputations for deployed well-developed
strategies for the use of learning technologies
and chief information officers on whom we
could rely to help us recruit and motivate
survey respondents. We are under no illusions
about the biased nature of our sample: it
reflects institutions we knew in advance were
investing, often substantially, in e-learning. We
also know that the respondents themselves
were neither a random nor a representative
sample of their administrative or faculty
colleagues. The members of the faculty, for the
most part, were early adopters; none could be
called diehards. The administrators recruited
Thwarted Innovation •••••••••••••••Page 20
for the project’s
panels were largely mid- to
upper-level technical staff responsible for
supporting faculty in their experiments with and
expanded use of learning technologies. The
reader is cautioned to keep the nature of our
sample in mind when considering responses to
particular questions and events—we certainly did.
In one important respect, our sample of
institutions is representative of the larger
population of degree-granting institutions of
postsecondary education in the United States.
Among the six are a community college, a public
comprehensive university, a public land grant
university; a major public research university; a
private liberal arts college, and a major private
research university. Four of these institutions
serve the Name-Brand/Medallion segment of the
market for undergraduate education, one serves
the Core market, and one serves the User-
Friendly/Convenience market.
Two False Starts
A roughly similar strategy was to be
employed at the six corporate
though only a single individual (usually a chief
training officer) was to respond quarterly to
inquiries. However, the rapid
slide into recession that coincided with the
launching of the project played havoc with this
measurement strategy. At first, the
participating trainers proved reluctant to report
on the degree to which their use of e-learning
was declining as their budgets were being
reduced and consolidated. Then, the training
officers with whom we had built relationships
began to disappear—the victims of corporate
reorganizations and downsizings.
Ultimately we abandoned our attempt to
track the corporate market for e-learning using
model and turned instead to a
series of indirect, web-based measures which we
describe in Chapter 5.
We also abandoned our attempt to establish
student panels. No matter what we tried, we
could not achieve consistent participation by a
group of students over an extended period of
time. In fact, we often failed to gather a
sufficient number of students on a participating
campus to allow us even to conduct the initial
interviews. We are not alone in having failed to
take account of student opinion and experience—
a lacuna that has warped most estimates of the
campus-based demand for e-learning.
The interview and Web-probe protocols used
Weatherstation Project
were designed
using two principal criteria. First and foremost,
we wanted to track change over time. We
thought that frequencies mattered much less
than the slope of change. Accordingly, most of
the questions we asked elicited answers that
could and did change over time—in both the test
and the actual use of the protocols.
Second, we wanted the content of the
questions to track closely with the conceptual
framework we were developing in order to chart
both the adoption and adaptation of e-learning
strategies, products, and services. We needed
questions that would allow us to track faculty
Thwarted Innovation •••••••••••••••Page 21
responses, in particular, in terms of e-learning’s
principal market niches:
Distance education;
•Facilitated transactions; and
Mediated learning and its four adoption
+Enhancements to traditional course/
program configurations
+Course management tools
+Importation of learning objects
+Development of new course/program
Faculty respondents were asked 17 yes/no
questions, largely tracking their use of specific
e-learning tools; administrators were asked 4
yes/no questions, primarily related to their
support for e-learning. Both faculty and
administrators were asked 14 questions asking
them to rate a particular attribute in terms of a
high/medium/low score. An example of the
first type of question asked respondents
whether they used multimedia presentations
(yes/no). An example of the second type of
question asked respondents to gauge the future
growth rate for e-learning as high, medium, or
low. Many of these latter questions focused on
the respondent’s perception of e-learning’s
importance in terms of institutional priority;
availability of technical staff and other
resources to support the development of e-
learning; benefits that derived from e-learning;
and student acceptance of e-learning as a
substitute for face-to-face instruction.
Appendices 1 and 2 to this report provide the
full set of questions asked of faculty and
administrative respondents.
We spent the first year of the project
building the interactive website, writing and
testing the interview protocol and subsequent
follow-up probes, recruiting the participating
campuses, and establishing the requisite panels
of faculty and administrative respondents. In
all, each respondent was contacted four times
over the course of 15 months—the initial
interview, the chance to correct/change how
that interview was entered into the project’s
database, and twice with follow-up probes via e-
mail. From the outset, we established two tests
that had to be met: respondents would have to
continue to respond; and they would have to
take seriously our admonition to think carefully
about how their experience with e-learning was
and was not changing over time. The overall
response rate from probe to probe exceeded 80
percent on all six campuses—most of the
missing respondents were either faculty on
leave or technical staff who had either left the
institution or had taken jobs with other
All of the available evidence suggests they
did change their responses in accord with
changes in their experiences with e-learning.
Indeed it is the presence and distribution of
those changes that tell the real story of e-
learning on college campuses. It is the story to
which we turn in the next chapter.
Thwarted Innovation •••••••••••••••Page 22
By design,
Weatherstation Project
was an experiment in
measurement. What we ultimately sought was a reliable strategy for
charting the evolution of the market for e-learning. The campus
were our principal innovation—and largest success. Although
we tracked campus experiences for only 15 months, involved only six
campuses, and had to abandon our efforts to track student experiences, what
we can report is that the strategy works. Once established, the faculty and
administrative panels proved to be stable, engaged, and interested in the
outcomes. Response rates were uniformly high. Respondents took
care to report where they wanted––and, just as importantly, where they did not
want––to change their responses. We can also report that the set of questions
and probes successfully captured the range of experiences that respondents
were having as they experimented with e-learning. The only question we now
wish we had asked was a specific query about their use of PowerPoint.
One of the byproducts of our testing of the campus
was a set of first findings. We wanted to be sure that we could extract from the
data a set of strategic stories telling us what was happening to e-learning on
these six college campuses. Again, we were aware of—and we caution the
reader to remember—just how small our sample is and therefore how tentative
we must be in drawing conclusions. The analysis that follows is therefore
meant to be illustrative of the power of a data collection strategy such as the
one employed by
Weatherstation Project.
Tracking e-learning’s Four Adoption Cycles
We begin with basic frequencies of faculty use for the principal elements of
the four often intertwined and almost always overlapping adoption cycles
associated with e-learning. Five of the questions asked faculty whether they
Chapter 4: First Findings
Thwarted Innovation •••••••••••••••Page 23
required one or more of
the following e-learning
enhancements to
traditional course formats:
student use of web-
based materials;
purchase of a textbook
with a CD-ROM or
access to a proprietary
website managed by
the textbook’s
student use or
purchase of “off-the-
shelf” software
packages; and
student participation in
e-mail discussions.
As Figure 4a indicates,
in at least one of their
courses, almost all faculty
members in the sample
required students to use
web-based materials; three
out of four used
multimedia presentations,
principally PowerPoint; 60
percent assigned textbooks with e-learning
supplements; nearly the same percentage
assigned off-the-shelf software packages; and just
over half required their students to participate in
e-mail discussions.
Figure 4a. e-learning Course Enhancements:
Faculty Respondents’ Frequency of Use
Figure 4b. e-learning Transaction Systems:
Faculty Respondents’ Frequency of Use
0% 25% 50% 75% 100%
Used computer-based
Used course management
Two campus
questions focused
on faculty use of two core elements of an e-
learning transaction system: course management
tools and computerized assessments. Nearly 80
percent of the faculty respondents reported
having used a course management tool,
principally Blackboard or WebCT. Just over half
0% 25% 50% 75% 100%
e-discussion required for
Off-the-shelf software used
Textbook with
disc/proprietary website
Multimedia presentations
Required students to use
web-based materials
Thwarted Innovation •••••••••••••••Page 24
said that they had used computerized
assessments (Figure 4b).
Finally, faculty respondents were asked
whether they had developed an e-learning
course object or a comprehensive e-learning
course. Seventy percent reported they had
developed a course object, while 44 percent
reported they had developed a comprehensive e-
learning course (Figure 4c). It is the answers
to these questions that demonstrate just how
skewed the
sample is. Most
faculty respondents were in fact early adopters
of e-learning—not innovators and experimenters
per se
, but rather early pioneers intrigued with
e-learning’s potential and hence willing to be
among the first to serve as institutional guinea
pigs. In other words, what we have in the
sample is the innovation’s
leading edge, at least on the six campuses
participating in the project
Accordingly, if you want to know where e-
learning is heading, watching the leading edge
proves to be a useful strategy. From this
perspective, what becomes most important is
the direction or slope of change that occurs over
time. Figures 4d through 4f report the degree
of shift in respondents’ reported usage of e-
learning elements over the course of the
project’s 15 months of operation. Given the
pace at which faculty ordinarily change their
teaching patterns, it is not surprising that there
was little change, although it is interesting to
note the growth of e-discussions (Figure 4d).
More surprising was the reported growth of
8 percent in the use of course management
systems and the even larger growth of 11
percent in the number of faculty who reported
using computerized assessments (Figure 4e).
Remember, we are charting the experiences of
faculty who are at the leading edge of e-learning
utilization. The fact that better than one in ten
of these early adopters over the course of a
single year began using computerized
assessments becomes an important marker for
charting an emerging market niche.
There was also a steady but slow increase in
the rate at which these faculty invested their
own time and effort in the development of
Figure 4c. Major Investments of Time:
Frequency of Faculty Responses
0% 25% 50% 75% 100%
Developed comprehensive
e-learning course
Developed e-learning
course objects
Thwarted Innovation •••••••••••••••Page 25
learning objects (e-learning’s basic building
blocks) and the development of comprehensive
e-learning courses (Figure 4f). From the more
open-ended aspects of our initial interviews with
faculty, we learned that a substantial number
had received institutional support to develop
these programs—in the form of technical staff,
development funds, reduced teaching loads, and/
or summer salaries. By the time the
panels were in place, the
Figure 4d. e-learning Course Enhancements:
Shifts in Faculty Respondents’ Frequency of Use
-15% -10% -5% 0% 5% 10% 15%
Used computer-based
Used course management
Figure 4e. e-learning Transaction Systems:
Faculty Respondents’ Shifts in Frequency of Use
yp q y
-15% -10% -5% 0% 5% 10% 15%
Multimedia presentations used
Textbook with disc/proprietary
website required
Off-the-shelf software used
Required students to use web-
based materials
e-discussion required for
Thwarted Innovation •••••••••••••••Page 26
percentage of faculty reporting that they could
expect a reduction in their standard workloads
to support e-learning development had dropped
to 13.2 percent. Over the course of the next 15
months, that share would be halved again to
reach 6.6 percent. There was a similar decline
in the reported level of department support for
developing e-learning components.
Attitudes and Expectations
The campus
made it equally
possible to track the changing attitudes and
expectations of respondents. We asked all
respondents, administrative staff as well as
faculty, to gauge the overall benefits associated
with e-learning—high, medium/moderate, low—
and then to estimate e-learning’s potential for
achieving economic efficiencies and opening up
new student markets. Finally, we asked the
respondents to estimate a future growth rate
for e-learning, broadly defined.
Not surprisingly, this faculty group of early
adopters thought e-learning offered substantial
(moderate to high) benefits, although their
estimates of its potential for realizing
efficiencies and opening student markets was
still in the medium/moderate range (Figure 4g).
It is, however, the shifts of opinion over 15
months that tell the more interesting tale.
Overall, there was a slight erosion in the
estimates of e-learning’s growth; a modest
increase in the respondents’ estimates of e-
learning’s capacity to serve new markets; and a
modest-to-substantial increase in their estimates
of e-learning’s potential for achieving efficiency
gains (Figure 4h). There was almost no shift in
the overall benefit associated with e-learning,
suggesting that faculty respondents were
primarily in the process of redefining the focus
of those benefits.
Because the
tracked how individual participants responded
in each follow-up to all of the questions in the
protocol, we could also track how often they
changed their evaluation of e-learning. Figure
4i displays those changes for our four “benefit
-15% -10% -5% 0% 5% 10% 15%
Developed e-learning
course objects
Developed comprehensive
e-learning course
Figure 4f. Major Investments of Time:
Shifts in Frequency of Faculty Responses
Thwarted Innovation •••••••••••••••Page 27
questions.” Note
that, while only a
slight negative
shift in the
collective estimate
of e-learning’s
likely growth rate
is apparent, in
fact nearly one in
four respondents
had changed his
or her mind. The
overall measure
shifted only
slightly, because
nearly as many
thought increased
growth was likely
as thought decline
would occur. On
that issue, then,
there was
volatility.On the
question of e-
potential to yield
efficiency gains,
there were also
one in four respondents changed their
answers—but those changes were more clearly
unidirectional. For every respondent who
predicted that e-learning’s potential for
Figure 4g. e-learning’s Value Components:
Faculty Respondents’ Score
Figure 4h. e-learning’s Value Components:
Faculty Respondents’ Shifts
e-learning's current rate of
Potential for more efficient use of
Capacity of e-learning to serve
new student markets
Benefits of e-learning
-25% 0% 25%
e-learning's current rate of
Capacity to serve new student
Potential for more efficient use
of resources
Benefits of e-learning
Thwarted Innovation •••••••••••••••Page 28
efficiency gains would decline, five thought the
potential would increase. The
seemed to report that, in the months ahead,
those responsible for explaining and defending
e-learning—the innovation’s early adopters—
would again be on the lookout for the economic
efficiencies e-learning had, in the past, so often
promised. The volatility surrounding the
anticipated growth rate for e-learning can be
interpreted in two ways: a major re-evaluation
was underway which would lead either to more
optimistic forecasts or, conversely, to a further
diminishing of e-learning’s prospects.
An Interpretative Frame
With this last example in place, we can now
summarize the interpretative range of the
project’s measurement strategy. In the first
instance, the panels provided a way to gauge a
frequency of use (or, the overall interpretive
score) for e-learning, reflecting the collective
experience of the early adopters who made up the
panels on each campus. These
results are not substantially different from those
reported by most surveys, including the one
Sizing the Opportunity
. The principal
difference in our study is that we know we are
reporting the experiences and opinions of e-
learning’s early adopters—that is, the innovation’s
leading edge rather than its lagging center.
The response shift measures capture the
direction and to some degree the momentum of
change that is underway in the market. The
volatility measures are an indication of likely
changes to come—in some cases, allowing the
interpreter to predict the direction of change (as
in the case of the search for economic
efficiencies) and in other cases, serving as an
Figure 4i. The Volatility of e-learning’s Value Components:
Percentage of Faculty Respondents Changing Their Estimates
0% 10% 20% 30%
e-learning's current rate of
Capacity to serve new student
Potential for more efficient use
of resources
Benefits of e-learning
Thwarted Innovation •••••••••••••••Page 29
alert that changes, probably
unpredictable ones, are in process (as
in the case of the changing estimates
of e-learning’s likely growth rate).
Yet another way to view this
interpretative frame is to note that the
frequency and score measures
principally represent what happened
yesterday; the shift measures call out
what is happening now; and the
volatility measures identify the areas
and sometimes the direction of future
change (Figure 4j).
Shifting Institutional Priorities
We can now use this three-part framework to
ask how institutional priorities were changing
across 2002 and 2003 on the six campuses
participating in
Weatherstation Project
. Three
questions focused on priorities. To what extent is
e-learning an
institutional priority?
To what extent is it a
budgetary priority?
To what extent was
departmental support
(or school support,
for the liberal arts
college in the sample)
available for faculty
interested in
developing e-learning
for the classroom or
the Web?
Figure 4k reports answers for these three
questions. The vertical scale indicates the
average score (high, medium/moderate, low)
associated with each question, as answered in
the spring of 2003. Interestingly, among early
adopters there was a clear sense that e-
learning’s institutional priority was higher than
Figure 4j. Interpretive Frame for
Campus Weatherstation Data
Figure 4k. e-learning as a Priority:
Score, Shift, and Volatility in Faculty Responses
-25% 0% 25%
Value Score
Degree of Shift
MODERATE High Volatility
High Volatility
Moderate Volatility
Thwarted Innovation •••••••••••••••Page 30
its budgetary priority. We believe that
respondents felt institutional leaders would
provide support by encouraging faculty to
experiment more readily than they would
commit the institution’s discretionary funds to
the cause. Respondents also reported that
departments provided more support to e-
learning than line items in their budgets,
although they offered less verbal support than
their institutional leaders. Here, we suspect, the
departments found themselves caught in the
middle—wanting to be helpful without having
the necessary resources to follow through.
Figure 4k also suggests that, overall, e-
learning’s early adopters were becoming more
pessimistic—for all three questions, respondents
reported a shift toward a more negative stance.
Again, it may be the volatility measures that
provide the most meaningful indicators. The
two questions that most directly affect faculty
in the classroom—budget and departmental
support—also occasioned substantial volatility.
The designation of “high volatility” signals that
more than 20 percent of faculty respondents
had changed their mind over the course of the
year. But the volatility involved switches in
both directions. On the question of budget
priority, almost as many faculty respondents
reported an increase as reported a decrease.
The responses to the institutional priority and
departmental support questions were more
unidirectional. In the case of the former, there
were nearly three negative shifts to one positive
shift; and on the departmental support question,
the ratio was two-to-one in favor of a more
negative estimate of e-learning’s growth rate.
Faculty vs. Administrative
The administrative staff that comprised the
other half of the campus
were asked the same questions tracking
attitudes, expectations, and judgments regarding
e-learning’s place within their institutions. For
the most part, the responses of administrative
staff, who are responsible for providing
technical support to their institutions’ e-learning
initiatives, matched those of the faculty’s early
adopters. In one important aspect, however, the
administrative responses were different. On all
but one question, administrative volatility was
substantially higher than faculty volatility.
Remember that volatility is different than
opinion shift—in the case of the latter, there is a
pronounced direction to the changes respondents
made from their original responses. In the case
of volatility, it is a matter of change without a
specific direction—with some respondents
becoming more negative and a nearly equal
number becoming more positive. What Figure 4l
indicates is an element of growing uncertainty on
behalf of technical staff. They were less certain
that they knew e-learning’s future growth rate,
less certain regarding whether or not e-learning
could promote the more efficient use of resources,
and even less certain about the availability of the
workshops which, for the most part, they were
responsible for offering. On 9 of the 15 questions,
more than one out of five administrative staff
Thwarted Innovation •••••••••••••••Page 31
changed their answers over the course of the 15
months tracked by the project. On four of the
questions the degree of change exceeded 25
percent. By comparison, faculty responses
seemed almost ploddingly stable.
Making Sense of a Mosaic
This increased sense of volatility among
administrative staff offers a clue about the
evolution of e-learning on these six campuses
between 2002 and 2003. The tracking data
suggest first and foremost that the chill of budget
reductions was settling over e-learning. There
was a growing perception that e-learning’s
priority within institutional budgets was
declining. There was a noticeable increase in the
economic benefits of e-learning that respondents
hoped would come from achieving greater
efficiencies and expanding the student market—
principally, we gathered, by offering more online
course to remote learners. What everybody
sensed was that e-learning programs were
increasingly going to have to pay for themselves.
The response of faculty early adopters to
these changing circumstances was to pull back,
noting in the process that e-learning was less an
institutional as well as budgetary priority. It
was less likely to receive direct support from
their departments, and less likely to provide the
extra incentives—release time, summer support,
travel funds—that had been important in
persuading them to invest their discretionary
time developing e-learning courses and course
objects 15 months prior.
The administrative staff simply vibrated—
their jobs were on the line. When we talked to
Figure 4l. Faculty vs. Administrator Response Volatility
0% 15% 30%
e-learning's current rate of growth
Potential for more efficient use of resources
Availability of workshops
Faculty overload as deterrent to e-learning
Anticipated student discomfort
Budget priority
Intellectual property rights a faculty concern
Support for faculty developing e-learning
Campus store as a source of e-learning software
Capacity to serve new markets
Institutional priority
Faculty familiarity with e-learning
Frequency of own use of e-learning
Benefits of e-learning
Percent Changing Responses over 15 Months
Thwarted Innovation •••••••••••••••Page 32
the administrative staff about what was
troubling them, we were met with two types of
answers. The first was that looming budget cuts
would undo all the good work they had been
able to accomplish over the last five years. The
second was that they were going to be left
holding the bag—expected to continue to support
faculty and student efforts at a time when
resources were being withdrawn rather than
being added to their programs.
Their questions were simple and to the
point. Who was going to make Blackboard and
WebCt work on their campuses now that the
administration was touting how many courses
had this online component? Who would train
the faculty just beginning to experiment with
tools beyond PowerPoint? And from where
would the energy as well as resources come to
introduce and then integrate the new products
that inevitably would attract the attention and
enthusiasm of faculty early adopters?
If administrative staff needed confirmation
of just how much at risk they and their
programs were, they needed only to consult the
daily e-mail report published by
of Higher Education.
For as long as any of us
could remember, those briefings had included a
section on Information Technology, featuring a
variety of items tracking the exigencies of
technology issues—including e-learning—on the
American college campus. The content was
there on Friday, October 17, 2003—and was
gone the following Monday.
The Chronicle
put a
good face on it, suggesting that technology had
become so ubiquitous that it no longer needed a
separate section in the daily briefing.
Ubiquitous or not, after October 17,
became a much less interesting read
if one’s focus was information and educational
technology. No doubt about it—the world of e-
learning was changing and not necessarily for
the better.
Thwarted Innovation •••••••••••••••Page 33
The reduction and then demise of our corporate
forced us
to adopt an alternate strategy for tracking the general market for e-
learning. Unable to query the customers of e-learning, we instead shifted
our attention to the providers—those who sold their wares to corporations, to
organizations other than corporations, and to both entities.
Our methodology was remarkably simple and straightforward. The first
step was to build a master list of providers—specifically, those providers with
websites. The next step was to classify each provider in terms of a set of
standard characteristics: market segment, business focus, specialization, and
product range. The
team completed the classifications, with two
members visiting the websites of each candidate separately and independently
Results were then compared; where necessary, the websites were revisited, and a
judgment about their classification ultimately rendered. The team
also randomly
revisited the websites, spending additional time with selected providers to ensure
that the classification scheme was capturing the right information. In all, 262
providers of e-learning products and services were identified and classified. (The
full classification scheme and results are presented in Appendices 4 through 7.)
The Shape of the Provider Market
The classification process began by determining to which market segment or
segments the provider marketed its products and services:
t, and/or
direct to the consumer,
where the consumer is assumed to be
an individual. Providers could offer products and services in more than one
market segment—and, in fact, most do. Roughly one-third of providers market to
a single segment, one-third to two segments, and one-third to three or more
Chapter 5: The Corporate
Market for e-learning
Thwarted Innovation •••••••••••••••Page 34
The segment in
which most
providers sell their
products or services
consists of
businesses, both
small and large.
Only 1 of every 5
providers offers
products and
services that they
believe will have
minimum or no
appeal to businesses.
This corporate
market for e-
learning also
contained the largest
group of providers
(20 percent)
concentrating on a
single market
segment (Figure 5a).
Half of the e-
learning providers
we tracked offered
some products and
services designed
at least in part to appeal to educational and
academic customers—principally schools,
colleges, and universities. But just 10 percent
of these providers specialized in the academic
market segment. While a third of the market’s
e-learning providers sought to serve
governmental agencies, none could afford to
specialize in that domain. Finally a third of the
market supplied products and services directly
to individual consumers (Figure 5b).
The various combinations of market segments
served by e-learning providers confirm how the
market for business-related products and
services dominates. It is, as one observer noted,
Figure 5a. The Provider Market for e-Learning
Corporate Academic Government Direct-to-Consumer
Percentage of Providers Marketing to each Principal Segment
Academic Only
Plus Direct-to-
Consumer Only
9% 9%
Percentage of Providers in each Market Segment
C = Corporate
= Academic
G = Government
D = Direct-to-Consumer
Figure 5b. Detail on Provider Market Segments
Thwarted Innovation •••••••••••••••Page 35
the only place there is money––although, given
the state of the economy, not very much is likely
to be earmarked for e-learning!
By the winter of 2003, the market for e-
learning had also transformed itself from one
that initially focused on content into one
increasingly dominated by providers with a
greater or equal focus on services, including
consulting help, hosting, and
the design and management
expertise needed to produce
customized e-learning
programs. By February, the
number of providers
exclusively marketing services
was almost twice the number of
those exclusively providing
content—and the number of
providers who sought to do
both accounted for 70 percent
of the provider population
(Figure 5c).
A second way to
characterize e-learning products
is to identify those that were
primarily designed to appeal to
corporate customers. Among
providers selling to businesses,
three specific product lines
were dominant: information
technology, customization, and
consulting. More than half of
the providers of e-learning to
businesses offered products
focusing on information
technology; among providers whose principal
appeal was to non-corporate customers, just 21
percent offered products and services focusing on
information technology. There were similar tilts
for customization—the promise of designing and/
or delivering a program customized for an
organization’s need—and for consultation. In the
case of customization, 47 percent of providers
Providers who Serve only Non-Corporate
Providers who Serve Corporate Customers
Percentage of Providers Supplying each Type of Product
Information Technolo
IT ++Consulting Customization
Figure 5d. Corporate e-learning’s Principal Foci
Content Content and Service Service
Percentage of Providers with each Product Focus
Figure 5c. Provider’s Product Focus
Thwarted Innovation •••••••••••••••Page 36
serving businesses offered to customize their
offerings, versus just 9 percent for providers
specializing in non-corporate products. In the
case of consulting, the split is 41 percent versus
7 percent (Figure 5d).
So dominant were
these patterns that we
came to define the trilogy
of information technology,
consulting, and
customization as
signature of the corporate
e-learning market. All but
16 percent of providers
serving businesses offered
at least one of these
specialties; a third offered
two of the three, and 17
percent offered all three.
All told, more than half of
the providers serving the
corporate market provided
two or more of these
specialties (Figure 5e).
The range of
characteristics we used
to describe and classify
providers of e-learning
make possible a rich
variety of analyses
detailing specific market
niches. For example,
Figures 5f through 5h
focus on the 20 percent
of providers offering
assessment tools—
exams, certifications, test writing, and test
preparation. As shown in Figure 5f, most of
these enterprises offered tools and products
associated with learning objects and
None Just One Two of Three All Three
Percentage of Providers to Corporate Clients at each Level
Figure 5e. The Corporate e-Learning Signature:
Combinations of IT, Consulting, and Customization
10% 11%
Degrees Credits Courses Certificates Learning
Percentage of Assessment Tool Providers
Who also Provide Learning Products
Figure 5f. Learning Products Sold by Providers of
Assessment Tools
Thwarted Innovation •••••••••••••••Page 37
certificates, while
relatively few offered
products associated with
formal degree programs
or college credits.
Many providers of
assessment tools (31
percent) offered training to
their clients; relatively few
provided hosting services
(Figure 5g).
Finally, providers of
assessment tools were
most likely to be found
serving the health
care industry.
surprisingly, just
a third of
assessment tool
providers offered
products related
to the information
industry (Figure
Tracking the
We developed market classifications for e-
learning providers to help us track changes in
the market itself. Our goal was a parallel
measurement strategy—again using frequencies,
shifts, and volatility to complement our
strategy for tracking the
Figure 5h. Business Focus by Providers of Assessment Tools
IT Communications Sales Health Care
Percentage of Assessment Tool Providers
by Business Categories
Figure 5g. Services Sold by Providers of Assessment Tools
22% 23%
Hosting/ASP Cours e
Customize Consulting Training
Percentage of Assessment Tool Providers
Who also Provide Services
collegiate e-learning market. What we needed
was a set of variables or characteristics whose
change over time would help us gauge the
rhythms of the market.
Thwarted Innovation •••••••••••••••Page 38
We came to focus on the Web itself, asking
how often and to what extent our selected
providers were changing their websites. The
assumption was that the majority of these
changes would reflect decisions by the provider
to offer new products and services or to revamp
their current presentations to better appeal to
the market. On a weekly basis, the
submitted the URLs from
our master list of providers to a BullsEye search
engine, which was configured to indicate which
sites had changed in the last week. The BullsEye
probes counted a variety of changes: new pages,
new HMTL code, new graphics, even automated
date changes. In the latter case we had to
assume that such alterations were randomly
distributed. Again, we were interested in the
distribution of change rather than the absolute
number of changes. Within this framework, we
were able to count the changes for any category
of providers, calculate an average, and then
compare that average with the previous week’s.
The results were interesting without being
definitive. Given the novelty of the measure
itself—average weekly website changes—we had
to guess at what would count as a significant
level of change. After some experimentation, we
adopted the following rule of thumb. Where the
average number of changes for a given week for
all providers in that market segment was one or
two per week, we said that market segment or
niche was in the “white zone.” Averages of more
than two changes per week earned a market
segment or niche the label “red zone.” We also
noted when a segment or niche seemed to be
balanced between both zones.
The measure worked, in the sense that it could
identify market changes and rhythms. Figures 5i
through 5k display this market tracker for
January 2003. In Figure 5i, only the Direct-to-
Consumer market segment is in the red zone,
though the
Corporate Plus,
and Academic
Plus segments
are on the
One week
later, as
captured in
Figure 5j, most
of the balloons
had ascended;
by the close of
the month,
Corporate Academic Direct-to-
Academic Plus Direct-to-
Consumer Plus
Market Segment
Average Changes per Week
Figure 5i. Activity Tracker, 1/13/03
Thwarted Innovation •••••••••••••••Page 39
shown in Figure
5k, only the
Corporate and
segments had been
left behind in the
white zone. Flip
quickly through
the three figures,
and the market’s
animation becomes
volatility in all but
two segments.
This market
tracker can display
data on any subset
of the market—
niche, product line,
consulting versus
service offerings,
specialty, or
business focus. It is
also possible to
expand the
comparing not
weeks (as above)
but rather months (as below, in Figure 5l). The
measure is still “Average Weekly Web Changes,
but the comparison is made across the last
quarter of 2002. Two of the market segments—
Corporate plus Government, and Corporate plus
Direct-to-Consumer—have experienced dramatic
swings, shifting downward from average changes
per week in excess of 5.5 in October 2002 to
less than 2.5 in January 2003. The remaining
market segments displayed in Figure 5l were
relatively stable.
In addition to tracking average weekly
changes in each provider’s website, we also
calculated the proportion of any given set of
providers who changed their websites in a
particular week. Figure 5m provides a graphic
Corporate Academic Direct-to-
Academic Plus Direct-to-
Consumer Plus
Market Segment
Average Changes per Week
Figure 5j. Web Activity Tracker, 1/20/03
Corporate Academic Direct-to-
Academic Plus Di re ct-to-
Consumer Plus
Market Segment
Average Changes per Week
Figure 5k. Web Activity Tracker, 1/27/03
Thwarted Innovation •••••••••••••••Page 40
of the
for e-
in each
changed their websites during the last week of
September 2002. Much of the core of the
corporate market remained stable, with just over
one quarter of the providers changing their
websites. Two sets of providers, however,
showed signs of atrophy—the set of providers
attempting to provide products to all four
segments (Corporate, Academic, Government, and
Direct-to-Consumer) and the set of providers
seeking to bridge the difference between Corporate
clients, Academic customers, and individuals
marketed to Directly. Two other segments seemed
in the throes of frenzied activity, the C plus D
group and the C plus A plus G group. What isn’t
clear is whether such activity signals market
growth or market churning.
Googling the Market
To derive a second measure of market
change, the
team turned to
Google, the predominant Internet search engine.
Using Google’s advanced search features, each
week the team
would input “e-learning” plus a
specific product category—for example,
“education” or “business and investing” or
“humanities.” For each category, for each week,
the probe produced a total number of “pages”
that then became the entry in the
database. On a weekly basis,
that database made two calculations for each
entry—the category’s share of the total number
of “pages” for that week; and the week-to-week
change in the category’s number of pages.
The Google tracker operated from February
2002 through May 2003, similar to the
operation of the campus
for a
total of 15 months. From the data produced by
the campus
, we could deduce a
strategic story of changing attitudes and
expectations being shaped, on the one hand, by
the budget chill creeping across higher education
and, on the other, by e-learning’s failure to
Figure 5l. Shifts in Average Weekly Web Changes by Providers:
October 2002 to January 2003
-4 -3 -2 -1 01234
Corporate Only
Thwarted Innovation •••••••••••••••Page 41
promote a fundamental pedagogical change in
the classroom. The data from the Google tracker,
as in the case of the data from the BullsEye
probes, largely reflected more of the same for
the corporate market—not currents or directions,
but rather ripples in a pond that was being
drained by an economic recession centered in
manufacturing and technology.
The overall shape of the e-learning market
as reflected in the Google probe underscores just
how much of that market is centered on
corporate America. Business and Investing,
Technology, and Computing and the Internet
account for 55 percent of the activity.
Education garnered 10 percent, Science and
Mathematics just 2.7 percent, and the
Humanities and Social Sciences just a trace (less
than three-tenths of one percent each). Figure
5n displays those distributions for two points in
time: February 2002 and May 2003. Perhaps
the most important point is to note how little
changed—most of
the graphs
balloons sit on
top of one
another for both
points in time.
The big loser
was the Education
category. The
winners were
Technology and
Government, Law,
and Politics.
Again, caution is needed in interpreting these
results. The growth in Technology pages from
providers offering e-learning products and
services is not necessarily a sign of growth in
the size of that market segment. The more likely
interpretation is that, given the downturn in the
fortunes of companies in the technology
business, the e-learning providers of technology-
related products and services were expanding
their search for new customers—that is, the
more likely explanation is market churning. One
should use the same lens for interpreting the
results displayed in Figure 5m.
The Google probe makes possible a display
that can be best likened to the output of an EKG
(Figure 5o). Again, the basic conclusion is one of
relatively constancy. Just three major peaks and
one trough appear over the course of 15
months—and, each time, there was a reversion
to the mean. The first (A on Figure 5o) was a
spike in the summer of 2002 in e-learning Web
Figure 5m. Percentage of Providers in Each Segment
Changing Their Websites
Thwarted Innovation •••••••••••••••Page 42
activity connected
to Technology and
to Computing &
Internet. However,
this spike was
immediately by a
trough led by the
same two
categories (B). The
second peak
occurred late in the
fall of 2002 (C)
and involved the
four categories of
principal interest to colleges and universities: the
Arts, Science and Mathematics, the Humanities,
and the Social Sciences. That peak also subsided,
leaving these four categories collectively at the
same 5 percent level with which they began the
tracking period. Finally, at the tail end of the
tracking period, there was a third peak (D) led by
Government, Law, and Politics, Computing &
Internet, and Business & Investing.
Do We Have a Market Tracker?
We cannot conclude, as we did in the case of
the campus
, that our market
trackers worked, nor can we claim to know what
was happening to corporate e-learning or its
providers. It is possible that our market trackers
simply missed substantial trends. It is also
possible that the nature and severity of the
recession produced a dramatically contracting
market for e-learning—one that our trackers
missed as well.
But is also just as plausible that the market
trackers worked—and that what they have to tell
both providers and consumers of e-learning
represents an important insight into the future
shape of the emerging market for e-learning:
that is, this market is dominated now and will
likely to continue to be dominated by providers
who offer products to businesses, both large and
small. Currently, to succeed in this market,
providers must offer a host of services, including
consulting and customizing learning products.
Our reading of the data suggests that over the
next several years providing services will prove
to be more profitable than supplying content.
The other well-defined, seemingly successful,
and ostensibly stable market niche is comprised
of firms and educational enterprises that sell
directly to individual customers—the distance
education niche. Their products are not very
soph-isticated, but their attention to detail and
Figure 5n. Percentage of Google Pages by Learning Subject:
May 2002 vs. February 2003
Internet Products
Red = May 2003
Blue = February 2002
Thwarted Innovation •••••••••••••••Page 43
to customer interests is becoming a hallmark of
their successes.
The other identifiable market niche with
“legs” is comprised of firms offering
computerized assessments—tests, exams for
licensing agencies, test-prep, and remote access
to standardized testing protocols like the SAT,
GRE, GMAT, and TESOL. The evolving nature of
this niche parallels the growing attractiveness
of computerized assessments on college
campuses—though it still is not clear just how
often either faculty members or their
institutions will have the financial wherewithal
to purchase these products.
The educational segment—once thought to be
among the market’s leaders—is actually getting
smaller. Though many providers advertise
learning objects, there is little evidence that they
are much in demand. Instead, the market
remains focused on bread-and-butter
applications—in Business and Investing,
Technology, and Computing & Internet. One gets
the sense when reviewing the products being
offered that “innovation is out and survival is in.
From this perspective, then, the general
market for e-learning looks very much like the
market tracked by our campus
not very expansive; dominated by the suppliers
of transaction systems and consulting; and still
waiting for the innovation to take hold.
The obvious question, then, is why hasn’t e-
learning taken off? Why are there relatively
few successful innovations? Why doesn’t
content matter more? Why should the market’s
educational segment be declining rather than
growing? We set out in the next chapter to
provide answers to these questions.
Figure 5o. Weekly Changes in Number of Google Pages:
February 2002-May 2003 (smoothed)
Business & Investing
Computing & Interne
Government, Law & P
Health & Medicine
Products & Services
Science & Mathemati
Social Sciences
Thwarted Innovation •••••••••••••••Page 44
Perhaps the most productive way to decipher what happened to e-
learning—and, in the process, to answer the questions we posed at the
end of the last chapter––is to examine the three basic assumptions
that defined its promise, as well as why those assumptions proved to be
particularly troubling:
1. If we build it, they will come.
2. The kids will take to e-learning like ducks to water.
3. E-learning will force a change in how we teach.
A fourth assumption, related more to the potential for e-learning to build
bridges across learning communities, could be added to this list: electronically
mediated learning would lead rapidly to the development of international
networks linking both scholars and learners.
Assumption 1: “If we build it, they will come.”
As with most innovations, those responsible for the experimentation that
yields an initial product simply assume that “If we build it, they will come”—
that their customers will recognize the value of their product as soon as it
emerges on the market. Almost all of e-learning’s first applications began in
precisely that way, as individual experiments whose interesting results led e-
learning’s first innovators to believe that they would attract the attention of
other experimenters and eventually the interest of the practice community. Not
surprisingly, then, most descriptions of both the spread and the potential of e-
learning derive either from catalogs of interesting experiments or from
collections of successful applications.
The best catalog tracking the rise—and, on occasion, the fall—of e-learning
experiments is Carol Twigg’s
The Learning MarketSpace,
which she describes as
“A quarterly electronic newsletter . . . highlighting ongoing examples of
Chapter 6: e-learning’s
Troubling Assumptions
Thwarted Innovation •••••••••••••••Page 45
redesigned learning environments using
technology and examining issues related to their
development and implementation.” Because
Learning MarketSpace
funds as well as reports
on experiments using e-learning in American
collegiate classrooms, its electronic pages
provide a unique glimpse of the growing
sophistication of available strategies and
programs. Much of the content focuses on the
development of course or learning objects—the
principal building blocks of any program
offering electronically mediated instruction,
whether on the Internet or through some other
form of electronic distribution.
The best collection of course or learning
objects has been assembled by MERLOT, an
acronym that stands for Multimedia Educational
Resource for Learning and Online Teaching.
What MERLOT wanted to become was a readily
available, low-cost, web-based service to which
individual experimenters could post their
learning objects and from which interested
practitioners could download objects to use in
their courses. A key component of the original
design was to develop a user community whose
members would regularly rate and evaluate the
quality and usability of the learning objects
available through MERLOT. While the latter
goal proved elusive in practice, MERLOT
nonetheless became a unique repository that
The Weatherstation Project
to track the
changing composition of e-learning’s user
community as well as the shifting emphases of
e-learning’s subject matter.
From June 2001 to January 2003, the
visited the MERLOT
website on a bi-monthly basis. MERLOT itself is
a marvel of careful documentation and reliable
programming—features that allowed us to ask a
series of critical questions: Who were
MERLOT’s members? Which fields of study
were best represented? Which disciplinary
communities? How fast was MERLOT both
growing and changing?
The answers to these questions echoed those
we had received from our
panels. Over the course of 15 months of
tracking, the number of MERLOT’s registered
members grew steadily at the rate of 2.5
percent per month. From June 2001 through
January 2003, MERLOT’s registered members
nearly tripled, growing from just over 4,500 to
just over 11,600. Faculty members were the
largest group, growing from more than 2,700 to
more than 8,000 (Figure 6a).
The growth was impressive; however, the fact
that MERLOT’s registered faculty numbered less
than 10,000 out of more than 1,000,000 total
teaching faculty in the U.S. (of whom roughly
half were full-time faculty) meant that MERLOT’s
total market penetration amounted to less than
one percent. Like the members of our own
panels and respondents to the
Sizing the Opportunity
survey, MERLOT
primarily tapped the opinions and interests of e-
learning’s innovators and early adopters.
Tracking MERLOT helps to document the
degree to which the most complex of e-
learning’s adoption cycles—the one focusing on
Thwarted Innovation •••••••••••••••Page 46
learning objects––has yet to take off. In general,
the learning objects posted to MERLOT are not
becoming more sophisticated; and, while the
number of MERLOT’s visitors and members
continues to grow, collectively they represent
but a small portion of e-learning’s potential
adopters. Users
continue to share
what they have
produced themselves
without exhibiting
much interest in
rating or evaluating
what others are
offering. There is no
feedback loop, no
evident connection
between the suppliers
and consumers of
learning objects.
Indeed, if one follows MERLOT’s postings as we
did, one comes away with the feeling that there
really are no e-learning consumers at all—only
innovators and inventors eager to showcase
what they have accomplished.
Just as important, tracking MERLOT
suggests that the
distribution of e-
learning’s early
adopters has
constant over the
last two years.
Course materials
posted to the
continue to be
dominated by just
two fields:
Business, and
Figure 6a. MERLOT Members
June 2001 September December March 2002 Ju ly October January 2003
Number of Members
Figure 6b. MERLOT Course Materials by Field
June 2001 September December March 2002 July October January 2003
Percentage of Materials Online in MERLOT
Science and
Social Science
Thwarted Innovation •••••••••••••••Page 47
Science and
Together, these
fields account for
nearly 60 percent
of all the learning
objects available
through MERLOT.
The principal shift
in the number of
learning objects
posted over these
15 months was
also largely
between these two
categories, with Business growing at the
expense of Science and Engineering (Figure 6b).
Tracking MERLOT’s disciplinary trends
suggests that this shift was largely occasioned
by a decline in the physics’ community
domination of—and perhaps interest in––
MERLOT in particular and course objects in
general (Figure 6c).
Inspecting the actual learning objects posted
to MERLOT reveals a second important aspect of
e-learning’s trajectory: there has yet to emerge
any sense of a dominant design in course
objects—the kind of dominant design that is
almost universally characteristic of successful
innovations. In the realm of technology there
are at least three dominant designs that can be
cited as examples. The first is the evolution of
spreadsheet software—beginning with VisiCalc,
proceeding through Lotus-1-2-3, and ending with
Microsoft’s Excel. Different products, different
internal designs, but all adhering to the basic
concepts of a spreadsheet consisting of rows and
columns. The second example of a dominant
design is the emergence of the Apple-pioneered
use of the graphical user interface—a dominant
design that every developer of user-friendly
systems now employs as a matter of course. The
third is the kind of sophisticated web-crawler
Google pioneered, which ultimately provided the
service itself a dominant market position.
Within the realm of e-learning in general,
two dominant designs have emerged.
PowerPoint now supplies the dominant design
for course enhancement materials—that is, for
e-learning’s first adoption cycle. For e-
learning’s second adoption cycle focusing on
transactions, Blackboard and WebCT course
management systems supply the dominant
design. But in the realm of learning objects,
anything goes. The range of modalities remains
June 2001 September December March 2002 July October January 2003
Percentage of MERLOT Materials On-line
Health Science
Information Technology
Teacher Education
World Languages
Teaching Well Online
Community for Academic Tech Staff
Figure 6c. MERLOT Course Materials by Discipline Community
Thwarted Innovation •••••••••••••••Page 48
so broad as to be wholly confusing. There is
still no sense that if “I know how to use one
learning object I basically know how to use all
or most learning objects in my field.” But that
is precisely what most e-learning users want,
largely because they know that the interfaces of
most of the software applications they use have
achieved that kind of transparency through the
application of a dominant design.
Carol Twigg in the most recent issue of
Learning MarketSpace
offers an important
summation of what
The Weatherstation Project
has now documented. Wistfully listing her
comments under the header “Build It, But Will
They Come?” she writes about MERLOT and
MIT’s OpenCourseWare Project:
This approach has several drawbacks.
Entries are selected and mounted by
interested individuals, but the materials
are not tied to improved student learning
outcomes. Many of the included learning
objects are intended for specific (and
possibly unique) upper division courses
that are not necessarily part of the
curricula at other institutions. Other
materials are designed for sophisticated
students and may not be relevant to a
more diverse student body at other
institutions. In addition, these projects
tend to assume that more options are
always better. MERLOT cites “links to
thousands of learning materials” as one
of its benefits, yet only a tiny subset has
been evaluated by anyone other than the
contributors. Most importantly, these
projects lack a methodology for transfer
to other institutions. Their strategy of
hope-for-the-best has been tried many
times in the past and failed (e.g.,
programs supported by Apple and IBM in
the 1980’s and 1990’s, and attempts by
national organizations like Educom).
Twigg, C. (July 2003)
The Learning
What Twigg refers to as a “hope-for-the-best
strategy” of transfer and dissemination is a
good description of e-learning’s current
predicament—and an explanation of why this
innovation’s champions have built a field of
dreams that, for the most part, has proven to be
attractive only to themselves.
Assumption 2: “The kids will take
to e-learning like ducks to water.”
Two years ago, most faculty or staff
members within a university community would
have been nearly unanimous in their assessment
of whether students would be able to utilize
computer-based learning—as part of a course
either on the Internet or in a classroom using an
electronic course management system or
learning objects. Indeed, they would be
incredulous that you made such an inquiry.
interviewers posed this
question in the fall of 2001, they were regularly
told: “Not a problem—the kids take to e-learning
like ducks take to water. After all, they love
games and technology, are dismissive of
professors who seem to have trouble navigating
Blackboard, and think that PowerPoint is state of
the art.
When asked, however, how comfortable
students would be if, for a particular course or
program, e-learning were substituted for in-class
instruction, the members of
campus panels were less sure. Eighteen months
ago, just over half of the administrative staff
surveyed—for the most part administrators with
responsibility for supporting faculty in their
Thwarted Innovation •••••••••••••••Page 49
role as teachers—said students would have little
or no trouble if e-learning was substituted for
in-class instruction. One-third of the group said
students would have some, but not a great deal,
of trouble; and just 15 percent said most
students would likely have a lot of trouble. A
year later the distribution of opinion among
administrative staff in the
panels was roughly the same: 46 percent said
there would be no problem; 41 percent said
most students would have some but not a lot of
trouble substituting e-learning for in-class
instruction; and 11 percent said most students
would have difficulty.
The similarity of the two distributions,
however, obscures the fact that one of every
four administrators in the panels changed their
opinion over the course of a single year—with
15 percent saying they now believed students
would have more trouble, and another 10
percent saying that students would actually
have less trouble. What is important to note
here is the volatility of the responses. Among
administrators, only the questions about e-
learning’s market position and institutional
priority generated a greater degree of change
over the course of a year.
Faculty responses generally mirrored those
of their administrative colleagues, though in
more muted tones. When first asked if they
thought most students would have trouble
substituting e-learning for in-class instruction,
the faculty members who were part of the
panel broke nearly into
thirds: 37 percent said students would have
little or no trouble; 32 percent said most
students would have some, but not a lot of
trouble; and 31 percent said most students
could have a lot of trouble with the substitution.
As with their administrative colleagues, faculty
opinion on this issue was noticeably volatile.
How many faculty changed their mind over the
course of the year? The answer is nearly one
in five, although again the overall distribution of
opinions remained roughly the same.
In the spring of 2003, the
team visited three of the campuses that had
participated in the project: Foothill College in
California, Hamilton College in New York, and
the University of Texas-Austin. In sessions with
panel members, the team asked why such
volatility in opinion was evident on the issue of
whether students would have difficulty in
substituting e-learning for in-class instruction.
The answers reflected a growing appreciation of
the fact that initial assumptions about e-
learning were being modified by actual
experience—along with a sense that no one had
ever asked the students whether or not they
Several weeks after the team’s visit to
Austin, there appeared in the
Daily Texan
opinion piece by one of the University of Texas’
senior honor students. Her column is worth
quoting in some detail, not because it proves in
and of itself that students are becoming
distrustful of what she called “teaching
technology,” but because it gives voice and
language to those doubts.
Thwarted Innovation •••••••••••••••Page 50
The fairy tale of e-learning assumes that
classroom technology enhances the
learning experience for both the
professor and the students. The reality of
such educational technology is far from
ideal. Often poorly integrated into a
course, its use skews the balance of
content and technology and lessens
dynamic interaction among students and
between students and faculty. . . .
The use of teaching technology can
quickly transform into a pedagogical
crutch. In an upper-division linguistics
course last fall, the daily lecture
consisted of no more than a PowerPoint
presentation and printed handouts of the
same display. This un-innovative
approach reduces the role of the teacher
to a mere conduit that transmits ideas
into student depositories.
Particularly troubling are the choices
of lower-division language classes to
implement technology that might allow
for a greater quantity of students but
lessens the quality of the education. . . .
A prime example of the increasing
pervasiveness of classroom technology is
the electronic textbook. The e-book
makes technology the primary
educational tool, even though many
students seem to prefer to use
technology as a secondary source.
Consider the case of Management 320F
last fall when the chosen text was
electronic. Professor Victor Arnold
initially ordered enough print copies of
the textbook for less than a quarter of
the class. Students could buy a download
version of the e-book or purchase a
password that would allow a page to be
viewed a maximum of four times. Yet
one-third of the class opposed the e-book
and lobbied for more print copies to be
Isensee, L. (January 28, 2003)
The Daily
, University of Texas-Austin
The University of Texas also provided an
important clue as to why the students’ interest
in games and their quick adoption of most
computer-based technologies did not translate
into an interest in e-learning. One of the senior
managers of the University CO-OP, the
university’s megabookstore, told the
team to check out “the kind of
software the kids were buying.The team did,
conferring with the bookstores on each of the
campuses participating in
and then turning to
Chronicle of
Higher Education’s
monthly tracking of the
“Best-Selling Software at College Bookstores.
The results were fascinating. In June
2003, for example, basic Microsoft
products accounted for five of the ten
best-sellers. Number seven on the best-seller
list was the leading anti-virus software, Norton,
reflecting the heightened concern over a raft of
viruses and worms then infecting machines
worldwide. The remaining four? In order, they
are: Adobe Photoshop, Adobe Acrobat,
Macromedia Studio MX, and Macromedia
Dreamweaver MX. Photoshop is used for
editing, enhancing, and optimizing photographs.
Acrobat allows the reader to read and prepare
PDF files. Dreamweaver allows the user to
construct sophisticated websites. And
Macromedia Studio MX, to quote the product’s
own website, “provides professional
functionality for every aspect of Web
development and includes the newest versions
of Dreamweaver, Flash, Fireworks and
FreeHand.” What this last set of software
products has most in common is the capacity to
allow users to prepare and distribute complex
presentations. Or, as the manager of the Texas
CO-OP reminded the
team, this
software is principally about showing off.
Thwarted Innovation •••••••••••••••Page 51
The implication, borne out in subsequent
interviews, is that student fascination with
computers and software has three major
components. They want to be connected,
principally to one another. They want to be
entertained, principally by games, music, and
movies. And they want to present themselves
and their work. As most faculty in the U.S.
have learned, students have become almost
obsessively adroit at “souping-up” their papers,
which they submit electronically and which they
festoon with charts, animations, and pictures.
As one frustrated professor who had just spent
a half-hour downloading a student’s term paper
was heard to remark,All I wanted was a
simple 20-page paper—what I got looks
suspiciously like the outline for a TV show.
Most promoters of e-learning simply missed
all of this devotion on the part of students to
complex presentations of self. The students
they saw in their mind’s eye were gamers who
would love simulations, who would see in the
computer a tool for problem-solving, who would
take to e-learning like ducks take to water. And,
in fact, there are some students just like that,
though, for the most part, they are concentrated
in engineering schools. The most successful e-
learning experiment was Studio Physics
developed by Jack Wilson, then at the Rensselaer
Polytechnic Institute (RPI). Studio Physics is
taught wholly on the computer in specially
designed “studios” where students work in two-
person teams on upwards of 25 computers.
Faculty circulate throughout the studio, providing
help and instruction as needed, as each student
pair works through a complex set of problems
and computer simulations designed to teach the
basics of introductory physics.
The program worked at RPI—and at more
than a dozen other institutions—because the
curriculum itself was problem-based, because
simple graphics could be used to simulate
physical properties and rates of change, and
because the students themselves saw Studio
Physics as an example of the kind of system
they had come to this engineering school to
learn to develop. Yet, this set of characteristics
is hard to match for other curricula. It is also
important to point out that Studio Physics
remained a group activity. The students came
to class, and they worked directly with their
partners and the faculty assigned to the Studio.
No one was isolated—no one was off in a room
by him- or herself with just a computer and a
set of e-learning exercises.
The importance of an actual, physically
intact learning community can be demonstrated
in another way. Three of the universities
participating in
Weatherstation Project
launched extensive programs of distributive
instruction that used web-based e-learning
modules as the principal means of instruction.
By intention and design they were to be
outreach programs capable of enrolling part-
time adult learners who were distant from
campus. What each of these universities
discovered, however, was that better than 80
percent of those enrolling in the e-learning
courses were full-time students living on
campus. Some apparently took these e-learning
Thwarted Innovation •••••••••••••••Page 52
courses because they were interested in or
curious about computer-based instruction. Most
students, however, enrolled in these e-learning
courses because they were “convenient.
Because they were on campus, the e-learning
experience was neither remote nor detached,
but simply there.
Assumption 3: “E-learning will
force a change in how we teach.”
One of the more hopeful assumptions
guiding the push for e-learning was the belief
that the use of electronic technologies would
force a change in how university students are
taught. Only bureaucratic processes have
proven to be more immutable to fundamental
change than the basic production function of
higher education. Most faculty today teach as
they were taught—that is, they stand in the
front of a classroom providing lectures intended
to supply the basic knowledge students need.
Those who envision a changed, more responsive
learning environment have argued that the most
effective instructor is not the “sage on the
stage,” but rather the “guide on the side.
Learning, they have argued, works best when it
is participatory. Students can become effective
problem-solvers only when they have mastered
the art of critical thinking and have acquired
the discipline necessary to be self-paced
learners. Constant assessment and feedback
are critical, so that both student and instructor
can determine, before it is too late, whether the
student is mastering the necessary material.
E-learning seemed more than ready to
satisfy each of these goals. As Studio Physics at
RPI demonstrated, within fully integrated e-
learning courses faculty are in fact guides—and
designers and mentors and conveners. They are
not presenters, unless they happen to have
filmed themselves performing an experiment or
conducting a simulation and then made those
images available on their students’ computers.
The student pairs represented exactly the kind
of interactive learning groups that educational
reformers envisioned. The feedback was
immediate and continuous. Students knew if
they had the right answer or were at least
proceeding in the right direction as soon as they
submitted answers to the problem sets on which
they were working. What the designers of
Studio Physics also learned is that there could
be no hidden assumptions—no relying on one’s
intuition or past experience to know when and
how to introduce new topics. For the first time
many of the faculty involved in Studio Physics
had to spell out their teaching strategy as well
as think through what kinds of learning
strategies their students were likely to bring
into the Studio.
Alas, Studio Physics is the exception,
not the rule. For the most part,
faculty who make e-learning a part
of their teaching do so by having the electronics
simplify tasks, not by fundamentally changing
how the subject is taught. Lecture notes are
readily translated into PowerPoint
presentations. Course management tools like
Thwarted Innovation •••••••••••••••Page 53
Blackboard and WebCT are used to distribute
course materials, grades, and assignments—but
the course materials are simply scanned bulk
packs and the assignments neither look nor feel
different. Even when the text book comes with
an interactive CD-ROM or when the publisher
makes the same material available on a
proprietary website, most faculty do not assign
those materials. Only modest breakthroughs
have occurred—in the use of e-mail to
communicate rapidly and directly with students
and in the adoption of computerized testing
materials, many of which provide a more
robust, but still static, means of evaluation.
A number of people are coming to believe
that the rapid introduction of course
management tools have actually reduced e-
learning’s impact on the way most faculty
teach. Blackboard and WebCT make it almost
too easy for faculty to transfer their standard
teaching materials to the Web. While
Blackboard’s promotional materials talk about
enabling faculty to use a host of new
applications, what the software promises up-
front is less dramatic: the ability for them “to
manage their own Internet-based file space on a
central system and to collect, share, discover
and manage important materials from articles
and research papers to presentations and
multimedia files.All faculty really need are
the rudimentary electronic library skills that
most have already mastered. Blackboard and
WebCT allow the faculty users to respond, when
asked,Are you involved in e-learning?” by
saying, “Yes, my courses are already online!”
The rapid introduction of PowerPoint as e-
learning’s principal course enhancement tells
much the same story. PowerPoint is essentially
“clip art” e-learning—in the sense that it allows
the instructor to import graphics and graphs
from other mediums, including the instructor’s
old lecture notes. Illustrated lectures do not
constitute electronically mediated learning any
more than courses that use Blackboard or
WebCT to distribute learning materials without
introducing learning objects.
Even the most adventurous and committed
faculty members often approach the use of e-
learning in ways that lessen its general impact
on the curriculum. On each of the campuses
participating in
Weatherstation Project,
faculty were initially recruited to experiment
with e-learning, supported by technical support,
summer salaries, and the ability to make their
e-learning course on any subject of interest to
them. With this level of support, most of the
courses were well-designed, technically
sophisticated, and, given the faculty members’
freedom to teach what they wanted,
idiosyncratic. Once the course had been offered
for two or three years, the faculty member
often moved on to other topics and different
experiments, having satisfied his or her own
interests and curiosities. Then the courses
died—simply because no one wanted to teach
someone else’s e-learning syllabus. What these
universities began to discover is that they
constantly had to make extra incentives
available to faculty in order to involve them in
e-learning. When the expenditures of those
Thwarted Innovation •••••••••••••••Page 54
funds became too expensive, the institutions
dropped the incentive programs and witnessed a
general flattening of e-learning adoptions and
experiments. All but forgotten, by then, was the
idea that e-learning might lead to a more
general reformation of both teaching and
learning styles.
A Fourth Assumption
More hope and anticipation than
assumption, the belief that was held by many of
e-learning’s early proponents was that
electronically mediated learning would lead
rapidly to the development of international
networks linking both scholars and learners.
On the scholarly side, many of those networks
now exist, leading to lively exchanges, shared
research, and cooperative investigations. On the
e-learning side, however, the big news at any
moment concerns what is about to happen
rather than what has actually been
What is better understood now is that most
e-learning takes place within national borders
and contexts, reinforcing the fact that place
remains of paramount importance. Little is
actually known in one country about the e-
learning capacities of other nations unless those
products are advertised on the Web in English.
Over the last two years, Professor Motohisa
Kaneko of Tokyo University and his colleagues,
principally Naoki Ottawa of Todai and Fujie
Yuan at the National Institute of Multimedia
Education (NIME), have employed probes to
analyze Japanese e-learning websites that are
similar to those used by
. Two conclusions are evident. First,
Japanese web-based e-learning is in its infancy,
and the products remain both limited in variety
and rudimentary in style and design. At the
same time, the Japanese Web probes make clear
that what has market appeal in Japan can be of
little interest to the American market. For
example, one of the largest product categories
among the Japanese websites is language
instruction and acquisition—a subject that is
simply not present on U.S. e-learning websites.
When e-learning products begin to penetrate the
market, they usually do so by appealing to
immediate, often very local, needs. Eventually,
no doubt, there can be a merging of interests
and products. In the beginning, however, it is
differentiation and specialization along lines
defined by national cultures and local
proclivities that matter most.
There are two important exceptions to this
generalization. The first involves tests and
examinations that students require if they seek
admission to an American or international
university, principally the SAT and TOEFL.
Prometric and its Japanese affiliate R-Prometric
do have internationally configured networks
spawned by the need to ensure the fair and
efficient administration of these exams. But
Prometric—and similar electronic-based testing
organizations—serve rather than link their
customers. To the extent that there is a
network, it is of providers rather than learners.
The second exception is the development of
a variety of high-cost, high-prestige programs of
Thwarted Innovation •••••••••••••••Page 55
business education, usually leading to the MBA,
involving some of the western world’s best
known universities and business schools.
Initially the most visible as well as the first to
launch a well-conceived and well-financed set of
products designed to serve a worldwide market
for business education was Cardean University,
a joint venture of five major business schools—
Stanford, Columbia, Carnegie Mellon, Chicago,
and the London School of Economics—and
UNext, a major Internet education company.
The problem was that the web-based products,
despite the prestige and visibility of Cardean’s
sponsors, never attracted the volume of
students it required to be a successful business
More recently, Universitas 21 has sought to
make a web-based, but nonetheless top-end,
business education available to students in
developing countries, offering MBAs at roughly
20 percent of the price of the in-residence
programs that the sponsoring universities offer.
A different set of institutions—for the most part
either present or former British Commonwealth
universities—forged a joint venture with the
Thomson Corporation, the single largest
economic enterprise with major investments in
programs of e-learning. Launched just this past
August, it is too early to tell if Universitas 21’s
educational offerings will attract students in
sufficient numbers to sustain the enterprise.
Already, however, the skeptics have cast their
doubts. As
The Chronicle of Higher Education
at least one online-education expert says
that the consortium may have set its
expectations too high. “What sells in
education is price and name,” says A.
Frank Mayadas, director of the Alfred P.
Sloan Foundation’s grant program for
online education. A new entity like
Universitas 21 Global may not be needed,
he says, now that many well-known
public and private universities offer
distance-education degrees that students
anywhere in the world can take.
Olsen, F. (August 28, 2003),
Chronicle of Higher Education
What Mayadas should have added, however,
is that while readily available, such courses also
have problems enrolling sufficient numbers of
students to recoup their initial investment.
The promise of an international community
of learners accessing a common set of
educational products and thus becoming a true
network without borders is not less appealing—
but fulfilling that promise remains a somewhat
distant goal.
Thwarted Innovation •••••••••••••••Page 56
As part of our work for
Weatherstation Project,
we have been
examining the thwarted nature of the e-learning revolution, asking,
“Why did the boom go bust?” The answer derives, first, from our
development of a conceptual framework to answer the question (Chapters 2 and
3); then, from our analysis of the market, based on the campus
and our tracking of e-learning across the Web (Chapters 4 and
5); and, finally, from our parsing of what we saw as e-learning’s troubling
assumptions (Chapter 6). The answer itself goes something like this.
E-learning, particularly in the United States, attracted a host of skilled
entrepreneurs and innovators who sought, as their most immediate goal, to
establish early prominence in an industry that had yet to be defined. They
sought to achieve market position quickly, lest others get there sooner and close
the door behind them. In seeking that advantage, they were aided by two
phenomena particular to postsecondary education and to the times. First, the
boom in commercial investments in e-learning enterprises followed more than a
decade of experimentation by faculty with the use of computers in teaching—a
good example was the development of “Virtual Shakespeare” at Stanford
University. A few experiments even flowered into commercially successful
products such as Maple and Mathematica, applications designed to teach students
calculus using electronically mediated instruction. While such work involved only
a minority of faculty, they were enough to advocate the new technology and
assure university leaders that the expertise needed for e-learning ventures was
available. As it turned out, however, that experimentation proved to be too
narrow to feed the e-learning boom that followed.
The dot-com boom provided a second major impetus. It spawned rosy
estimates of the market for Internet-based services—Michael Moe’s
extrapolation of a trillion-dollar market was only but one of a dozen or more
Conclusion: What’s Next?
Thwarted Innovation •••••••••••••••Page 57
highly publicized claims. Assured by the
technology’s advocates that the necessary
expertise was in hand or soon would be,
entrepreneurs both inside and outside
traditional postsecondary education rushed to
market with e-learning ventures. A veritable
feeding frenzy ensued, with large amounts of
time, effort, and capital committed to e-learning
development and marketing.
In retrospect, the rush to e-learning produced
more capacity than any rational analysis would
have said was needed. In a fundamental way,
the boom-bust cycle in e-learning stemmed from
an attempt to compress the process of innovation
itself. The entrepreneurial enthusiasm produced
too many new ventures pushing too many
untested products—products that, in their initial
form, turned out not to deliver as much value as
promised. Some successes were recorded and
certain market segments appear to remain
robust and growing, particularly the
transactional segment dominated by course
management systems like Blackboard and WebCT
and more recently receptive to computerized
testing routines like those developed by
Prometric. But overall the experience with e-
learning has been disappointing.
There were many after-effects to e-learning’s
inevitable crash, though perhaps the most
dangerous was that the experience jaundiced the
academy’s view concerning the actual value of
technologies promising electronically mediated
instruction and the market’s willingness to
accept new learning modalities. The hard fact is
that e-learning took off before people really knew
how to use it—before anything like a dominant
design was even on the horizon. Missing, in the
first instance, was a proven knowledge base of
sufficient breadth to persuade faculty that
adaptation was necessary. As a result, e-learning
entrepreneurs assumed a much higher level of
risk than they bargained—and not surprisingly,
most ended up paying the price.
Contextual Changes
In many ways the underlying message of
our report is that it is high time for “e-learning”
to get real—in a dual sense. Those who
promote, fund, and ultimately depend on e-
learning need to talk less and succeed more.
And those early adopters need to understand
that their success depends as much as the
context in which they operate as on the power
of the technologies they employ.
Necessary Changes Within the Academy
Itself. The first set of necessary conditions
involve changes within the academy itself.
The future of e-learning—particularly for
full-time, residential students—is linked to
the pace of educational change and reform.
The full potential of e-learning and
electronically mediated instruction will not
be realized unless there is an
acknowledgment, on the part of a large
number of faculty, that there is need to
substantially improve educational quality,
especially for undergraduates. What is
required is a commitment to organized
quality processes that transcend curricular
innovation, stress technology as an
important tool for improvement, and do not
Thwarted Innovation •••••••••••••••Page 58
assume things are going well, absent
evidence to the contrary.
•A Methodology for Calculating Costs and
Efficiencies. Once a significant number of
institutions, including a fair share of market
leaders, have determined they need to
improve the quality of their educational
programs and that e-learning can serve as a
means to that end, these institutions will find
themselves addressing questions of costs and
efficiencies. What adopting institutions will
require is a methodology that allows the
calculation of the economic contributions as
well as the costs of on-campus e-learning—
and how those contributions and costs
compare to those of more traditional forms of
on-campus instruction.
Less Rigid Tradeoffs Between Costs and
Quality. With the necessary educational
incentives and costs analyses in place, the
final step in this on-campus process will be
for institutions to better understand—and
hence be able to articulate and make a
central feature of their strategies and
plans—how e-learning can allow for a less
rigid set of trade-offs between costs and
quality. It requires a fundamental change in
a mindset which heretofore assumed that
education’s production functions are largely
fixed—that is, a change to one part requires
corresponding changes to all other parts,
because the relationship between inputs and
outputs is fixed. In the final analysis, what
the widespread adoption of e-learning
requires is a broad willingness on the part
of adopting institutions to search for more
flexible combinations of inputs: people,
facilities, and technology.
More Persistent Links Between Corporate
and Collegiate Education. Perhaps the
largest unknown is what will happen to
corporate training and education now that
the economy is once again growing. If that
growth results in substantial labor
shortages, everyone will be looking for ways
to speed up and make more efficient the
ways in which the labor force acquires new
skills. In the training depression that
accompanied this recession, e-learning made
some important inroads. Will they be
preserved and expanded? Will for-profit
collegiate education continue to expand and
will entities like the University of Phoenix
provide the bridge between corporate and
collegiate education? Will there be a
merging of efforts or the continued
development of what amounts to almost two
separate industries?
Technological Changes
The next set of necessary conditions for the
growth and expansion of e-learning focuses on
the technologies that make electronically
mediated learning feasible.
•A Dominant Design for Learning Objects.
First, there needs to emerge a dominant
design, particularly for the learning objects
that are e-learning’s building blocks. It is
not just a matter of making them more easy
to create—although that end is important—
Thwarted Innovation •••••••••••••••Page 59
but also more interchangeable and more
easily linked with one another. In
envisioning this context, it helps to think of
a railroad marshalling yard in which the
cars are the learning objects being
assembled behind locomotives that are the
user-interface drivers of an efficient e-
learning system. The marshalling yard only
works if the cars all have the same gauge
and have common couplers.
•A Technological Focus on What Students
Really Want. At the same time, it is
important for e-learning designers to resolve
questions regarding what students expect
from e-learning, as an extension of their
interest in other technologies. Here, we
require ways to motivate students to learn
using the technologies and to bring human
interaction into the equation in optimal ways.
Market Conditions
Finally, because e-learning was presented as
an innovation that could be financed through
venture capital and market revenues, there will
have to be some successes stories here as well.
More Market Successes. More specifically, e-
learning needs a substantial number of
showcase ventures that generate revenue
growth sufficient to sustain continuing
innovation without continuous infusions of
capital. In this arena, nothing will succeed
like success.
•A Real Market for Learning Objects. At the
same time, there needs to develop a robust
and growing “market” for e-learning objects.
Economies of scale in e-learning depend
critically on the ready importation of
learning objects. Finding, acquiring, and
using such objects in courses needs to
become an accepted element of faculty
These, then, are the conditions necessary for
e-learning to expand and flourish. We count
ourselves among the optimists who believe
electronically mediated instruction will become a
standard, perhaps even dominant, mode of
instruction. But we also understand that
progress over the next decade is likely to be
slow, probably best described as plodding. The
technology’s skeptics, emboldened by the fact
that, to date, e-learning’s failures have been
much more prominent than its limited
successes, will challenge each new product and
innovation. Ultimately, however, the lure of
anywhere-anytime learning will prove
irresistible—educationally as well as financially.
The next step will be to use the power of e-
learning to establish the networks without
borders that an increasingly fractured global
community desperately needs.
Three Practical Steps to Start the
It could be said that the revolution—though
slow in gaining momentum—has been launched.
The challenge at hand involves the acceleration
of e-learning’s adoption. Three practical steps
are required before e-learning and electronically
mediated instruction can achieve its full
Thwarted Innovation •••••••••••••••Page 60
Develop a Catalog of Lessons Learned. First
and foremost, the industry needs a catalog
of lessons learned. Our hope is that this
report represents a start in that direction.
•Map the Obstacles still to be Overcome.
Second, we will need a more realistic
mapping of the obstacles that must be
overcome—in terms of the technology itself;
in terms of assuring that universities in
particular become platforms of adoption as
well as sources of innovation and invention;
and in terms of achieving the market
conditions necessary for growth. In this
report we have also tried to provide an
initial enumeration of those conditions.
•Move Ahead in Developing Dominant Designs
and Global Networks. Finally, e-learning in
all four of its innovation cycles requires a
set of realistic strategies for developing the
dominant designs and the global networks
that will make it possible for e-learning to
come of age—and to signal its broad
Not the End of the Story
Despite the travails of the last several years,
e-learning has retained a core of true believers
who argue, still forcefully and occasionally
persuasively, that a revolution is at hand––that
the computer will do for learning today what
printing did for scholarship in the fifteenth
century. Don’t be fooled by the failures and
false steps, they proclaim. The best is yet to
More quiet, and also more numerous, are
the pragmatists who point out that e-learning is
alive and has in fact spurred a host of
important educational changes, probably best
symbolized by the widespread adoption of
course management tools such as BlackBoard
and WebCT. Money is being spent. Smart
classrooms are being built both on campuses
and businesses. Collegiate faculty and corporate
trainers are successfully integrating
electronically delivered learning materials into
literally thousands of courses focusing on both
traditional and non-traditional subjects. What
these pragmatists have come to understand is
that e-learning is evolving in ways that few had
We count ourselves among the pragmatists.
We believe the story of e-learning is still
unfolding—no one really knows what tomorrow
will bring, although we suspect that computer-
based learning technologies will continue to
serve as a major catalyst of innovation. The
underlying information technologies on which e-
learning depends are themselves too ubiquitous,
and the people attracted to having them serve
as learning platforms too smart, for us not to
take seriously the prospect that major changes
will flow from their efforts.
Thwarted Innovation •••••••••••••••Page 61
Appendix 1: Survey Responses of Faculty
Appendix 2: Survey Responses of Administrators
Appendix 3: Clients Served by Providers Engaged in e-Learning
Appendix 4: Content Offered by Providers Engaged in e-Learning
Appendix 5: Delivery System and Product Types by Providers Engaged in e-Learning
Appendix 6: Items Offered by Providers Engaged in e-Learning
Appendix 7: Services Offered by Providers Engaged in e-Learning
Appendix 1
Survey Responses of Faculty
Round 1 Responses
Round 2 vs Round 1
Round 3 vs Round 2
Round 3 vs Round 1
Currently there is a reduction in the traditional
workload for faculty in your department engaged in e-
76 66 10 14% 1% 3% 5% 0% 5% 9% 1% 8%
Currently there is funding dedicated to support e-
learning activities in your department/school.
74 24 50 37% 5% 1% 4% 0% 4% 9% 4% 5%
Currently there are technical staff in my
department/school dedicated to support e-learning
77 11 66 04% 3% 1% 5% 3% 3% 9% 5% 4%
In my department/school there are currently awards
for pedagogical innovation using new technologies.
73 48 25 43% 3% 0% 1% 1% 0% 4% 4% 0%
I have used a course management tool like
Blackboard, Prometheus, or WebCT.
77 22 55 03% 3% 0% 5% 5% 0% 8% 8% 0%
I have used off-the-shelf software packages such as
Dreamweaver, Maple, JMP or another statistical
77 32 45 04% 3% 1% 1% 1% 0% 5% 4% 1%
I have customized, off-the-shelf software for use in
77 45 32 05% 3% 3% 4% 4% 0% 6% 5% 1%
I have used multi-media presentations combining
text, voice, and video/digital images.
77 21 56 03% 1% 1% 0% 0% 0% 3% 1% 1%
I have developed e-learning course objects. 77 26 51 04% 4% 0% 1% 1% 0% 5% 5% 0%
I have developed a comprehensive e-learning course.
77 46 31 03% 3% 0% 1% 1% 0% 4% 4% 0%
I have required students to purchase software tools
such as Excel, SAS, SPSS, JMP, MAPLE, etc.
77 63 14 01% 1% 0% 1% 1% 0% 3% 3% 0%
I have required students to participate in electronic
discussion groups.
77 38 39 05% 4% 1% 4% 4% 0% 6% 6% 0%
I have required students to use web-based materials.
77 10 67 00% 0% 0% 4% 4% 0% 4% 4% 0%
I have used computer based assessment instruments
(tests or other forms of evaluation) in one or more
73 43 30 47% 7% 0% 5% 4% 1% 12% 10% 1%
I have assigned text books that include interactive
discs or access to a proprietary web site (password
77 29 48 03% 0% 3% 5% 4% 1% 8% 4% 4%
I have made assignments requiring students to use
the discs or proprietary web site that come with the
text book.
77 52 25 05% 4% 1% 5% 4% 1% 10% 8% 3%
I have supplied students other interactive discs or
77 44 33 05% 4% 1% 1% 1% 0% 4% 4% 0%
Appendix 1
Survey Responses of Faculty
Round 1 Responses
Round 2 vs Round 1
Round 3 vs Round 2
Round 3 vs Round 1
No Opinion
What is the frequency of my own use of e-learning
77 25 30 22 010% 8% 3% 8% 5% 3% 18% 13% 5%
What is the familiarity of faculty in my department
with e-learning?
75 24 31 20 24% 1% 3% 3% 3% 0% 7% 4% 3%
What is the current rate of the growth in e-learning
75 11 37 27 213% 7% 7% 10% 3% 8% 21% 8% 13%
How great is the value or benefit from e-learning?
77 438 35 06% 4% 3% 3% 1% 1% 9% 5% 4%
What is the capacity of e-learning to provide
opportunities to use resources more efficiently?
75 21 33 21 211% 8% 3% 12% 11% 1% 22% 18% 4%
What is the capacity of e-learning to serve new
student markets?
73 17 25 31 411% 8% 3% 5% 4% 1% 16% 12% 4%
To what extent is faculty overload responsible for the
reluctance of some faculty to experiment with e-
75 22 26 27 211% 11% 0% 5% 4% 1% 15% 13% 1%
How great is the concern among faculty about the
intellectual property rights of teaching material?
74 37 27 10 35% 5% 0% 9% 9% 0% 14% 14% 0%
What degree of discomfort would your students have
with the substitution of e-learning for face-to-face
75 28 24 23 211% 7% 4% 8% 4% 4% 19% 11% 8%
What is the degree of school/ department support
for faculty developing e-learning courses or course
77 16 32 29 010% 3% 8% 13% 5% 8% 21% 6% 14%
How much priority is given to e-learning initiatives
relative to other budget priorities in the school/
59 24 17 18 18 7% 3% 3% 17% 5% 13% 21% 6% 14%
To what extent are there workshops to introduce,
teach, train faculty to use of e-learning?