Abstract

The Second This article describes a future environment for

information systems in which information re-
Information places technology as the primary focus of firms'
efforts. The scenario is provoked by an aca-

Revolution demic research finding, and thus the article also

points to a way in which IS researchers might
have a greater impact on practice. The contri-
Thomas H. Davenport bution of the IT industry to the world economy is
also a focus of this vision, in that decreased IT
University of Texas at Austin sales contribute to a global recession.

Introduction

To some, it represented a worldwide cataclysm;

to others it had only moderately negative im-
plications. But almost all those involved - a
group that eventually included business people
in most industrialized nations -eventually came
to admit that the IT Recession of 1998 had many
long-term positive outcomes for the world
economy. These beneficial changes were
eventually known as "The Second Information
Revolution"; some called it "the real information
revolution," implying that the first one had
primarily been about technology rather than
information.

Trouble on the Horizon

The problem had been brewing for a couple of

years. The IT industry - semiconductor, compu-
ter, networking, and software firms - had been
producing higher-capacity products for decades.
Like the automobile firms of the 1960s, each
year the vendors' products were the equivalent
of "longer, lower, wider, with even more horse-
power." In 1998 Intel's latest processor had
eight million transistors, and was supposedly
1500 times more powerful than the original IBM
PC. Microsoft's Windows 98 allowed file names
of 1024 characters and let 64 programs run con-
currently in 64 megabytes of main memory. The
same firm's Office 98 word processing program
allowed simultaneous spelling, grammar, read-
ability, typographic, and time-to-print checking.
There were so many functional icons on the
word processing screen that little room was left
to display any text.

At the enterprise applications level, many

companies were installing so-called "Enterprise

The DATA BASE for Advances in Information S y s t e m s - Fall 1996 (Vol. 27, No. 4) 85
 Resource Planning" packages, but there panies spending more on IT did not perform
seemed to be few success stories. SAP AG, the better than their competitors; firms with more
leading vendor of such packages, released its advanced IT infrastructures also had no ap-
R/4 system, which basically included every app- parent competitive advantage. Individuals with
lication an organization could want, all integrated more powerful personal technologies were not
with a common object-oriented database. The judged by their supervisors or their peers to be
problem was that few if any customers had yet more productive workers. In fact, many such
completed their implementations of the R/3 "power users" were viewed as spending too
package. At the same time, many companies in much time on technology, and not enough on
the middle and late 1990s were completing their their real jobs. Senior executives at companies
moves to a client/server architecture. Many had with more powerful and expensive technologies
found, however, that the supposedly cheaper were less satisfied with their information environ-
and more powerful client/server systems were ments in their firms than those who spent less on
actually more expensive to build, more difficult to technology. While these results only confirmed
maintain, and more prone to failure. Company many managers' suspicions, they shook the IT
IT managers felt that they had to replace com- industry to its core.
puters and networks, and install new versions of
software, on a continuous basis. At both the PC Thus began what came to be known as the first
and the enterprise level, companies were be- "IT Recession," an economic downturn that
ginning to feel that they were on an IT treadmill, lasted several years and affected the US, Japan,
moving faster and faster with no "off" switch. and Europe. The first sign of the recession
came when the stock prices of technology firms
As early as 1996, there were indications of a - particularly such technology leaders as Intel,
substantial problem in the corporate computing Microsoft, Hewlett-Packard, Cisco Systems, and
world. A study of corporate PC use found that personal computer firms such as Compaq and
the average PC cost more than $13,000 per year Dell - plummeted. Their sales results for the
to use and support. Over 20 percent of the second quarter showed sharp declines in corpor-
average PC user's time at the computer was ate and individual purchases of technology.
spent "futzing," or attempting to make the Since IT companies had fueled the growth of the
machine cooperate with the user's desires. US and Japanese economies over the previous
There were mutterings about lowered produc- decade, these countries' GDP figures showed no
tivity. More and more users were beginning to growth in the second quarter, and even shrunk in
complain that they got too much electronic mail, the third and fourth quarters of 1998.
voice mail, faxes, and computer-originated paper
documents. The IT organizations in companies were also
shaken by the results of the research. Many
Academics Create a Crisis CEOs and other general managers decided to
"clean house," according to press reports. "I'd
In mid-1998, however, came the precipitating never been comfortable with those people any-
event for the recession. A consortium of leading way," one executive at an automobile firm
IT academics from various universities had been stated. At least two companies asked for the re-
formed in late 1996 to study the relationship bet- signations of all of the workers in their IT
ween IT and business success on multiple functions. Several firms sought outsourcing
levels. In April 1998, the consortium released its proposals, but others that had already out-
preliminary report to the companies that had sourced commented publicly that, while their IT
sponsored the study. Although the report was service might be somewhat cheaper, it wasn't
ostensibly proprietary to those firms, it was much different from before.
quickly leaked to the business press, and be-
came the cover story for two business mag- A Silver Lining in the Clouds
azines.
With this dramatic turnover in company IT per-
The study's results suggested that there was no sonnel began the first positive implications of the
relationship between investment in IT and bus- recession. New management teams came into
iness success across multiple industries. Com- IT functions; most managers had substantial

86 The DATA BASE for Advances in Information S y s t e m s - - Fall 1996 (Vol. 27, No.4)
experience in other functional areas. Later re- valent before the recession. Vendors had gotten
search found that the most common background the message: the dominant marketing emphasis
for these new information managers was mar- by the year 2000 was on practicality, cost
keting (clearly an information-intensive function reduction, and reliability, rather than increased
itself). Many of these managers took a fresh power and speed. Computer companies
look at the function's objectives and orga- stressed their long "mean-time before failure"
nizational structure. Because knowledge man- cycles and ease of installation and maintenance.
agement was a popular concept at the time, Network vendors focused on reliability and ease
"Information and Knowledge Management" be- of network management. Software vendors
came a popular new name for the function. introduced slimmed-down versions of popular
programs that met the needs of the common
Indeed, the common element among many user rather than the most sophisticated cus-
companies' approaches to IT was an emphasis tomers.
on information rather than technology. "Putting
the '1' back in 'IT'" became a popular phrase. Academics Document the Changes
These reconstituted information management
groups often involved combinations of library, A follow-up study in the year 2002 by the same
market research, and communications functions. group of academic researchers whose work had
It was common to undertake studies to identify triggered the recession found interesting and
information needs - independent of the tech- heartening results. Again the focus of their study
nology used to supply the information. Instead was the relationship between IT and individual
of creating complex data models that no users and organizational productivity. Economists had
could understand, analysts "shadowed" workers already suggested that the productivity of firms
to find out what information they really used, and had begun to increase substantially in 1999.
had sessions with groups of users to identify The IT academics demonstrated convincingly
ways in which their information environments that it was the change in orientation to IT that led
could be improved. These sessions addressed to the productivity increases. Individual users
a wide range of issues - not just technologies or now had familiar, reliable tools, and spent much
automated systems, but such topics as how or- less time tinkering with their computers, explor-
ganizational structures, politics, physical facility ing new software, and dealing with technology
layouts, and business process designs got in the problems. Infrastructure failures declined sub-
way of effective information. stantially; with fewer new technologies to
address, some IT groups had weeks on end
The IT-led recession continued for several years, without any network or data center failures.
largely because many firms decided that they System development finally matured as a dis-
would not upgrade their technology unless it cipline - largely because programmers finally
seemed absolutely necessary. Some com- had a stable technical environment in which they
panies developed a consensus that they would could do their work. Because the technology
"make do" with their existing technology base for environment didn't change much - and perhaps
five years. As one prominent CEO noted, because developers had pragmatic orientations
"We're going to spend five years catching up and business rather than technical backgrounds
with the technology we already have." Given the - the idea of software reuse took off in firms,
lack of demand, many vendors couldn't afford to even those that did not employ object-oriented
maintain the torrid pace of new product develop- technologies. Some of the most productive firms
ment they had maintained before the recession. even had libraries of COBOL modules and
Some computer industry executives admitted routines for the purpose of reuse.
privately that they had wondered why their
customers put up with two-year product ob- The Rise of "Appropriate
solescence cycles. Intel, for example, let five Technology"
years lapse between the introductions of its P7
and P8 microprocessor architectures. It took awhile for the IT management industry to
adapt to the new environment; but eventually
The new products that did emerge from the IT consultants, service providers, and the press
industry had a different flavor than those pre- caught on. In early 2004, a book was published

The DATA BASE for Advances in Information S y s t e m s - Fall 1996 (Vol. 27, No. 4) 87
called Appropriate Technology. It described ex- technical change in the early part of the new
amples of companies that had profited through century, these companies began to focus on
the use of conventional, tested technologies for how better to package, market, and sell their
their IT architectures and applications. One firm, information. New information products and ser-
for example, had built a company-wide inventory vices proliferated, from traditional and new
level alert system using simple e-maU tech- providers. Even record companies, which had
nology. The use of Web browsers (by 2004, put much of their efforts into such unsuccessful
these were considered old hat) for all information technical delivery vehicles as digital cassettes
interfaces had greatly simplified reporting and mini-disks, returned to their music and iden-
systems in another company. The author also tified new recording stars to reverse a decade-
described how one of the most successful long decline.
"virtual companies," with its employees working
from their homes all around the world, had deve- New financial information services abounded.
loped highly sophisticated information networks Banks, for example, discovered that they knew
using by-now-obsolete Pentium processor com- enough about their customers' financial profiles
puters. "You don't need a Ferrari to drive to the to offer them comprehensive financial planning
grocery store," the company's CEO stated. services - without the need for time-consuming
data entry by the customer. Credit card com-
Providers of IT education, once oriented to train- panies offered their customers tax preparation
ing people on how to use the latest generation of services - again with no customer data entry, as
hardware and software tools, were forced to long as key payments were made by credit card.
change focus. Instead of teaching about the use Financial fraud declined substantially; companies
of technology, new courses were developed that had automated fraud detection before the turn of
focused on the effective use of information. the century and found that computers could not
Managers and workers were taught how to detect fraud as well as humans. But new fraud
identify their own information needs, how to detection approaches relied on both human and
share information effectively with others, and computer-based screening. Similarly, the
how to make information-based decisions. pattern recognition and "data mining"
Universities developed course offerings involving technologies being explored by many firms in the
how to search for information and use it late 1990s had not proven to be as successful as
effectively, consistent with a newly popular intelligent humans and computers working in
emphasis on lifelong learning. A popular article concert.
in Harvard Business Review described twenty
ways to add value to information. This article
was well received by many middle managers, Work Habit and Lifestyle Changes
whose ranks had swelled again after reductions
in the 1990s. The middle management role had When companies and organizations jumped off
come to be viewed as largely involving the the technology treadmill, individual workers also
creation of effective information environments for began to change their behaviors with regard to
business units and functions. Instead of simply information and technology. By the late 1990s,
passing information up and down the hierarchy, many workers had felt that technology had led to
middle managers were evaluated on how they a substantial lengthening of their work hours.
added value to information by synthesizing, Even in their so-called "home" lives, workers
editing, and analyzing it. were answering e-mail, catching up on voice
mail, and having global videoconferences at all
Even information provider companies were hours of the night. With cellular phones, pagers,
affected by the change in the IT management and laptop computers, vacations now meant just
climate. In the fast-changing technology en- another place to do work. Divorce rates zoomed
vironments of the 1980s and 1990s, much of as family members became more stressed with
their energies had gone into creating access to less time to work out problems.
their information through new technologies and
interfaces - e.g., PC's and the Internet - and Again, a book crystallized the mounting dis-
creating client/server networks for their internal satisfaction with the technology-enabled life.
IT architectures. With the decreasing pace of Called Take Control of Your Information Destiny,

88 The DATA BASE for Advances in Information Systems-- Fall 1996 (Vol. 27, No.4)
the 2004 book counseled workers to shut off their private networks. Companies had also
electronic devices and pay attention to their grown tired of the furious pace of product deve-
spouses, children, and hobbies while at home. lopment in Internet browsers, Web-page man-
Companies adopted some of these recom- agers, and network-based programming lan-
mendations as policies after realizing that any guages. The worldwide IT recession slowed this
reasonable work/life balance had been de- pace somewhat, but not as much as in other do-
stroyed. A prominent study in 2003 by edu- mains of the IT world.
cational researchers found that children who
spent significant time using multimedia devices In 2004, the problem reached a crisis point
had less creative, vivid imaginations. Thereafter, when, due to excessive traffic, the Internet
sales of video games and educational software became virtually unusable. A worldwide group of
began to decline, and those programs that still Internet user representatives was convened; the
thrived encouraged children to use their imagin- group determined that Internet traffic could no
ative powers. longer be free. An Internet tariff was established
based on the number of bits sent or requested
These changes in information behaviors also ex- across the network and the time of day at which
tended into the work environment. Information they were transferred. The tariff was put into
overload had become a critical problem; many place immediately, and by mid-decade Internet
workers received over 100 messages per day, traffic had dropped considerably. Casual, aim-
and the level of communications from other less browsing of the Web almost disappeared,
media had not declined. For decades tech- and the network became primarily a business
nologists had predicted that automated filters or and educational tool (schools were exempted
electronic agents would deal with the information from paying the tariff).
overload problem. They were still making such
predictions in 2005, but by then many com- With response times improved, businesses
panies decided to take matters into their own began to return to the Internet. As with the rest
hands. Workers were given instruction on when of the IT field, emphasis began to shift toward
to use e-mail, and were told to avoid un- using the network and its applications as a
necessary copying of messages. Substantial means of accessing and displaying information
peer pressure was exerted on those who vio- and knowledge. There was a flowering of cre-
lated norms of e-mail behavior; every e-mail ative uses of the Internet for marketing and cus-
program could automatically generate a "why did tomer service applications; the use of the tele-
you send me this message?" response. Some phone declined precipitously for order-taking and
companies established charging systems for e- front-line service calls, and even when cus-
mail, which gave individual workers a com- tomers did need to speak with a person, they
munications budget. The greater the number of could rapidly access one by pressing the "Inter-
messages sent (and, in some cases, the more net phone" button on a service screen.
senior the recipient) the more quickly the budget
was used up. Some firms also established pe- A mental transformation with regard to the
riods during the day in which e-mail was not to Internet also took place in the latter half of the
be sent or read. Most employees greeted the century's first decade. The emphasis at the turn
policies with relief. of the century had been on the marvelous ability
of the Internet and the Web to increase access
A Return to Internet Sanity to information. Given the widespread "infoglut"
of the new decade, however, access was no
The Internet, of course, had been a key source longer deemed such a positive attribute. Inter-
of information overload. By the early 2000s the net-based information providers instead began
Internet had become extremely congested de- to focus on how to engage the attention of the
spite several attempts to bulk up the infra- information consumer. For their part, companies
structure. Many firms had concluded that the and employees that used the Internet as a
network was no longer well suited to the req- business tool began to focus on how information
uirements of business and had built up their own could be sought, acquired, and used to improve
Intranets. If they wanted customers to access business decisions. Several companies even
their systems, they simply gave them access to established policies whereby decisions by

The DATA BASE for Advances in Information S y s t e m s - Fall 1996 (Vol. 27, No. 4) 89
 individual managers were assessed in retrospect technology by itself cannot solve business pro-
to determine whether relevant information had blems. In fact, the technology itself never really
been acquired and used. achieved the same level of prominence and visi-
bility that it had attained in the mid-1990s.
A Return to Random Information Computers came to be viewed as "information
appliances," capable of doing useful things with
By the year 2007, a clear shift was underway in information and knowledge, but not nearly as
thinking about how to provide information to indi- valuable as humans in making effective use of
viduals. With the decline in technology orien- those resources. Technology vendors grew at
tation, researchers and practitioners were able to the rate of most other industries, and their gro-
focus more on how individuals dealt most effec- wth became somewhat cyclical as capital
tively with information, and how to produce spending rose or fell with the broader economy.
desirable information environments. Most magazines and even Web pages devoted
to breathless technology prose went under.
For example, a key assumption at the turn of the Business publications emphasized business,
century was that the more customized the infor- and only discussed technology in proportion to
mation environment, the better. Online informa- its size and importance as one industry in the ec-
tion services, for example, endeavored to pro- onomy.
duce news services tailored to the specific inter-
ests of the reader. Advertisements, product In many ways, information technology in the
offerings, and promotional campaigns were inc- early 21 st century became like automobiles - un-
reasingly "mass-customized" for the individual deniably useful, and a major force in the eco-
reader. nomy, but not quite as fascinating as they once
were. IT, like cars, still had their enthusiasts, but
Contrary to expectations, however, many read- they were a minority of the population. For most
ers found that these customized information people, computers and cars were simply ways to
environments were boring. As one reader com- get things done.
mented in a newspaper article describing the Later on in the 21 = century, IT began to re-
trend, "One of the reasons I read newspapers semble an even earlier technology: electricity.
was because I liked coming across items that Hailed as a transforming technology early in the
surprised me. With this "Daily Me" system (the 1900s, electricity was entirely taken for granted
name of one of the customized news services), I
by mid-century. By the year 2020, information
never learn about anything really new." In re- technology had been around for almost seventy
sponse to these complaints, both online and years and was similarly assumed to be available.
print-based information services returned to Like electricity in the 1990s, IT improvements
more varied content. The options of "random"
had plateaued in recent years; silicon had
and "as deemed important by editor" appeared reached its limits. Much of what could be
frequently in content selection menus. Even the automated had already been done. Specialized
paper newspaper staged a bit of a comeback in business functions for dealing with it began to
circulation figures.
recede, just as Vice Presidents for Electrification
had in the previous century. Information tech-
Back to Business as Usual nology, like electricity, had become something
everyone could use, and was deeply embedded
By the latter half of the new decade, the shift to in everyone's life. Managers, workers, and citi-
an information-focused business culture was zens were no longer absorbed by the technology
well established. By that time the IT product in- or its manufacturers; instead, the Second
dustry had returned to growth and profitability, as Information Revolution was devoted to the use of
firms replaced aging infrastructures, and vendors this once-wondrous tool in the service of
offered products with real innovations in func- information.
tionality.
About the Author
But the techno-hype of the past never returned. Thomas H, Davenport is professor and director
Managers and workers had finally realized that of the information management program at the

90 The DATA BASE for Advances in Information S y s t e m s - - Fall 1996 (Vol. 27, No.4)
University of Texas at Austin. He works in the
areas of information, knowledge, and process
management. He was previously director of
research at Ernst & Young's Center for Business
Innovation and a consultant and researcher at
McKinsey & Company and CSC Index. His
Ph.D. is in sociology and organizational behavior
from Harvard University. He makes no claims to
be able to predict the future, but he did write the
first article and the first book on the topic of
business process reengineering and several of
the earliest articles on the topic of knowledge
management.
E-mail: tdav@notes.bus.utexas.edu

The DATA BASE for Advances in Information S y s t e m s - Fall 1996 (Vol. 27, No. 4) 91