Advanced information systems case analysis

Y E

L M

A G

C Y

A N

B L

A C

K

november 2006 141

n the information era, the best of times are the worst of times. Computer

hardware keeps getting faster, cheaper,

and more portable; new technologies

such as mashups, blogs, wikis, and busi-

ness analytic systems have captured the

imagination; and corporate IT spending

has bounced back from the plunge it

took in 2001. In 1987, U.S. corporations’

investment in IT per employee averaged

$1,500. By 2004, the latest year for

which government data are available,

that amount had more than tripled to

$5,100 per employee. In fact, American

companies spend as much on IT each

year as they do on offices, warehouses,

and factories put together.

However, as IT’s drumbeats become

louder, they threaten to overwhelm gen-

eral managers. One of the biggest prob-

lems companies face is coping with the

abundance of technologies in the mar-

ketplace. It’s hard for executives to fig-

ure out what all those systems, applica-

tions, and acronyms do, let alone decide

which ones they should purchase and

I

There are three

categories of IT, each

of which provides

different organizational

capabilities – and

demands very different

kinds of management

interventions.

Mastering the Three Worlds of Information Technology

by Andrew McAfee

C L

A U

D IA

N E

W E

L L

T O O L K I T

how to successfully adopt them. Most

managers feel ill equipped to navigate

the constantly changing technology

landscape and thus involve themselves

less and less with IT.

Adding to executives’ diffidence, cor-

porate IT projects have often delivered

underwhelming results or been outright

failures. Catastrophes – such as the one

at American pharmaceutical distribu-

tor FoxMeyer Drug, which went into

Chapter 11 and was sold in 1997 when

a $100 million IT project failed – may

be less frequent today than in the past,

but frustration, delay, and disappoint-

ment are all too common. In 2005, when

IT consultancy CSC and the Financial

Executives Research Foundation con-

ducted a survey of 782 American exec-

utives responsible for IT, 50% of the

respondents admitted that “aligning

business and IT strategy” was a major

problem. The researchers found that 51%

of large-scale IT efforts finished later

than expected and ran over budget.

Only 10% of companies believed they

were getting high returns from IT in-

vestments; 47% felt that returns were

low, negative, or unknown.

Not surprisingly, any fresh IT pro-

posal sparks fiery debates in board-

rooms. Some boards say “Why should

we bother? IT isn’t strategic, so it doesn’t

matter in a competitive sense. We should

be minimizing our technology expendi-

tures.” Others argue “Whether IT mat-

ters or not, we shouldn’t be doing it our-

selves. Companies are becoming virtual,

and software is becoming rentable, so

why do IT the old-fashioned way?” Thus,

executives try to delegate, outsource,

rent, rationalize, minimize, and gener-

ally remove IT from their already long

list of concerns.

But managers who distance them-

selves from IT abdicate a critical respon-

sibility. Having studied IT for the past

12 years, I believe that executives have

three roles to play in managing IT: They

must help select technologies, nurture

their adoption, and ensure their ex-

ploitation. However, managers needn’t

do all those things each time they buy

a new technology. Different types of

IT result in different kinds of organiza-

tional change when they are imple-

mented, so executives must tailor their

roles to the technologies they’re using.

What’s critical, though, is that execu-

tives stop looking at IT projects as tech-

nology installations and start looking

at them as periods of organizational

change that they have a responsibility

to manage.

Building an Effective IT Model Everyone who has studied companies’

frustrations with IT argues that technol-

ogy projects are increasingly becoming

managerial challenges rather than tech-

nical ones. What’s more, a well-run IT

department isn’t enough; line manag-

ers have important responsibilities in im-

plementing these projects. An insightful

CIO once told me,“I can make a project

fail, but I can’t make it succeed. For that,

I need my [non-IT] business colleagues.”

Managers I’ve worked with admit pri-

vately that success with IT requires their

commitment, but they’re not clear

where, when, and how they should get

involved.

That’s partly because executives usu-

ally operate without a comprehensive

model of what IT does for companies,

how it can affect organizations, and

what managers must do to ensure that

IT initiatives succeed. As HBS professor

Clayton M. Christensen and Boston Uni-

versity professor Paul R. Carlile point

out in their working paper “The Cycles

of Theory Building in Management Re-

search” (Harvard Business School, Feb-

ruary 2005), a good model or theory

does two things: It groups important

phenomena into categories, and, within

categories, it makes statements of cause

and effect. Yet even state-of-the-art mod-

els of IT’s impact consist only of state-

ments about individual technologies,

such as “CRM lets you get closer to cus-

tomers” and “SCM enables you to re-

duce inventory.” Such declarations don’t

help executives; they’re more akin to

sales pitches than statements of fact.

These assertions are also silent about

why technologies will deliver to compa-

nies the benefits they have promised.

Why will customers start confessing

their deepest desires to your customer

relationship management system? Why

will suppliers start delivering just in

time when you set up a supply chain

management system? Existing models

don’t help executives choose among

technologies, either. Every business

wants both to be closer to customers

and to keep inventory levels low – but is

it better to first invest in CRM or SCM

improvements?

One way to build a comprehensive

model is to place IT in a historical con-

text. Economists and business historians

agree that IT is the latest in a series of

general-purpose technologies (GPTs),

innovations so important that they

cause jumps in an economy’s normal

march of progress. Electric power, the

transistor, and the laser are examples

of GPTs that came about in the nine-

teenth and twentieth centuries. Compa-

nies can incorporate some general pur-

pose technologies, like transistors, into

products, and others, like electricity, into

processes, but all of them share specific

characteristics. The performance of such

technologies improves dramatically

over time. As people become more fa-

miliar with GPTs and let go of their old

ways of thinking, they find a great many

uses for these innovations. Crucially,

142 harvard business review | hbr.org

T O O L K I T • M a s t e r i n g t h e T h r e e W o r l d s o f I n f o r m a t i o n Te c h n o l o g y

Andrew McAfee ([email protected]) is an

associate professor at Harvard Business

School in Boston. Visit his blog at blog.hbs

.edu/faculty/amcafee.

Executives need to stop looking at IT projects as

technology installations and start looking at them

as periods of organizational change that they have

a responsibility to manage.

Y E

L M

A G

C Y

A N

B L

A C

K

general purpose technologies deliver

greater benefits as people invent or de-

velop complements that multiply the

power, impact, and uses of GPTs. For

instance, in 1970, fiber-optic cables

enabled companies to employ lasers,

which had already been in use for a de-

cade, for data transmission.

The complements of process GPTs are

organizational innovations, or changes

in the way companies get work done.

Research suggests that four organi-

zational complements – better-skilled

workers, higher levels of teamwork, re-

designed processes, and new decision

rights – allow process GPTs to deliver

improved performance. For instance, in

the early twentieth century, factories in

America replaced central motors driven

by waterwheels or steam engines with

newly invented electric motors. These

large motors were connected to a drive-

shaft, which was connected by belts to

the factory’s machines. At first, electric

motors were bolted onto the old drive-

shafts. As time went on, businesses built

smaller electric motors and connected

one to each machine. The new motors

gave companies the freedom to redesign

work flows. They were able to build

long, low factories instead of high, nar-

row ones, for example, and to arrange

machines in rows that later became as-

sembly lines. However, businesses had

to hire workers who were both more

skilled and better able to independently

make decisions at each station. Once all

the organizational complements to elec-

tric motors were in place, they maxi-

mized the technology’s impact and

boosted productivity in the U.S. manu-

facturing sector.

These insights are also true of IT, but

with one distinction: Information tech-

nologies, my research shows, don’t enjoy

november 2006 143

M a s t e r i n g t h e T h r e e W o r l d s o f I n f o r m a t i o n Te c h n o l o g y • T O O L K I T

the same relationships with the four

organizational complements that other

process GPTs have. Some information

technologies can deliver results without

the complements being in place; others

allow the complements to emerge over

time; and still others impose the com-

plements they need as soon as compa-

nies deploy the technologies.

Based on those variations, we can

classify IT into three categories. (See

the exhibit “The Three Varieties of

Work-Changing IT.”) Each offers com-

panies distinctive capabilities, delivers

unique benefits, and triggers organiza-

tional changes of different types and

magnitudes. This classification can help

leaders understand which technologies

they must invest in as well as what they

should do to maximize returns. It can

also indicate which IT initiatives are

going to be relatively easy to imple-

ment and on which projects executives

should focus. In that light, IT manage-

ment starts to look less like a black art

and more like the work of the executive.

The Three Categories of IT Executives often talk about the revolu-

tion that computers have brought about

in companies, but, as the IT model I’ve

described illustrates, that’s an oversim-

plification. IT sets off several kinds of

revolutions in organizations because

technologies fall into three distinct

categories.

Function IT (FIT) includes technolo- gies that make the execution of stand-

alone tasks more efficient. Word proces-

sors and spreadsheets are the most

common examples of this IT category.

Design engineers, accountants, doctors,

graphic artists, and a host of other spe-

cialists and knowledge workers use FIT

all the time. People can get the most

value from these technologies when

their complements are in place but can

also use FIT without all of the comple-

ments. For instance, an R&D engineer

can use a computer-aided design (CAD)

program to improve the way he does

his work without making any changes

in how the rest of the department func-

tions. Furthermore, FITs don’t bring

their complements with them. CAD

software, for example, doesn’t specify

the processes that make the most of its

power. Companies must identify the

complements FIT needs and either de-

velop them or allow users to create them.

FIT is powerful. Five years ago, Ducati

announced that it would enter the

MotoGP racing circuit in 2003. Its de-

signers kicked off a project to build a

suitable motorcycle in November 2001.

They started by using simulation soft-

ware to build and test virtual engines.

The simulations made the team realize

that a two-cylinder engine wouldn’t be

powerful enough to win races, so it de-

cided to build Ducati’s first four-cylinder

engine. The team finished designing

the engine in August 2002; a motorcy-

cle powered by the engine was zoom-

ing around test tracks two months later;

and the project was largely complete by

January 2003. The Italian company par-

ticipated in the MotoGP circuit in 2003

and outperformed most of its rivals:

Ducati placed second in the manufac-

turers’ standings, a ranking of compa-

nies that race motorcycles on the cir-

cuit, and its riders finished fourth and

sixth in the individual standings.

Ducati’s experience with FIT vividly

demonstrates the capabilities of this IT

category:

• Enhancing experimentation capacity.

Ducati’s engineers built thousands of

engines and motorcycles and compared

their performance without touching a

sheet of steel.

• Increasing precision. The company’s

designers came to trust the software

so much that if test results disagreed

with a simulation, they told me, the

first reaction was to mistrust the test

results.

Network IT (NIT) provides a means by which people can communicate with

one another. Network technologies in-

clude e-mail, instant messaging, blogs,

and groupware like Lotus Notes. NIT al-

lows people to interact, but it doesn’t

define how they should interact. It gives

people freedom to experiment instead

of telling them what they must do. Un-

like FIT, network IT brings comple-

ments with it but allows users to imple-

ment and modify them over time.

In 2005, investment bank Dresdner

Kleinwort Wasserstein introduced three

network technologies: messaging soft-

ware, employee blogs, and a company

wiki, a Web site that employees could

contribute to or edit without needing

permission or HTML skills. DKW’s peo-

ple generate data, get opinions, and find

answers by using the messaging soft-

ware to contact the firm’s traders and

analysts across the world. Many man-

agers write blogs or post comments on

others’ blogs. Some DKW directors see

the wiki as a way to deal with e-mail

overload and encourage their teams to

post agendas, to-do lists, and work in

progress on the wiki rather than circu-

lating them via e-mail.

As the DKW example illustrates,

NIT’s principal capabilities include the

following:

• Facilitating collaboration. Network

technologies allow employees to work

together but don’t define who should

work with whom or what projects em-

ployees should work on. At DKW, ad hoc

teams have formed because employees

read one another’s blogs. These teams

have used the wiki to accomplish tasks,

and they have disbanded without orders

from senior executives.

• Allowing expressions of judgment.

NITs are egalitarian technologies that

let people express opinions. DKW em-

ployees use blogs to voice their views

about everything from open-source soft-

ware to interest rate movements.

• Fostering emergence. “Emergence” is

the appearance of high-level patterns or

144 harvard business review | hbr.org

T O O L K I T • M a s t e r i n g t h e T h r e e W o r l d s o f I n f o r m a t i o n Te c h n o l o g y

Classifying IT into three types can help leaders

understand which technologies they must invest in

as well as what they should do to maximize returns.

Y E

L M

A G

C Y

A N

B L

A C

K

information because of low-level inter-

actions. These patterns are useful be-

cause they allow managers to compare

how work is done with how it’s sup-

posed to be done. Emergence is also

valuable for users. For instance, employ-

ees can easily search and navigate DKW’s

blogs and wiki for trends and data even

though nobody is in charge of making

them easy to use.

Enterprise IT (EIT) is the type of IT application that companies adopt to re-

structure interactions among groups of

employees or with business partners.

Applications that define entire business

processes, such as CRM and SCM – as

well as technologies, such as electronic

data interchange, that automate com-

munications between companies – fall

into this category. Unlike network tech-

nologies, which percolate from the bot-

tom, enterprise technologies are very

much top-down; they are purchased

and imposed on organizations by senior

management. Companies can’t adopt

EIT without introducing new inter-

dependencies, processes, and decision

rights. Moreover, companies can’t slowly

create the complements to EIT; changes

become necessary as soon as the new

systems go live.

In 2002, American retail drugstore

chain CVS became concerned about

the long wait times at its pharmacies

and reexamined two steps in its pre-

scription fulfillment process that it had

automated. Initially, its pharmacies

had performed the first step, a safety

check for drug interactions, one hour

before the customer’s desired pickup

time. After that, it checked whether the

insurer would pay for the medicine.

Despite automating the process, CVS

often was unable to resolve all of the

outstanding safety and insurance issues

by the promised pickup times, which ir-

ritated customers. CVS then decided to

reverse the order in which the steps

were executed. The change met with

resistance from many CVS pharma-

cists, who felt that since the drug safety

check was the more important of the

two, it should be the first step in the

process. The team that was rolling out

the project reasoned with the skeptics

but eventually realized that it would not

win them all over. So it instructed the

pharmacies to perform the insurance

review first, when customers dropped

off prescriptions, rather than immedi-

ately before pickup time. That allowed

technicians to work with customers

to correct small glitches, such as date

of birth errors in health insurance rec-

ords, that would prevent drug reim-

bursements and to warn people if they

were likely to run into bigger issues,

such as the nonpayment of insurance

premiums. The new sequence also let

CVS’s pharmacists incorporate the

safety check into their quality control

procedures instead of treating it as a

separate step. Redesigning the fulfill-

ment process cut wait times at CVS by

as much as 80%, which improved cus-

tomer satisfaction.

As CVS’s experience shows, EIT’s pri-

mary capabilities include the following:

• Redesigning business processes. Be-

cause CVS employees couldn’t fill pre-

scriptions until they had completed the

two checks in the new sequence, the re-

vamped fulfillment process wasn’t just

a good idea in theory – CVS employees

had to execute the process in that partic-

ular sequence. EIT gives managers con-

fidence that employees will execute pro-

cesses correctly.

• Standardizing work flows. Once com-

panies identify a complementary busi-

ness process, they can implement it

widely and reliably along with the EIT.

CVS rolled out its new process in 4,000

november 2006 145

M a s t e r i n g t h e T h r e e W o r l d s o f I n f o r m a t i o n Te c h n o l o g y • T O O L K I T

The Three Varieties of Work-Changing IT

IT Category Definition Characteristics Examples

Function IT IT that assists with • Can be adopted without complements Simulators, spreadsheets,

the execution of • Impact increases when complements computer-aided design,

discrete tasks are in place and statistical software

Network IT IT that facilitates • Doesn’t impose complements but lets them E-mail, instant messaging,

interactions without emerge over time wikis, blogs, and mashups

specifying their • Doesn’t specify tasks or sequences

parameters • Accepts data in many formats

• Use is optional

Enterprise IT IT that specifies • Imposes complements throughout the Software for enterprise resource

business processes organization planning, customer resource

• Defines tasks and sequences management, and supply

• Mandates data formats chain management

• Use is mandatory

outlets across the United States in less

than a year.

• Monitoring activities and events effi-

ciently. EITs can allow managers to get

an accurate and up-to-date picture of

what’s happening throughout the en-

terprise, often in something close to real

time. CVS’s software lets executives

know how many prescriptions are filled

every day in each location, how long it

takes to fill each prescription, and what

kinds of fulfillment problems employ-

ees had to tackle.

Managing the Three Types of IT Across the three IT categories, execu-

tives have three tasks. First, they must

help select IT applications that will de-

liver the organizational capabilities they

desire. Second, they must lead adoption

efforts that result in the creation of

complements for those technologies.

And third, they must shape the exploi-

tation of IT by ensuring that technolo-

gies, capabilities, and complements stay

aligned.

IT selection. Companies often select IT applications after one of their execu-

tives hears about a new technology and

wonders why his or her organization

hasn’t invested in it yet. This approach

is pervasive. How often do you hear,

“Shouldn’t we take a look at Technol-

ogy X?” or “Why can’t Technology Y do

that for us?” Companies will even in-

vest in a technology because everyone

else in the industry has purchased it or

because it comes with glowing recom-

mendations from consultants, analysts,

and journalists.

Trouble is, there’s an endless supply of

new applications, partly because of in-

novation and partly because of clever

rebranding. Companies can’t possibly

evaluate all the new applications that

cross their paths. Another, more funda-

mental, problem is that this method of

choosing applications reflects an outside-

in approach: Executives describe a tech-

nology that’s available in the outside

world and propose that it should be

brought into the company. No one stops

to think about whether the organiza-

tion actually needs the capabilities that

the technology offers. Between 1999

and 2001, American companies spent

$130 billion on IT they never used, ac-

cording to one estimate. An outside-in

mentality was surely behind much of

that waste.

A more sensible question for execu-

tives to ask is “What do we need IT to do

for us?” For instance, they might con-

sider, Do our company’s engineers need

to increase their experimentation capac-

ity? Do our sales and marketing depart-

ments need to collaborate more often?

Do we need to standardize fulfillment

processes throughout the world? Man-

agers should also set IT priorities. They

must decide, Is it more important to

have a single source of employee data or

to get weekly reports from the sales

force about client contacts? Would the

R&D department be better off if it

could conduct more simulations or if

it had an online space for brainstorm-

ing? Would it be more valuable to en-

hance the enterprise system by adding

a layer of analysis software or by extend-

ing it to suppliers through a private data

exchange? These are tough choices, but

they are appropriate ones for top man-

agement teams to talk through. (See the

sidebar “The IT Dialogue.”)

An inside-out approach puts the spot-

light squarely on the business before

evaluating the technology landscape; it

focuses on the capabilities that IT can

provide rather than on the technologies

themselves. A discussion among execu-

tives about capabilities will highlight

what the business most wants to be good

at – and it will show whether there’s

agreement about what the business

needs to be good at. Once the company’s

business needs are clear, the technolo-

gies it requires will come into focus. Typ-

ically, FIT delivers productivity and op-

timization, NIT increases collaboration,

and EIT helps standardize and monitor

work. Thus, when executives decide what

capabilities they need, they will know

what kind of IT to buy and the nature of

the initiatives they must manage.

In our 2004 case study “Enterprise IT

at Cisco,” two HBS colleagues, F. Warren

McFarlan and Alison Berkley Wagon-

feld, and I described how Cisco used the

inside-out approach to refocus the IT

selection process. Cisco realized that

there were drawbacks in its IT decision-

making process as it was trying to re-

cover in late 2001 from a fall in reve-

nues. CIO Brad Boston found that Cisco

had nine order status tools. Each of

them used data from different sources,

which used different definitions for key

terms. As a result, the systems couldn’t

give the company a clear picture of its

orders. There were similar problems in

the sales organization. Boston and his

colleagues realized that Cisco needed to

improve its standardization and moni-

toring capabilities, so they selected an

upgraded ERP system and a customer

database. They also decided to imple-

ment the new technologies across the

company even though it was costly and

time-consuming to do so. The ERP proj-

ect required three years to implement

and cost the company approximately

$200 million. Since Cisco couldn’t gain

the capabilities it wanted without those

technologies, however, it chose to invest

in them.

IT adoption. After IT selection, exec- utives’ attention turns to adoption: the

hard work of putting the technologies

they’ve invested in to productive use.

At this stage, managers’ main respon-

sibility is to help create the comple-

ments that will maximize IT’s value.

FIT doesn’t bring its complements with

it, so managers must find ways of iden-

tifying them. That’s what BMW’s chief

designer, Chris Bangle, did in the late

1990s when he wanted designers to use

computer-aided styling (CAS) software

in addition to paper, clay, and wood. As

Bangle explained to HBS professor Ste-

fan Thomke during an interview, the

146 harvard business review | hbr.org

T O O L K I T • M a s t e r i n g t h e T h r e e W o r l d s o f I n f o r m a t i o n Te c h n o l o g y

Once the company’s business needs are clear, the

technologies it requires will come into focus.

Y E

L M

A G

C Y

A N

B L

A C

K

designers were reluctant to use the soft-

ware, even though Bangle had hired

CAS specialists to work alongside them.

One day, Bangle declared that within

three months, the CAS team would have

to pay for itself – or he would sell the

team’s computers. He didn’t twist the

designers’ arms; he pressured the CAS

specialists and modelers. They helped

the designers adopt the software and

create new design processes. Bangle

knew he couldn’t force the technology’s

adoption or merely hope that comple-

ments would emerge. He had to allow

his team to discover new ways of work-

ing – although he could prod it a little.

There’s an interesting dichotomy in

executives’ roles when it comes to NIT

adoption. Because the use of such tech-

nologies is voluntary rather than manda-

tory, they make users feel more, rather

than less, in control of their work. As

a result, their adoption isn’t difficult.

However, managers still have to inter-

vene with new technologies, such as

groupware, wikis, and blogs, by demon-

strating how they can be used and by

setting norms for participation. Once

network technologies are properly es-

tablished, their use takes off, and the

challenge for managers is to refrain

from intervening too often or with too

heavy a hand.

In stark contrast to FIT and NIT, en-

terprise IT is hard for companies to

adopt. The benefits look great to people

at the top, but employees usually dis-

like EIT technologies. Unlike network

technologies, they don’t just enable

new ways of working; they dictate them.

Enterprise systems define new cross-

function business processes, impose the

processes on employees without allow-

ing employees to modify them, and

bring higher levels of oversight. Most

employees don’t like having new pro-

cesses dictated to them by a piece of

software and will use a variety of tech-

niques to prevent the adoption of en-

terprise technologies. Executives must

intervene forcefully throughout EIT

adoption efforts because new processes,

changed decision rights, and greater

interdependence come hand in hand

with these technologies.

november 2006 147

M a s t e r i n g t h e T h r e e W o r l d s o f I n f o r m a t i o n Te c h n o l o g y • T O O L K I T

The IT Dialogue

Business and IT leaders should meet periodically to discuss the state of

the company’s IT-based capabilities. During these meetings, the CIO’s role

isn’t to unveil new technologies but to collaborate with executives on capa-

bility development. The discussions should cover organizational footprints:

the geographic, functional, and divisional range over which the company

will deploy new technologies. The expense, difficulty, and time involved in

deploying IT – especially enterprise IT – increases as the footprint grows.

The conversation can be broken down according to IT category and should

address some specific questions.

Functional IT • Will any of the new software on the market enable our engineers,

scientists, analysts, and other workers to do their jobs more efficiently?

• Are our function technologies outdated? If so, why? What has changed?

Network IT • How do our people collaborate? Do we know what technologies they’re

using?

• Do we have ways of letting qualitative information flow horizontally and

vertically within the company as well as back and forth with customers

and suppliers?

• If we wanted to get broad feedback on an important topic, how would

we do it?

• How do we know what our people are working on and what they think

the hot topics are?

Enterprise IT • In what ways are our current processes not supporting the needs of the

business? Which ones need to be redesigned? Which ones should be

extended to our customers and suppliers?

• Are there best practices that should be embedded in our enterprise IT

endeavors so that they can be deployed more widely? How much more

widely? Do they need to be adjusted at all for new environments?

• Are there important business activities, events, or trends that we should

monitor? If we aren’t monitoring them, why not? Are the data unavailable

or stored across so many systems that the information is difficult to

assemble?

• What’s the most recent period that we could easily analyze? One hour

ago? Yesterday? Last month? Last quarter?

It’s important for executives to know how long a proposed IT effort

will take and what it will cost, but the ROI figure is nearly worthless. I’ve

seen many business cases prepared to justify IT investments. Virtually

all of them predicted increased revenues, higher profits, and lower costs.

And all of them ascribed the benefits entirely to the technology, making it

sound like a can’t-lose proposition. In reality, IT is never a sure bet because

of the complex interplay between technologies, capabilities, and comple-

ments. Smart companies spend little time predicting the financial benefits

that will accrue from IT efforts. They keep track of spending and milestones

and constantly check to see if they are on track to gain the IT-based capabil-

ities they desire.

In fact, the biggest mistake business

leaders make is to underestimate re-

sistance when they impose changes in

the ways people work. In 2002, a

Boston-based hospital set up an IT sys-

tem that replaced handwritten prescrip-

tions with online orders. The system in-

stantly checked doctors’ prescriptions

for harmful doses or drug interactions

and transmitted the orders to the hospi-

tal pharmacy. Even though studies had

demonstrated that the system would re-

duce medication errors, physicians bit-

terly resisted. They complained that the

computer-based process was slower and

less convenient than paper-based order-

ing and that the built-in error checking

didn’t work. They protested so strongly

that the hospital was able to roll out

the system in only a few departments.

Today, most of the doctors continue to

write prescriptions on paper and fax

them to the hospital’s pharmacy. The

system’s champions were caught com-

pletely off guard by the doctors’ reac-

tion to the monitoring and standardi-

zation capabilities that the hospital

sought.

EIT adoptions can give rise to several

kinds of problems. For example, EIT

projects often become delayed as em-

ployees and managers negotiate the use

of complements, such as new processes,

that the technology has imposed. Com-

panies often settle for solutions that are

more modest than originally planned

and gain only some of the capabilities

they had initially sought. Firms may

even abandon EIT adoptions altogether.

Even worse, some businesses don’t aban-

don an EIT project when they should,

which wreaks havoc on performance.

For instance, in the late 1990s, both Her-

shey and Nike implemented technolo-

gies that were a poor fit with their busi-

ness needs and processes. As a result,

the finances and share prices of both

companies suffered.

All the successful EIT adoptions I’ve

studied have used the same process for

avoiding failure, and all the unsuccessful

EIT adoptions I’ve studied have not used

it: They have decided at the outset how

key issues about configuration and

other aspects of the adoption will be

raised and how they will be settled. The

most important participants in this task

are not IT specialists or consultants but

business leaders from the areas affected

by the new technology. The more areas

there are, and the more their work is

being changed, the more the adoption

effort needs a seasoned leader. A mid-

level project manager doesn’t have the

formal or informal authority required

to make and implement these tough de-

cisions. At CVS, for example, the leader

of the EIT project was responsible for

both IT and store operations, so he had

the authority to deploy the new pro-

cess despite opposition from the chain’s

pharmacists. Similarly, despite Cisco’s

decentralized culture, the company set

up a business process operating com-

mittee (BPOC) that consisted of six se-

nior executives and the CIO. The BPOC

met throughout the EIT adoption effort

to make policy and process decisions

and to signal that Cisco wouldn’t back

away from establishing the comple-

ments that the technology needed even

though there was resistance within the

organization.

Leaders who successfully implement

EIT try to build consensus in the organi-

zation, but they’re also willing to push

ahead without having everyone on

board every step of the way. Their deci-

sive style runs counter to the usual ad-

vice about how executives should get

users to accept and own new technolo-

gies. For example, in 1999, when a mu-

tual fund company set up a CRM sys-

tem, it asked its salespeople to enter the

information about their meetings with

brokerages and institutional investors

into the system. The sales reps saw this,

correctly, as an attempt to capture

knowledge that existed only in their

heads. They refused to use the system,

which delivered little value to anyone

for years. The situation changed with

the arrival in 2001 of a new sales presi-

dent, who demanded that reps enter in-

formation into the CRM system, threat-

ened to withhold commission payments

from those who didn’t, and instructed

her direct reports to cross-check the

sales reps’ entries against expense re-

ports. The president’s policy was met

with stiff opposition, but the reps

quickly realized that they had to accede

to the demands of the new boss if they

wanted to continue working for the

company.

IT exploitation. A business leader’s third IT-related responsibility is to ex-

tract the maximum benefit from tech-

nologies once they are in place.

Companies can best exploit FIT by

fine-tuning organizational complements.

When HBS professors Marco Iansiti and

Alan MacCormack studied the 1995

America’s Cup sailing competition, they

found that all of the teams used simu-

lation software to help them design

their boat keels. Most teams worked

with universities and aerospace compa-

nies to build sophisticated simulations

and used either mainframes or super-

computers to do the work. They were all

beaten by Team New Zealand, which

used less powerful workstations but

brought the computers down to the

docks where its boats were built. The

New Zealand team also encouraged

experimentation and teamwork and

pushed keel modification decisions

down the organization. Because the

other teams didn’t do all of those things,

they couldn’t harness the full power of

the FIT.

Employees exploit older NITs such as

e-mail and instant messaging on their

own, but business leaders have a role to

play in exploiting newer technologies

like blogs and wikis. They can help sus-

tain and increase the use of comple-

ments to make the technology continu-

ally more effective, primarily by guiding

users. Darren Leonard, a managing di-

rector in the global equity derivatives

business at Dresdner Kleinwort, recalls

148 harvard business review | hbr.org

T O O L K I T • M a s t e r i n g t h e T h r e e W o r l d s o f I n f o r m a t i o n Te c h n o l o g y

The biggest mistake

business leaders make

is to underestimate

resistance when they

impose changes in the

ways people work.

Y E

L M

A G

C Y

A N

B L

A C

K

how he got his colleagues to use the

company’s wiki: “First, if a wiki has no

structure, it’s perceived not as an oppor-

tunity but as anarchy, and our people

have no time for anarchy. I went back

to my initial pages and rewrote them to

be a lot more directive. For example, I

made a page with the agenda for an up-

coming meeting and asked people to

add to it. Second, wikis have to be

clearly better than other ways of collab-

orating. There have to be uses [for them]

that demonstrate their power. One of

these uses came prior to a special senior

management meeting where we could

bring questions from our groups and get

them answered. I put up a page…asking

my [team members] what questions

they wanted me to ask on their behalf.

People used the page to post questions,

edit them, and discuss which ones were

the most important and why. That really

accelerated wiki use. Finally, old habits

are hard to break. The tendency is for

people to keep using e-mail because

that’s what they know….I have to [tell

them], ‘I’m not reading e-mails on this

topic. Use the wiki’ or ‘Everyone’s assign-

ments are on this page – use the same

page to report on progress.’”

Interestingly, EIT’s exploitation is

often easier than its adoption. Since

the work of imposing new processes is

done by this stage, the manager’s task

is to leverage already standardized data

and work flows. Few employees and

managers have problems with that;

they’re eager to get the most out of a

system that was so much trouble to set

up. Exploiting EIT sometimes requires

adding a new FIT on top of it. In the

mid-1990s, food services giant Sysco im-

plemented an ERP system and data

warehouse across its 80 regional busi-

nesses. Sysco’s executives realized that

because all of the companies were now

recording orders in the same way, it

was possible to analyze the standard-

ized data to answer two questions:

Which customers were most likely to

defect? and What other products could

it be selling to existing customers? Sysco

invested in business intelligence soft-

ware, which sits on top of the ERP sys-

tem, extracts data from it, and facilitates

its analysis. As a result, salespeople and

managers gained something akin to a

crystal ball that could provide two criti-

cal answers they needed.

Other companies have exploited en-

terprise systems by extending them to

customers, suppliers, and joint-venture

partners. That expands businesses’ mon-

itoring capabilities and provides levels

of control that they could otherwise

have achieved only by employing more

people. For instance, the $107 million

Argentine grain producer Los Grobo

uses an EIT system to track all the work

done on its farms. Los Grobo rents most

of the fields, and contractors plant,

spray, harvest, and oversee them. The

contractors enter their activities into

Los Grobo’s system through a Web inter-

face, which allows managers and spe-

cialists at the company’s Buenos Aires

headquarters to make informed deci-

sions about land management and yield

improvement. This platform has helped

Los Grobo grow its sales at a rate of 40%

per year since 2000 – without buying

november 2006

M a s t e r i n g t h e T h r e e W o r l d s o f I n f o r m a t i o n Te c h n o l o g y • T O O L K I T

more land or hiring as many employees

as it used to.

• • •

For a resource to have an impact on a

company’s competitive position, it must

be valuable, rare, inimitable, and non-

substitutable. Oil wells and diamond

mines meet the test; pencils and paper

don’t. What about IT? At first glance, it

would seem that all three IT categories

fail to meet these criteria. Vendors offer

a wide range of FIT, NIT, and EIT, so

these technologies are not rare and

seem to be highly imitable. However,

people often forget that while the soft-

ware itself might not be any of those

things, a successfully implemented sys-

tem isn’t easy to replicate. Because of

the managerial challenges inherent in

its implementation, IT meets all four cri-

teria when a company succeeds in ap-

plying a technology and, consequently,

gains valuable capabilities.

Reprint R0611J

To order, see page 159.

Harvard Business Review Notice of Use Restrictions, May 2009

Harvard Business Review and Harvard Business Publishing Newsletter content on EBSCOhost is licensed for

the private individual use of authorized EBSCOhost users. It is not intended for use as assigned course material

in academic institutions nor as corporate learning or training materials in businesses. Academic licensees may

not use this content in electronic reserves, electronic course packs, persistent linking from syllabi or by any

other means of incorporating the content into course resources. Business licensees may not host this content on

learning management systems or use persistent linking or other means to incorporate the content into learning

management systems. Harvard Business Publishing will be pleased to grant permission to make this content

available through such means. For rates and permission, contact [email protected].