1. Real Strategies for Virtual Organizing
33
ll. Venkatraman - John C. Henderson
Harmony among
three vectors —
customer interac-
tion, asset sourc-
ing, and knowl-
edge lei'erage —
and a strong IT
platform form
the strategy and
structure of a
business model
for the knowl-
edge economy.
N.VenkatramBn is the David
J.McGrath,Jr.Prolesior
of Management and John
Henderson is professor
ol maoagement, botfi at
Dostoo University School of
IVIanagemenl. In addition, N.
Venkatramao is principal and
John Henderson is Ihe direc-
tor at the Boston University
Systems Research Center.
As the pos-sihilities of the information rev-
olution chalienKe tr.iditional business
logic, conipanie.s :ire experimenting^ with
a wide :irr;iy tif strategic aUernati es and
organizational forms. The appropriateness
of tlie current business model rooted in
the industrial ectmomy is c]iiestionable.
Drucker lias outlined liis views of a
knowledge-based organization. Quinn has
documented the shift toward a ser'ice-
iiased econotny with a focus on intellect,
llamel and I'rahalad argue for a critical
PcK'us on core competencies and an orga-
nizational design tliat best leverages them.
Womack and Jones advocate a lean orga-
niz:uion. anci Mandy painis a shamrock
structure.' We could cite many more opin-
ions, but the message is cleiir: the current
models of strategy and stnicture are woe-
fully inadequate to meet the imminent
challenges of the information age.
During the pa.st Xo years, we imdertook
a systematic study to conceptualize the
architecture of 'irtual organizing. Here,
we present our views on the architecture
of the twenty-first century business model.
W'e choose the term architecture rather
pLiiposefully and define it as "providing a
framework for the conduct of life, not a
spec ification of what life should be.
Architecture shouki facilitate, guide, and
provide a context; it should not provide a
rigid blueprint for conduct."- Moreover,
Sloan Management Review
Fall I9SB
Venkatraman • Henderson

2. the "building sluuikl preferably be ahead of its time
when planned so that it will l5e in keeping with tlie
times as Iong as it stands."'
We reject a virtual organization as a distinct structure
(like functional, divisional, or matri.). Instead, we
treat virtualness as a strategic characteristic applicable
to every organization; our discussion then is applica-
ble to centur)'-old companies tliLit manufacture
cement, chemicals, and auios as well as to new
34 entrants in the fa.st-changing high-technology market-
place. We view virtualne.ss as a strategy that reflects
three distinct yet interdependent vectors:
• The customer interaction vector (virtual encounter)
deals with the new challenges and opportunities for
company-to-customer interactions. IT now allows cus-
tomers to remotely experience products and serx'ices.
actively participate in dynamic customization, and
create mutually reinforcing customer communities.
• The assel coufigttratiou vector (virtual sourcing)
focuses on firms' rec|uirements to be virtually inte-
gnited in a business network, in sharp contrast to the
vertically integrated model of the industrial economy.
I'irms using the Internet for busine.ss-to-business
transactions can structure and manage a dynamic
portfolio of relationships to assemble and coordinate
the required assets for delivering 'a!ue to customers.
• The kiioivk'cl^c k>ix'ra}>i' ector (virtual expertise) is
concerned with the opportunities for leveraging
diverse s<)urce.s of expertise within and across organi-
zational boundaries, IT now enables knowledge and
expertise to become dri ers of 'alue t reation and
organizational effectiveness.
No one vector adec|uately captures the potential
opportunities of virtual organizing; their iiik'rde/K'ii-
dence creates the new business model. We view vir-
tual organizing as a strategic approach that is singu-
larly focused on creating, nurturing, and deploying
key intellectual and knowledge assets while sourc-
ing tangible, physical a.ssets in a complex network
of relationships. We depart from the ciuTent litera-
ture on irtual organizing that proposes incremental
improvements to the biisiness logic rooted in the
industrial age. We develop our logic of virtual orga-
nizing by placing IT at the center. The powerful
convergence of compLiters and communication tecli-
nology and the emergence of the Internet are
enablers of this new business niodel. Nadler and his
colleagues indicated that the creation of effecti'e
architecture hinges on the use of structural materials
capable of implementing the architecture and dis-
cussed IT'S power in creating future organizational
Figure 1
Virtual Organizing: Three Vectors and Three Stages
Vectors and
Characteristics
Stage 1 Stage 2 Stage 3
Customer Interaction
(Virtual Encounter)
Remote experience of
products and services
Dynamic
customization
Customer communities
Asset Configuration
(Virtual Sourcing)
Sourcing modules Process
interdependence
Resource coalitions
Knowledge Leverage
(Virtual Expertise)
Work-unit expertise Corporate asset Professional community
expertise
Target Locus Task units Organization Inter-organization
Performance Ohjectives Improved operating
efficiency (ROD
Enhanced economic
value added (EVA)
Sustained innovation
and growth (MVA)
Venkatraman • Henderton Sloan Management Review

3. arcliilectuiv.' In n siiiiikir :iy. we helieve that the
emerging; urcliitcctiirc of virtual organizing i.s not
po.s.sihle, or constructed elTecUvely, without the .sig-
nificant power of IT.
The emerging global, digital economy
allows for establishing and leveraging a
two-way information link between a com-
pany and its customers — albeit through
remote and asynchronous mechanisms.
Each vector of our model ha.s three distinct stages
(see /'i,i>itrc' I). Stage one focuses on the task units
(such as customer ser'ice, purcliasing. or new prod-
uct developnieni). Stage two ibcLi.ses at the organiza-
tional level on how to coordinate activities to create
superior economic value. The third stage focu.ses on
the interoigunizational network to design and leer-
uge multiple interdependent commiuiiiies for innova-
tion and growth.
Tlie.se ihree vectors ha'e tratlitionally heen indepen-
dent: they focused on isolated functions — market-
ing, purchasing, and human re.sources, respectively,
with their idiosyncratic processes and information
.systems. For instance, marketing acti ities in the cus-
tomer interaction vector were supported hy telepho-
ny, call centers, and product simulations. Sourcing or
purchasing acti ities were siipporred l)y electronic
data intercliange (F.DI) and CAD/CAM integration
hetween suppliers and manufacturing processes.
Decision .support .systems and groupware supported
managerial work. There was historicalh' no common
Linilying platform to pull ihese ditlerem activities
together. Hoeer, the increased adoption of enter-
prise systems like SAP. Oracle, Baan. anti Peoplesoft.
comhined with the rapid acceptance of the Internet
protocols, offers the po.ssihilit)- of a common teclinol-
ogy platform.
Our view of virtual organizing integrates these three
hitherto separate threads into an interoperable IT
platform that supports and shapes the new husiness
model. The cliallenge is to ensure internal consisten-
cy across tlie three vectors and benchmark a profile
of viitualnc'ss relative to competitors and referent
companies in the marketplace. External benchmark-
ing hecomes critical as companies experiment with
differenl approaches to design.
Customer Interaction (Virtual Encounter)
The interactions of a company with its customers
(and end consumers) in ihe industrial age occiured
through a multistage distribuiion network involving
wliolesalers. retailers, customer sef'ice agents, and
franchisees. The predominant focus was on efficiently
distributing products in a linear fashion IVom manii-
factiuers to con.sumers. The emerging global, digital
economy allows for eslahlishing and leveraging a
two-way information link hetween a company and its
customers — albeit through remote and asynchro-
nous mechanisms.^ This link is possible not only for
constimer products and ser'ices hut also for htisi-
ne.s,s-to-husiness products and sen'ice.s.
Some customers are interacting with companies in
new, inntnative ways. They are testing procUicts al
early concept stages (for example, software and
advertising), while others are receiving e-mail replies
to prohlems not coered in product manuals (com-
puters, electronic equipment, and industrial prod-
ucts). Technologically savvy customers are constaict-
ing customized news feeds (www.snap.com), review-
ing and monitoring financial accounts (American
Express, Schwab, or Fidelity), exploring real-estate
properties (w^vw.centuiy21.com). or reviewing video
news clips (www.cnn.com). Moreover. 'irtual con-
sumer communities — which alter the traditional con-
sumer role — are emerging.''
Corporations face new strategic c]uestions: What are
the tenets of marketing for the information age? How
can we leverage the fiinctionalily of customer inlerac-
tions in crafting sLiccessfLil marketing strategy? What
is advertising s role under conditions of greater cus-
lomer pull and fine-grained segmentation? Should we
actively create customer communities or participate in
them unc)htrusi'ely? To address these and related
c|Liestions. we delineate three stages of customer
interaction: remote experience of products and scr-
ices, dynamic customization, and cListomer commu-
nities."
Remote Experience of Products and Services
To imdersiand this stage, let us examine Sears,
Roehuck's introduction of product catalogs, the first
successful attempt at creating a virtual product expe-
rience. It simulated key product features from a phys-
35
Sloan Management Review
Fall 1998
Venliatraman • HendeisDn

4. ic-al space (storefront) to another physical space
(paper), separated by time and distance. Custoniers
experienced the different products through the pages
of a catalog and made their purchase without actually
.seeing or touching the product. Over the years, the
catalogs Ixjcame glossier, mail-order sales increased,
and direct marketing became a major force. The infor-
mation age has created the home shopping television
network: consumers virtually experience product fea-
tures — signaling a shift from atoms to bits, in
36 Negroponte's words." 'W has allowed a more complex,
intricate demoastration of features and functionality.
Remote and continuous links witb cus-
tomers become critical as the concept of
brand identity and brand equity are rede-
fined.
Similarly, the concept of remotely experiencing cus-
tomer service is also not new hut is heing redefined.
Since the early 1980s, building supervisors have had
superior customer service from Otis Elevators' remote
elevator monitoring. Schliimherger remotely reads
utility meters; Sony provides PC owners with diag-
nostics and repair over ordinary telephone lines. GE
Medical Systems remotely supports lah technicians in
hospitals dealing with medical images, and GE
Aircraft Engines remotely tracks the performance of
aircraft engines during flight.
More recently, the Internet has accelerated and
redefined the possihilities of the remote product or
service experience, for example, the real-time mon-
itoring of shipments (www.fedex.com; www.ups.com)
or stock portfolios (www.yahoo.com). Airlines are
po.sting fares only for their freciuent flyer memhers
through their Weh links (see, for instance www.usair-
ways.com.) and hypa.ssing traditional travel agency
channels. Retail Iianks are creating high-tech links
with customers (www.weIlsfargo,com/home; www.
citihank.com); their challenge is to ensure that the
sites complement the .service delivery at their hanks.
The possibility of remotely experiencing products or
services ha.s major implications for developing the
required business Infrastmcture. Charles Schwah
redefined the brokerage husiness with a diiferent
price-point than full-service hrokers and offered a
portfolio of channels for customer service (1-800
numhers, hranch offices, lelehrokers, and e-Schwah).
Now, new entrants like E'Trade (www.etrade.com)
and Ameritrade (www.ameritrade.com), which do not
have a physical presence, further threaten Schvvah's
pricing staicturc. The challenge for companies in infor-
mation-intensive markets is to manage the velocity of
the shift from physical to electronic infrast mettires anci
compete effectively against those entrants that don't
Iiave llie constraints of a physical infrastmcture.
The same challenge is driving software distrihiitors,
such as Egghead, to reassess their di.stribution strate-
gy: during the past year. Egghead has closed more
than seventy traditional software outlets and has
opened its first store on the Net. Its on-line storefront
(vvwvv.egghead.com) features three main rooms —
Internet products, busine.ss products, and games —
with on-line product support. It is entirely reiinqiiish-
ing its pre.sence in the physical marketplace to focus
on the electronic marketspace.
The passive distrihution of standardized news
throLigh traditional channels is also being redefined.
Leading newspapers and magazines routinely allow
readers to hrowse their late.st issues on the Web:
Fi)iciucicil Times (www.ft.com), Comfntlenvorld
(vvww.compiitervvorld.com), /''oilnite (www.pathfind-
er.com), Vl//rc'f/(www.wired.com), and 'Ihe h'conontist
(www.economist.com). Nearly all are facilitating
interactions hetween editors or writers and readers.
Traditional selling of textbooks at the heginning of a
.semester is now relegated to an educational consul-
tant or a value-adding partner throughout the .semes-
ter. Leading textbook publishers are creating sup-
plementary Weh site links to connect students and
profe.ssors throughout a course (www.mhh.coni;
www.wiley.com; www.uol.com).
Product Companies. Clearly, consumer product com-
panies like Procter & Gamhle, Colgate, Kraft Foods,
and consumer durables and husiness-to-busincss
companies like General Motors, General Electric,
Allied Signal, and Caterpillar have crafted their husi-
ness strategies by leveraging physical assets and
developing powerful global hrands supported hy
ma.ss advertising and mass distrihution. But remote
links with customers apply equally well to these
companies. Remote and continuous links witli cus-
tomers hecome critical as the concepts of lirand iden-
tity and hrand equity are redefined.
Hen(tersan Sloan Management Review
Fall 1998

5. Kraft Intemclive Kitchen (www.kraftfoods.com) i.s an
example of :i consumer products company keeping in
touch with its consumers liy providing information-
based sen'ices like meal planners, recipes, lips, anc!
cooking tecliniques. Kraft's intent is to have remote con-
nection.s and interactions with consunicrs in new ways.
General Motors and its OnStar service use.s global
positioning system satellite technology and a hands-
free, voice-activated cellular phone to link the driver
and the vehicle with the OnStar Center. Advisers pro-
vide real-ti[ue, person-to-person services such as
emergency help linked througli automatic detection
of air-liag deployment, stolen vehicle tracking, route
support, remote diagnostics, remote door unlocking,
and a variety of travel and personal services. GM can
provide information-based services to the driver and
even remotely diagnose tlie car continuously for a
fixed service fee (www.onsuir.com).
Kvery company should assess how its products and
services can be experienced virtually in the new mar-
keting infrastmcture. A Wel) site is essential. The
question is liow best to use the Web's power to cre-
ate superior linkages with custoniers.
Dynamic Customization of Product and Services
The second stage ofthe customer interaction vector
f(K'uses on the opportunities and challenges in dynam-
ically customizing products anci services. Competitive
markets are rapidly eroding margins due to price-
based competition, and companies are seeking to
enhance margins through customized ofTerings." Our
view of dynamic customization is Ixised on three prin-
ciples: modularity, intelligence, and organization.
Modularity is an approach for organizing complex
products and processes efficiently.'" Intelligence
througli continuous information exchange with con-
sumers allows companies to create products and
processes using the lx.'st possible modules. More
importantly, dynamic customization of products and
.services requires an organization that is fundamentally
committed to operating in this new way.
Modularity. Product or service modularity reciuires
the partitioning of a task into independent modules
that function as a whole within an overall architec-
ture." The cla.ssic example of a modular product is
IBM's System/3f)0. The concept has heen extended to
other industries, l-or instance, in a.s.semhl!ng cars,
Toyota, BMW, Mercedes-Benz, GM, Porcl, and
Chrysler create product platforms that allow modular
Intelligent sites learn their visitors'
tastes and deliver dynamic personalized
information about products and services.
reuse. The benefits are not only lower unit costs but
al.so greater customer satisfaction; companies can
configure automobiles to customer requirements.
Modularity also works for services, especially infor-
mation-intensive services. Take, for instance, personal
news feeds (www.cnn.com; www.cnet.com) and cu.s-
tomized stock ciuote.s and tracking (www.schwab.com).
It is possible to reuse various modules of news and
inforiuation Eo construct a customized news service at
CNN Interactive anci a customized, interactive edition
of I'he Wall Street Jottntcii.
In place of .standard textbooks, instructors now assem-
ble textlxx)ks that suit their pedagogical style and
objectives liy .selecting modules from different source.*;.
McGraw-Hill's College Division pioneered this concept
through its I'rimis offering. Primis users can create a
custom textbook from modules within the Primis data-
ba.se and a.s.sociated tiataba.ses. Many other textbook
companies offer similar configurations.
Intelligence. Dynamic customiz^ation is rooted In the
deployment of intelligent agent .software, such as
Firefiy (www.firefly .com). Firefly Passport is available
to end users who rank Web sites as they surf the
Internet. Firelly uses the data lo create consumer pro-
files. Based on an automated collabonitive filtering
process, Wei") site operators can match users with simi-
lar profiles and make recommendations based on their
shared interests. The automated collabonitive filtering
process lets u.sers receive real-time, personalized list-
ings for items in a site's catalog of products, services,
or content. The result is intelligent sites that learn their
visitors' tastes and deliver dynamic personalized infor-
mation aboul products and services. As the Web
becomes the main marketing infrastaicture and as
intelligent agents such as Firefly Passport gain more
sustained use on the Web, there will be greater oppor-
tunities and pressures Ibr dynamic cu.stomization.
A primitive application of automated collaborative fil-
tering appears on sites such as Yahoo's personalized
site (my.yahoo.com), a movie recommendation site
(www.filmlmders.com), and Barnes and Noble's on-
line book.store (www.barnesandnoble.com).
37
Sloan Managemeni Review
FaU 1998
Venkatraman • Hendenon

6. The most profound aspect of interaction
in the virtual model is the emergence of
electronic customer communities.
Dynamic cii.stomization is po.ssihle .since these Web
.sites, .suppcjrtL'd hy ugL-nts. levfiagf in real timt*
each users experiences to guide others tt) rele'ant
38 content and products. The sites incorporate their
entire user ba.se. not ju.st Ilie thoughts of a lew
experts. Indeed. e'eiy .search engine is trying to dif-
ferentiate itself llirough its .suj-Jcrior ability to create
custom profiles.
Organization. Ultiinutely, niodularit)- and intelli-
gence are useless unless the organization de.sign is
geared to deliver products and serxices on a
dynamic, adaptive hasis. Organizaiions need to
change how they look at marketing processes,
shifting from an inside-out perspective to an out-
side-in perspective. For example, in personal com-
puters, Dell is clearly leading with its dynamic cus-
lomization offering. Dell s succe.ss relative to others
within the computer industry is attrihutahle to its
ahiiiiy to develop an entirely differeni busine.ss
model on the concept of building to order.'-
Similarly, the challenge for McOraw-IIill's Primis
division is to rnoe from textbook modularization
and go on to create an organization that can deliv-
er educational solutions for customers.
The challenge of dynamic customization rests on
greater decomposition of products and services into
modules that can be combined to deliver increased
functionality. The numher of Web transactions will
rapidly increase as electronic commerce goes main-
stream. Modules will be redefined continuously, and
more information will be captured as consumers
become comforiable with intelligent agents ihai work
for them. Over time, consumers will expect enhanced
customization for fair value in reunn for aggregators
consolidating detailed personal information.
Marketers will focus on as.seiiibling for modules that
deliver cu.stomized solutions.
Customer Communities
The most profound aspect of interaction in the virtual
model is the emergence of electronic cu.stomer com-
muniiies. The.se coninumilies signal a power sliih
from manutaciiirers to ciistoiiiers: the communities
are information-gathering and information-disseminat-
ing conduits. Pie iously, in the industrial economy,
consumers could not he effectively linked togetlier
acro.ss lime and space.
Hagel and Armstrong offer live defining characteris-
tics of virtual communities: distinctive focus, capacity
to post their content for access to the wider conimii-
nily, appreciation of member-generated content,
acce.ss to competing offerings, and commercial orien-
tations. They state. AVe cannot yet point to a single
e.xample of a virtual community that incorporates all
live of the defining cliaracterisiics."" We contend that
customer communities can exist with only the firsl
three characteristics. The lasl two characteristics —
competing offerings and commercial orientations —
apply lo only a subset (jf comnuinities.
Some customer communities refiect strong commit-
ment to brands. Harley-Daicison*s site (www.hadey-
davicison.coin) houses the Ilarley Owners Group
(IIOG) — a Weh siie for Harley owners to share their
stories and pictures. This community does not overtly
create new product sales liLii does maintain an
important link wiih core customers und enhance
brand identity.
Citibank — in partnership with The Mining Company
— has created a community on its Web site Ibr cus-
tomers to learn about Citibank's products and ser-
vices. It uses bulletin hoards, chat rooms, e-mail, and
oiher features so prospecti'e customers can interact
with current customers and receive testimonials
directly from them (www,citibank.com; WT'w,the-
miningco.com).
The Intel Pentium chip case exemplifies the role and
power of customer communities een when an orga-
nization does not .sell its products directly to llie end
user. The light-knit user comnnmity had access lo
information about poieniial problems with Intel's
Pentium chii> and forced Intel lo act. The community,
supponcci hy Usenet groups, made Intel rethink its
heiuistics for product recall. Intel nov.- has a news-
group feature within its site (newsgroups,intel.com)
and is activcK- participating in its customer group.
The source of the group s power came from knowl-
edge (about the performance parameters of the
chips), not from scale and volume of purchase,
I'nlike buying cooperatives. indi'iduals wiihin a cus-
tomer coninuinity retain their rights about brand pref-
erences and purchase decisions.
Venbtiaiiian • Hendeison Sloan ManagemenI Review
Fall 1993

7. Amazon.com (www.ama/.on.com) is often touted as
•An example of how to lewruge the power of cus-
tomer communities. .Amazon U:s created a 'irtual
bookstore with minimal physical assels and a partiCLi-
lar focus on customer relationships. Will Amazon
fatle away as other serious competitors like Barnes
and Noble (www.barnesandnoble.com). Borders
(www.Borders.com). and Microsoft (ww.books.com)
enter the fray? As long as ,-niazon is able to orches-
trate a tight-knit amimunily of avitl readers, who cre-
ate and di.stribute their own content (in the form of
re'iews). competitors will h:i'e difficulty dislodging
and disintegrating this comniunity. The challenge for
Amazon.com is to provide balanied rex'iews rather
than only f:i"orable ones.
Consumer communities are in the early stages.'"
Nevertheless, they could cxerci.sc significantly greater
power in the future and transform the role of market-
ing. .Should marketers orchestrate customer communi-
ties for their own products or invohx- a third party?
Companies such as Intel. Marley Oavid.son. I-gghead
Software (www .eggheacl.com). Tnuelocity (www.trav-
elocity.com). Toyota (owners.toyota.com'ent ranee.html)
and Apple (lompuler (www.apple.com/usergroups)
are forming customer communities around iheir prod-
ucts and brands.
At the same lime, some communities form without the
involvement of major .sellers. I'or instance, photogra-
phy (www.photoshopper.com). automobiles
(www.autoweb.com; www .cdmunds.com. www .net-
market.com), and .some specialized areas like water
utility (www.waleronline.com) and air pollution
(www.pollutiononline.com) gain credibility largely due
to their lack of ownership links to product and ser'ice
provitiers. These communities must generate and
maintain ihe consumers' tru.st as they collect personal
information while provitling aiue-added senices.
As 'irtual organizing becomes more widespread.
companies must recogni/e communities as part of the
value deiiveiy system and re.spond appropriately in
their strategies.
Questions for Managers
• Do you have a .strategy to virtually connect with
your customers that Incuses on capturing information
and Icv'eraging knowledge? 'Hiis contrasts sharply to
.setting up a distribution .sy.stem that focuses on effi-
ciently tlistributing pi'oducts anti scr'ic'es.
• Have -ou dex'elopcd apjiropriate niechanism.s so
customers can reach you twenty-four hours LI day,
seven clays a week? Can customers experience salient
product features remotely? Can cu.stomtrrs gain access
to a databank of answers to frecjuently asked ques-
tions so they can solve some problems themselves?
More importantly, how does your remote access
capaliility rank against your competitors and other
companies that may do business with your cus-
tomers?
• How extensive is your capability to cu.stomize your
products and services over their life cycles? While it
is becoming relatively easier to customize the product
or ser'ice at creation or delivery', the challenge is to
dynamically link to the customers to refine features
o'er time.
• Does your organization still reflect a make-to-sell
orientation rather than a make-to-order philosophy?
The latter is more in line with a sense-and-respond
approach to virtual encounters.
• Do you plan to be a passive participant or an
active orche.strator in dealing with emerging cu.stomer
communities?
• How are you assessing your progress in the cus-
tomer interaction vector as the marketplace demands
greater remote access, dynamic customization, and
participation in the customer community?
Asset Configuration (Virtual Sourcing)
The second vector focuses on accjuiring critical assets
and resources, a clear move away from vertical inte-
gration toward greater reliance on components
obtained from external markets. As they move from
an industrial economy, many corporLitions will con-
centrate on creating and deploying intellectual and
intangible assets while sourcing tangible, physical
assets from a complex business network.
Iffective contracting for complementar' capabilities
through a network of suppliers nnd .subcontractors is
a characteristic of virtual organizing.'' Davidow and
Malone note: "For a virtual corporation to succeed, it
mu.st be closely linked with its suppliers as to create
a shared destiny. . . . Ultimately, even the boundaries
between them will become indistinct."" Goldman et
ai. highlight the importance ofclo.se relationships
with suppliers anci subcontractors to realize efficiency
and flexibility — critical indicators of organizational
agility.'" Quinn describes Nike as a model for effec-
tive sourcing of manufacturing competencies: "Nike is
basically a research, design, and marketing ctjmpany
—• out.sourcing 100 percent of its athletic footwear
39
Sloan Management Review Venkatraman • Hendetson

8. manufacturing to numerou.s production partners'
abroad."'"
As corporations refocus on their core competencies
and move away from vertical integration, they will
obtain coniplementar' as.sets through interfirm rela-
tionships.'" For example, confeetioner' companies
like Ne.stle, Mars, and others get their packaging from
leading packaging and printing companies. Nike and
Reebok get their shoes, based on their proprietary
40 designs, from leaciing Asian contractors, Chrysler's
turnaround is often credited to its .sourcing capabili-
ties, reflected in its relatively higher percentage of
conlracted parts. Dell's success is attributed lo ils
superior .sourcing strategy. Hoeing developed its latest
plane, Boeing 777. with a portfolio of relationships
among subcontnictors and lead cu.stomers.
Nearly every m:ijor treati.se on the logic of new .stnile-
gy and competition di.scii.sses the requirements of
effective sourcing of competencies.-" However, this is
not u simple choice between niake-ersus-buy or ver-
tical integration but involves a continuous reconfigu-
mtion of critical capabilities asseinliled through differ-
ent relationships in the business network.'' IT allows
the po.ssibility of efficient .sourcing of standard mod-
tiles and creates opportunities for process outsourc-
ing. There are three stages of as.set reconfiguration —
sourcing modules, process interdependence, and
resource coalitions.
Sourcing Modules
The first stage in this ector deals with the benefits of
efficiently sourcing standard module.s or componenis.
Advances in the industrial age were based on prod-
uct modularity — building a complex product from
subsystems designed independently yet fimciioning
us a seamle.ss, integrated entity. GM. Ford, and
C;hrysler operate more like assemblers and integrators
of subsy.stems than traditional vertically integrated
manufacturers. Baldwin and Clark argue ihal in a
new age of modularity, the value-adding role of a
corporation is less in the manufactLire of a critical
component than in the creation of a product or ser-
vice architecture.--' Thus the designer assembles multi-
ple interlocking produci modules to deliver a superi-
or .solution, while controlling the architecture of the
subsystem and its role in delivering value.
For more than a decade, the ptjwer of electronic data
interchange (HDD has sup[xirted the ability lo effi-
ciently source modLilai' products. Wal-Mart pioneered
a new retailing tnocle! by leveraging a sophisticated
EDI network wiih its suppliers to radically lower ils
invenioiy le'el. In Negroponle's words. Wiil-Mart sub-
stituted bits (information) for atoms (inventory) to
enhance its operating margins to a level previously
unseen.-' F-DI has had a significant role in reducing
inventory levels in .supply chains within con.suiner
products (Procter & Gamble. Kraft General Foods),
athletic shoes (Nike, Reebok), apparel (Gap.
Benetton), aircraft puns (Boeing, GF), and computers
(Dell. Gateway).
The Web is pushing the ability to source standard
product modules even further. Cisco Sy.stems Inc.. for
example, expects to .sell SI.8 billion worth of Internet
products. Boise Cascade Office Products (www.
bcop.coni) has deployed an extranet for its largest 600
corporate customers, anci its margins haw doubled
cc^mpared to its traditional openitions. Dell is .selling
more than $3 billion in computers over the Web.
mainly to corporate customers like Shell anci Boeing.
Geneinl Electric"s Trading Process Network (TPN)
started as a .ser'ice to streamiine GE's procurement of
.standard products within GK Lighting. It has now
expanded to cover a broader range of business units.
The network links more than 2.SOU GF trading part-
ners and accounts for more than SI billion of pro-
curement in 1997. By initial estimates. TPN has
reduced the procurement-cycle time by SO percent,
procurement-process costs by 30 percent, and actual
material costs by 20 percent. GH is expected to
soLirce more than S5 billion of procurement annually
through TPN by 2000.
Sourcing Logic. Serices like GF.'s TPN force man-
agers to constantly ask: What assels can we obtain
from outside without loss of competitive advantage?
When should we revise our sourcing logic?
The selection of a.ssets is complicated because their
criticality changes over time. In the 19H()s. IBM con-
ceptualized the personal c-oiiipiiler market around ils
proprietaiy architecture. It purchased operating sys-
tems from .Microsoft (MS-DOS and OS/2) and
microchips from Intel. However, as the technology
architecture evolved. IBM was left withoui a critical
asset in the clianged marketplace. The market shifted
from :i hardware standard that favored IBM to a new
standard based on .software (Windows) connected to
chips (Intel). So, while IBM's original sourcing deci-
sion might have been efficient and prudeni, il could
Venkatramaa • Hendeison Sloan Management Review

9. not adapt ils sourcing logic to the changing market
conditions.-'
Process Interdependence
The next stage focuses on the interdependence of
busine.ss proces.ses across organiziUionai boundaries.
External specialists can CLirn" OLII informution-
intensi*e inisiness jirocesse.s witiiout ioss oi controi.
During tile past few years. se'erai speciaiisi business
process firms liae emerged in tiie areas of account-
ing. inentoiy controi, customer senice. cali-centers.
dataiiase anaiysis. teiemarketing, anci itjgistics.
Business process outsourcing wiii continue to
increase as more speciali.st iirni.s emerge in the recon-
figured business network. I-or exampie:
• DirecTV, which delivers aiiout ISO channels of
sateliite television programs, is working witli
MATRIXX. a iuisiness unit of Cincinnati Beli. to deiiv-
er superior customer ser ice. From its dedicated facii-
ity in Suit Like City, L'taii, MATRIXX provides tele-
phone customer ser'iee and sales support, wliicli
includes ecjuipnient deaier referrais. programming
sign-up, and compiete customer .senice and account
management. MATRIXX is tlie extension of DirecTV
operations iiy providing more than 2.000 dedicated
customer service representatives wlio iiandie about
20 mliiion caiis a year. Tiiis marks tlie first time a
major corporation has entrusted its entire customer
ser"ice operation to an outside telephone marketing
company on such a grand scLile. DirecTV iiiis inte-
grated its processes with MATRIXX. thereby leverag-
ing its expertise at managing customer reiLitionsiiips.
• Aiiegiance Corjioration recently spun off from
Baxter Healthcare. The company, known for its leg-
endary ASAP system, created ;i separate eniiiy to
focus on logistics and malerials management process-
es.-* In Miircii 1996. Baxter announced an unprece-
dented agreement to work with Methodist Healthcare
System of San Antonio, a joint-venture partner with
Columbia Ilealtiicare System. L'nder the agreement.
Baxter operates a ser-ice center to support the con-
solidation and outsourcing of seiected funclions witii-
in the Metiiodist Heaithcare System and to implement
process iinproements to reduce costs. More impor-
tantly. Metiiodist is Ie eraging Baxter's expertise in
three proces.ses: distribution and logistics, operations
efficiencies, and resource management. Furthermore.
Baxter's compensation is directly tied to the perfor-
mance ie'eis of these processes. Aiiegiance
Corporation's stock price has more than doulileci
since the spin-off, anci the firm has a market capital-
ization of mcjre than $2 billion.
• Kraft Foods is iiiterlwining its marketing processes
with the data ccjilectlon anci analysis at ACNieisen. a
ieading information provider in tiie consumer pack-
aged gotjcis industiy. The early avaiiaiiiiity of market-
ing data tiirough process integration i'>etween tiie two
organizations aiiows Kraft Foods lo respond to mar-
keting trends c|uicker anci more effectively than its
competitors. Kraft has benefited by reconfiguring its
marketing processes across organizational boundaries.
• National Semicoiiciuctor (NSC), tiie tiiirtcentli
iargest computer-chip maker, entereci an agreement
ill whicii FedEx manages its iogistics operations. In
tiie eariy 1990s, NSC reaiized that a major driver of its
operationai inefficiency was its outmoded process for
mo*ing in'entory. Since NSC could not aciiie'e the
iiest-in-ciass process ievels of a iogistics company
without taking vaiuable resources away from product
design and manufacture, it outsourced its processes
to FedEx. NSC. tiirougii its iink with I'edFx. has
iiiipro'eci its process performance: it nioes products
from facton- to customer in about four days. ;md its
distriijution costs have failen from 2.6 percent of rev-
enues to 1.9 percent.
• KedEx. on tiie otiier iuuici. has iiecoiiie a j")iuyer in
tile eiectronics commerce anci iogistics maiketfiiace
ij' piu-sicLiliy mo'ing goods from far-Hung manufac-
turers to their gloixii customers. Its new ser'ice.
Virtual Orcier. takes and processes orders, inciuding
arrangements for siiipping. provides access to ship-
ment status, and iuinciies customer c|ueries. Indeed.
its goal is to become the airline of the Internet. It has
combined electronic ct)mmerce and logistics opeiii-
tlons and accjuired Caiii)er Logistics to ensure a f'uil
range of sei-vice capai>iiities.
l^rocess reengineering captured managers* imagina-
tion in the 19HOs and early 1990s for one rea.son:
inipro-ed operating margins. Now. another shift.
process outsourcing, is "the deiegation of one or
more business [processes to an external provider wiio
tiien owns, manages, and administers die seiected
processes based on measurable metrics."-" This is
attractive because of greater asset utilization. As
processes become more standarciized and as tiie mar-
ket matures witii more staiiie participants, many cor-
porations wiil recognize the criticality of business
process outsourcing. Specialized firms like MATRIXX
41
Sloan Managemenl Review
Fall 19gB
Venkatraman • Headenon

10. and FedF.x can seamlessly carry out customer service
anci iogistics: at tiie same time, extranets can ensure
liiat linns do not lose controi of their processes.
E'ery corporation siiouici assess tiie iieiieflts anci risks
of cariying out these processes themselves. Tiius IT is
fundamentally affecting the business scope ol most
organizations.-^
Resource Coalitions
The third stage in the asset configuration vector
42 focuses on the establishment of a resource network.
in wiiicli tiie firm is part of a 'ii5rant, dynamic net-
work of compiementaiy capaiiiiities. A corporation
becomes, not a conventional portfolio ot products or
businesses, but a portfolio of capabilities and rela-
tionsiiips. Every organization is either impiicitiy or
expiicitly positioned in a network of resources wiiere
it acc|uires compiementary capaiiiiities. Corporations
are increasingly relying on external sources not only
for support activities liut aiso for critical resources.
Positioning a firm within a iiroader netwcjrk of
resources in the marketplace is a driver of competi-
tive advantage. Consecjuently. tiie strategic ieadership
ciialienge is to orciiestrate an organization's position
in a ciynamic. fast-changing resource network. Three
examples heip iiiustrate the ix-nefits of re.source
coalitions:
• When viewed as a portfolio of capabilities through
relationships. Nike is a new breed of organization
that has positioned ilself slrategicaiiy within a net-
work of complementaiy resources. Nike a.ssemiiies
liie rec[uireci capabilities tiirougii a coaiition. inciud-
ing production suiicontractors in Asia, ad agencies
(Weiden-Kennedy), web suppcjrt (Vivid Publisiiing).
retaii outiets (Eootlocker). exciusive contracts witii
athietes (Michaei Jordan. Tiger Woods. Andre Agassi),
preferred contracts witii |)rofes.sionai teams (Daiias
Cowboys), and universities (Ohio State footixiii.
University of North Caroiina basketbaii). Nike's
adaniage is based c^n the superior orchestration of
its position in the resource coalition. As it.s competi-
tors strive lo construct similar coalitions to neutralize
Nike's acivantage, Nike's success wiii dei^enci on
proactively aciapting its network to responci to new
market rec|uirements.
• Charles Schwab has successfully eoleci from :in
undifferentiated discount broker it) a key playei"
through its offering, OneSource Schwab positioned
itself within a resource coalition composed of inde-
pendent financial acK'isers (who acl on behalf (jf indi-
vidual in'esiois). a broad array of financial product
companies (mutual funds, stocks, initial public offer-
ings, debit cards, anci iife insurance), provicieis of
researcii reports. ;iiici se'ei*Lil teciinoiogy flriTis (suji-
porting its e-sciiwaii seivice). This vigcjrous network
underpins Schwab's coiupetiti'e acKantage. Wliiie
inciividual eiements of the network can iie imitated,
Schwaii's aiiiiity to integrate tiie different fLicets
cohei'cntK is clistinctie. Ils future Micce.s.s wili be
fundamentally ba.sed on its ability to adapt this net-
work to ciianging market and competitive ccjdditions.
• As part of Amazon's attempt to create a powerful
iiusiness modei. ii is orchesti'aling afflliateci sites as
associates to sei-ve as extencied bookstores. The asso-
ciates establish their own book.stores on their Web
sites or place banners and k)gos to direct traffic to
Amazon's site, ln return, they receix'e a conimi.s.sion
on books purchased ihrough their relerrals. l-'or
instance, a search on AkaX'isia triggers an opportunity
to order books on the selected topic. Amazon's suc-
cess will be significantly affecled by ils ability lo cvn-
trally position ilself in the constantly reconfiguring
resource network.
Participation in Resource Coalitions. In the irtual
integration model, one firm does not dominate all
others in a network. 'I'luis es'eiy firm balances ils
leadership position relative h) one set of resources.
with secondary" roles related to olher c-omplemeiitary
resources. F'or instance. IBM is dominant in the main-
frame coiiipuler arc-hitec-[ure while it follows stan-
dards established in other areas (e.g., Vinck)ws-Inlel
architecture). Siiiiilariy. I-'ideiity distriixites its funds
tiirougii Sciiwab wiiile aggressi'eiy promoting its
own Funds Network as an alternative. Those compa-
nies tiiat can carefuiiy anaiyze tiieir relative ciepen-
dence on otiier piayers in tiieir resource coaiition and
ensure their unic|ue capabilities will be successful.
Eery organizalion is dejienclent on lelatioiisliips
for assembling complementaiy capabiliiies.
StarliLicks has eiiieiged as a major coffee retaiier
witii saies of more tiian SI biilion and a 6.3 per-
cent return on in'estment. Moiv impressi'e is its
sttjck market capitalization — more than SS billion
in June 1998. Instead of committing its internal
resources to build a brand image. Starbucks has
fbrmeci a iriual network of complementaiy players
into a coalition of I'niteci Airlines, .Marriott anci
Westiii Hoteis, Star Markel. Pepsi, and Barnes and
Noble, among otliers. In doing so. it has created a
Venkatraman • Henderson Sloan Management Review
Fall 1998

11. le'el of brand ecjuity far greater than its ievei of
resource depioyment.
Shifting Value Drivers. The importance anci reiati'e
priorit)" of resources ciianges o-er time. IB.M, Apple,
and Motorola pooled their resources to create the
PowerPC. However, they could not establish a iable
alternalix'e to Vindows95/NT anci aiiandoneci tiie
coaiition. Now tiiey are part of tiie resource network
dominated by Microsoft anci Intei. More recentiy,
IBM. Sun. anci Oracle formed a network to establish
the Network Computer (NO as an alternative to the
Wintei piatform.
Such coalitions highlight another impor-
tant characteristic of virtual organizing:
the hlurred distinction between competi-
tion and cooperation.
Sucii coaiitions iiigiilight another important charLicter-
istic of virtual cjiganizing: the blurred distinction
Iietween competition and cooperation. E'eiy compa-
ny is positioned witiiin A resource network and
simuiiancously plays both competitive and coopera-
tive roles. Sheli and Amoco iia'e pooied most of
their w-est Texas oii fields to make upstream opera-
tions more efficient, while competing witii eacii otiier
in rcftaiiing. BP and ARCO Aiaska iiave a cooperative
agreement on maintenance, operations, procurement
transpcjrt, and driiiing. while the' compete in other
areas of the value chain. Blurring boundaries are also
evident as Netscape. Sun. Microsoft. Oracie. Appie.
IB.M. Digital, Intei. Lind others simuitaneousiy tr' to
estaiiiisii superiority anci ciistinction on liie Internet
wiiiie recognizing tiie need for interoperaiile technoi-
ogy architecture. Knowing where and how to add
value is important tt) strategy deveiopment in tiiis
new game of co-opetition (tiie iogic of comi^ining
cooperation and competition).*
Creating a resource coaiition anci nax'igating under
fast-ciianging market conditions is not simpie. Riiie
makers in a resource coaiition can iose liieir ability to
set rules wiiile internal and externai conditions
change, as in the IBM PC case. Similariy, Appie.
which iiad a central position in the resource coalition
for tiie personai ciigital assistant (wiien it introduced
Newton), iost its position to Pilot. Hewlett-Packard,
and Casio. Similarly, Xerox did not capitalize on its
early leaci in tiie graphicai user interface. Digitai's
resource coaiiiion faitered. whiie .Microsoft, Sun. anci
Intei gained tiieir positions due to superior resource
networks.
Questions for Managers
• Does your organization lia'e a logic for sourcing
that distinguishes the assets to manage inside from
those that can be a.ssembled via relationships in the
business network?
• Do you ha'e a .sy.stematic approach fbr identifying
the modules that you can obtain from external part-
ners? How efficient is your sourcing process com-
pared to companies such as GE's TPN?
• Do you consider process outsourcing as the best
way to allocate internal resources to high-valuc-
:idding areas? .More importantly, do you still carry' out
processes iniernaliy tiiat may best be done outside?
• Can your sourcing strategy dynamicaiiy respond to
make-partner-buy decisions? Is it structuring and
iiKuiaging a portfbiio of relationships for obtaining
the ret|uired capabilities?
• Can you create interdependencies within your
processes across organizational i')oundaries? .^re tiiey
seaniiess and supporteci by IT?
• How weli arc you iiaiancing dependence on part-
ners in tiie resource coaiition witii tiieir dependence
on you? Are the coaiition partners favoring you over
your competitors (wiio may aiso overiap witii tiie
network fbr accessing siniiiar resources)?
• How are you asse.ssing your progress in tiiis vector
as you stri"e to efficientiy source moduies. reconfig-
ure processes, and orchestrate a superior position
w ithin the resource coalition? Do you liave a score-
card of financial and operational metrics to monitor
your performance?
Knowledge Leverage (Virtual Expertise)
The current trend is toward greater empioyment
opportunities in smalier companies, fewer permanent
or guaranteed joiis. more work done at remote ioca-
tions, greater reiiance on contract workers, and
greater miiization of teams. Drucker has asserted that
companies are siiifting "from the command and con-
trol organization, the organization of departments and
di"isions. to the information-based organization, the
organization of knc;)wledge specialists."-" The basic
economic resource is not land, physical resources, or
cajiital but knowledge and inteiiectuai assets."' In
43
Sloan Management Review
Fail 1998
Venkatraman • HendersDn

12. Quinn's view of an inteiiigenl organization rooted in
knowledge: "These knowlecige assets can disinterme-
diate iiureaucracies, draniaticaiiy iower overiiead
costs, support rapid execution of strategies, and suii-
stantially increase the iearning rate of empic^jyees and
tiieir response to customers."" l^otii iinpiy tiiat effec-
tiveness comes from ieveraging inteiiect and knowi-
edge rather than from economies of scale in opera-
tions or physical sources of acivantages.
44 This vector complements the logic of virtual organiz-
ing in the previous two vectors. Here we focus on
tiie possii^iiilies and meciianisms fbr ieveraging
expertise at many levels. Senior managers face these
questions: How do we recognize anci ieverage
knowiedge as a corporate a.s.set? How siiould we
manage knowledge workers wiien iuiman resource
poiicies are geared toward prociuction anci adminis-
trative workers? How do we design processes for
ieveraging knowledge when most proce.sses are
ciesigned fc^r cosi optimization and efficiency? What
mechanisms shouici we use for ieveraging lacit
knowledge, since we have focused previously on
explicit (codified) knowledge? What incentives anci
compensation practices fiike stock options) wili
attract and keep knowiedge workers? To address
these and related questions, we delineate three stages
of knowledge leverage: iiarnessing work-unit exper-
tise, recognizing knowledge as a corporate asset, and
gaining access to broad professional community
expertise.
Work-Unit Expertise
Recently, the redefinition of tasks has been accompa-
nied by more work being distributed across time and
distance. More tasks can be decomposed so they can
be done at different lcjcations and time periods (for
example, customer service, order fulfillment, or new
product development). However, their effectivene.ss is
ensured througii the depioyment of integrateci controi
systems supported iiy groupware.
The expertise to carry out the tasks of a work unit, lo
achieve performance improvement as a leam, cannot
be achieved iiy individuals. American Airlines
deployed an expert .system. Smarts, to complement
its reservation system, Saiire. .so sales repre.sentatives
couici develop target proniotionai programs to
increase its market share anci profitability.'- Merriii
Lyncii's succe.ss is attriiiuted to its aiiiiity to codify
and distrii>ute higii-ievei anaiyticai skiiis to its differ-
ent work units.''
Hewlett-Packard faceci the challenge of capturing the
knowledge from tiie use anci support of its compiex
gioiial products. Customers neecied support thai
involved constantly changing technical configurations.
In 1995, HP created a case-based reasoning tool to
capture teciinicai support knowledge for possiiiie use
i")y sen'ice and support personnel worlciwide. This
initiative iias made tiie distributed service work unit
effective: tiie average time for caiis has been reduced
two-thirds, and tiie cost per cali has iieen lowered SO
percent.''
An organization s aiiiiity to make proce.s.ses effective
is increasingiy supported by impro'ements in IT
functionality like groupware (Lotus Notes), ideocon-
ferencing, and intriinets tliat facilitate teani-leel coor-
dination and exciiange of information and knowi-
edge. Teams de'eiop efTecti'e routines fbr siiaring
knowlecige anci expertise. Consulting companies have
created successful modeis of knowiedge siiaring
w'itiiin teams tiiat are not in tiie same time zone.
Pharmaceutical companies iiae cieployed new prod-
uct development teams that take advantage of the
twenty-foilr-iiour ciock by ciistriiiuting work to teams
in differenl time zones. Teams in financiai senices
firms work around the ciock to create products and
service clients in the global financiai marketpiace.
Case-iiaseci reasoning tools (like tiie HP case),
expert systems, neurai nets, and the Weii aiiow
firms lo capture anci ieverage knowledge to these
ciisiributed processes on a giobai basis. The real
challenge in maximizing work-unit expertise lies
not in ciesigning the technological platform to sup-
port group work but in designing the organization
structure and processes. What is the best way to
bring different perspectives into a task team? How
do we balance tlie need for colocation w itii the
rec]uirement lo get the iiesl possiiiie expert? How
do we design tiie performance assessment system
for tiie work unit so that it encourages teamwork
ratlier than individual performance? Many compa-
nies iike Honcia. Unilever, Joiinson <!<: Johnson,
Motorola. GH. Boeing. Kodak, Xerox, and others
are experimenting witii structures, processes, and
technologies that maximize work-unit expertise as
they move away from functionaliy baseci organiza-
tions to a process-driven approach.
Corporate Asset
Tiiis second stage focLises on harnessing the collec-
tive expertise acro.ss work uniis. rather than within
fenk3traman • Hendeison Sloan Management Review

13. units. The main focus is less on tangible and codified
knowledge and more on tacit knowledge to collec-
tively leverage. Both Xerox and IiP treat knowledge
as a corporate asset, one of the new drivers ot bot-
tom-line profits and top-line growth.
• Xerox captures knowledge about new ways of ser-
vicing its machines. Many different knowledge workers
— customer service engineers, field .sendee engineers,
and product .specialists — created the Eureka prcx'ess,
which rests on a database called Tips. Key to this
prcK'ess is the emphasis on validation of expertise and
knowledge. After a "tip" (an idea or suggestion to .solve
a problem) is submitted, tip authors can monitor its
progress through the validation proc-ess. and the origi-
nator's contribution is recognized. They can obsen-e all
die comments and interact with alidators. The tip is
seen not as the asset of an isolated department or a
function but ofthe corporation.''
• Starting in 1995, BP has experimented with a vinu-
al teamwork program. When ecjuipment fails in the
North Sea, the drilling engineers haul the faulty hard-
ware in front of a tiny video camera connected to
BP's virtual teamwork .stations. Using a satellite link,
they can call a drilling equipment expert in Aberdeen
who can visually inspect the part while talking to the
shipboard engineers. Thus distributed expertise is vir-
tually brought to bear on the problem tliat recjuires
time-sensitive response."' According to John Browne,
group chief executive, British Petroleum: "Advances
in communications technology — particularly high
bandwidth communication and information systems
— allow the people working on a field in the North
Sea to talk directly and to share experience with the
people working offshore on a field in the Gulf of
Mexico 5,000 miles away."''
Knowledge is a corporatewide asset that
should be systematically managed.
Consulting organizations like Arthur Andersen.
McKinsey, KPMG, and others are also focusing on
identifying, capturing, and leveraging their knowl-
edge assets. Arthur Andersen combined convergent
and divergent systems; the former focuses on codi-
fied information translated into knowledge and com-
municated through standard channels at prespecified
inten'als; the latter is a knowledge-creating system
based on information not codified but disseminated
in real time. Similarly, the World Bank is reposition-
ing itself as a provider of knowledge-driven services
(www.worldbank.org). According to James Wolfensohn.
president of the World Bank: "We need to become, in
effect, the Knowledge Bauk.'""
Common to these examples is the notion that knowl-
edge (or intellectual assets) is a corporatewide asset
that should be systematically managed. At the U.S.
Army's Center for Army Lessons Learned (CALL) dur-
ing the 1994 Haiti invasion, CALL experts inter'iewed
soldiers about incidents with mobs and confrontation
with local authorities, obsen'ed after-action reviews
(AARs). and read intelligence reports to compile
les.sons. CALL developed twenty-six training scenarios
for replacement units getting ready for Haitian duty;
in the following six months, the units encountered
twenty-three of the twenty-six scenarios. Similarly,
the CALL process helped the army in Bosnia with
early reports of incidents and the consequent distilla-
tion of likely actions." Companies like Motorola. GE,
Stc'clcase, BP," and GM'' are beginning to u.se the
AARs developed by the army — each developing
their own scenarios or lessons.
Professional Community Expertise
In the third stage, the focus is on the community of
professional expertise — well beyond the domain of
a focal organization. For instance, 1,000 staff people
at HP collaborate with more than thirty of their part-
ners and suppliers through a virtual team platform,
desktop collaboration, information-sharing tools, and
videoconferencing. Through coaching, BP has been
able to transfer tacit knowledge acro.ss lioundaries.
Organizations are increasingly leeraging the exper-
tise in the extended network (suppliers, customers,
partners, alliances, and so on) and the broader pro-
fe.ssional community. For instance, .some firms retain
a core of experts on their payroll while contracting
for other skills as needed. LInlike contract workers
who smooth out fluctuating demand, this expertise is
for specific tasks and is often connected electronical-
ly. The knowledge of lawyers, medical practitioners,
advertising executives, investment bankers, and tech-
nology gurus is often leveraged within an extended
community through electronic networks.
Recent advances in IT — especially the deployment
of a global, accessible, collaborative infrastaicture —•
fundamentally enable this virtual leveraging of exper-
45
Sloan Management Review
fall mi
Venkatraman • Hendeison

14. tise in the professional network. It is possihle to get a
second opinion on a medical procedure by allowing
other experts to 'iew the complete medic'al liistor-
and associated images from remote locations. Multi-
ple experts from different locations can simultaneous-
ly interpret the meaning of an unforeseen ex'ent like
a currency crisis in Mexico or lairope or realignment
in the e.xchange rate between the K.S. dollar and the
Japane.se yen. As com|-)anies identify antl integrate
expertise from multiple sources, they face the clial-
46 Icnge of how best to compensate anci motivate
employees with expertise. The traditional employ-
ment c-ontract might have to be replaced with a more
specific contract linked to the le-erage of expertise.
Kmergent x'irtual ct)mmunities .sen'e as a primar>'
mechanism for collecting and legitimizing know-ledge
and expertise. The Motley Fool (www.fool.com) is
one example; more .sLicli communities will likely
emerge.
Amazon is strategically positioning itself in the
knowledge network to indutle book reviewers and
critics, technology de'elopers who constantly pLisli
Amazon to the cutting edge of the Web platform,
logistics experts who optimize the physical distribu-
tion of books, and sociologists who stud- the emerg-
ing patterns of electronic communities. Amazon
needs a wide range of experti.se if it is to be at the
top of this new business model. The knowledge net-
work from the community is more important as it
expands be-ond books to incliEde movies. CDs, anci
oiher products ibr the same customers. Its aiiility to
Ie'erage the knowledge and expertise of its commu-
nity might help Amazon maintain its leadership posi-
tion.
Questions for Managers
• Does your organization recognize the importance
of knowledge and intellectual assets in creating
-aiue? If .so. what are the mechanisms (formal and
informal) for implementing?
• How well are your task units fteams and groups)
le-eraging their collecti-e expertise? What characteris-
tics of the IT platform support knowledge leveniging
%-ithin the unit?
• Do you treat knowledge as a corporate asset? If .so.
how is knowledge linked to organizational effectiveness?
• Are you effecti-e in positioning your prtjfessional
expertise within a comnumity rele'ant for your busi-
ness? What is your position in the community, rela-
tix' to your competitors?
• Wliat qualitative anci quantitative indicators are you
adopting to better le'erage knowledge in creating
value? What companies cio you Lise for benchmarking?
Nine Challenges for Transition
The implications of making a transition to the new
business model are:
1. Shifting value drivers. i-A ei"j- corponition should
develop a systematic approach to recognizing and
responding to shifts in what dri'e.s value. Microsoft,
Intel. Starbucks, Amazon.com, 'ah()o!. IW l-!xplora-
tion, and others are focusing c^n kntjwledge as dri-
ers of their business capabilities. The stock market
values these firms for their intellectual assets more
than for their physical assets. Recognizing the shift
anci responding to it is key.
2. Designing the new business model. Tlie business
model is a coordinated plan to design strategy" along
all three vectors rather than leading in any one 'ec-
tor. Strategies should ret-ognize the interdependencies
among the three vectors.
3. Governing beyond outsourcing. Sourcing has
become a primaiy aspect of the strategic logic.
Companie.s should cie'elop a strategic approach to
g{)ernance that constandy shifts the management of
tangible assets to the market 'here the market is
[iiLiture. At the same time, they sht)uld focus on the
go'ernance of 'alue-addeci intellectual assets inside.
4. Interacting with customers for knowledge lever-
age. Distribution channels are a strategic link with
c-ustomers for gaining ai-cess to critical knowledge.
Customers increasingly participate in the product or
senice development proce.sses much earlier and pro-
'ide useful information for enhancing prodLict 'alue.
More importantly, participating in the customer com-
munities enables a company to be in the midst of
information flows about new product features.
5- Navigating across muitipie communities. Market
leaders are differentiated by their ability to positicju
themsel-es in a network of communities — cust<}mer
communities, resource coalitions, and professional
communities. Companies need to play arious roles
-itliin these communities — sometimes active, other-
wise passive: setting rules sometimes while folto'ing
rules at others; and competing and cooperating with
the same firm at different times or in different settings.
Venkatraman • Henderson Slaan Management Heviaw
Fnll m^

15. 6. Deploying an integrated IT platform. Market lead-
ers sliould design ati integrated IT platform for the
new virtual busine.ss tnodel. They should experitnent
with various tec-hnological alternatives as they .stt-ive
to align their business stt-ategy with the IT strategy.
The Internet (iticluding inti-atiets and extninets)
tnakes possible a common platform that allows for
inter- and intraorganizational coorditiation oti a scale
not preiousiy possible. IT is both a cause (dist-uption
of tnat-ket equilibrium) and a solution (building
blocks for the new organizational logic). The technol-
ogy t-ec[uiretiients for [lie tht-ee 'ectors are converg-
ing, thus com]iellitig managet-s to take a more central
look at the IT platfortii. .Senior managers catinot .sim-
ply delegate resjionsibility to opet-atitig managers to
execute ati informatioti strategy- that supports the
business strategy. The rec|uitetnetit of continually
aligtiitig business atid IT stnitegies is now tnore
itnpoiiatit thati ever before.
7. Allocating resources under increased uncertain-
ty. Leaders are difierentiated by how- they allocate
critical t-esources. llnder conditions of relative certain-
ty, cotiipanies can adopt preciictable models of
resout-ce allocation (for example, the discounted cash
tlow and traditional net presetit 'alue calculations).
Howeer. the future is expected tcj be Lincertain.
Leaders should adopt the di,scipline ofthe real
options Lipproach to Licti'ely allocate resources.'- That
a|-)pt-oacli brings the disciplitie of financial markets to
strategic decisions and calls for setiior managers to
actively manage iiuestments to respond to changes
in external atid internal conditiotis.
S. Designing an organization for knowledge lever-
age, tiompanies are experimenting wiih matiy different
organizational forms that overcotne the limitations of
exi.sting .stnictures and processes. The underlying logic
for new fonns should be to leverage knowledge acro.ss
multiple le'els — within and across boundaries. There
are no .strict or rigid principles for designing organiza-
tions that leverage knowledge, but cotnpanies will con-
verge on sotiie principles as new entities design organi-
zations without the constraints of the old principles.
9- Assessing performance along multiple dimensions.
Managers should tnove away frotn static, tnyopic tnet-
rics like market share, as they Ix'cotne tneaningless with
changes in tnarket definition. Sitnilarly, they should sup-
pletnent accounting metrics like ROI, ROA, and ROE
with tnore contetnponiry metrics like EVA (econotiiic
'alue added), MVA (market value added), and metrics
that capture the share of new wealth creation. Clearly,
Yahoo! and Atnazon.com are not succe.ssful if we look
at tlieir profitability nutnlx^rs but are successful frotn a
stcK'k market valuation viewpoint.
Conclusion
While vittual organizing is a powerful concept, "irtu-
al organization" is an imfortunate term. It connotes
impo.ssibilities such as a "hollow corporation" or a
liroker." Virtual organizing as a concept focuses on
the importLince of knowledge anci intellect in creating
'alue. Our approach recognizes three interdependent
ecEots — irtual encounter, virtual sourcing, and vir-
tual expertise — supported by a powerful, integrated
IT platfortn. The strategic logic for the new business
tnodel is rooted in the interdependence atnong the
three vectors. It will be difficult — if not itnpossible
— to craft an effective strategy and structure without
harmony atnong the three vectors.
47
References
This paper is based on a research project with a grant
from the Advanced Practices Council (APCI Df the
Society for Information Management (SIMI tilled.
"Avoiding the Hoiiow: The Bjilding Blocks of Virtual
Organizing." We thank Madeline Weiss, Bob Zmud, Lee
Sproull. and members of the SIM-APC and Boston
University Systems Research Center for sharpening the
framework and developing the management implica-
tions. This work was also supported by the National
Science Foundation under Grant No SBFl 9422284
IPrincipal Investigator. N. Venkatraman). Any opinions,
findings, conclusions, or recommendations are the
authors' and dc not necessarily reflect the views of Ihe
National Science Foundation. We also thank Lee
Sproull for stimulating our thinking about knowledge
ne^vorl^s and the role of communities and PR
Balasubramanlan for useful comments.
• 1 P. Drucker, Post-Capitalist Societ
England: Butterworth Heinemann, 1993);
JB. Quinn, Intelligent Enterprise {New "{QI]!:. Free
Press, 19921; and
G. Hamel and C K Prahalad, Competing for the
Future IBoiXon: Harvard Business School Press,
1994).
Womack and Jones coined the term "lean organiza-
tion" to extend the original concept of lean produc-
tion. See.
J P Womack, O.T. Jones, and 0. RODS, The Machine
That Changed the WorldWevi York: Rawson
Associates, 1990).
C Handy. The Age of Unreason {BosXon. Harvard
Business School Press, 1989).
• 2. See, for instance:
D. Nadler, M Gerstein, R, Shaw, and associates.
Organizational Architecture: Designs for Changing
Organizations (San Francisco: Jossey-Bass. 1992).
• 3. S.E. Rasmussen. Experiencing Architecture
(Cambridge, Massachusetts MIT Press. 1991),
• 4. Nadler etal. (1992).
• 5. For an overview of the characteristics of the
new marketspace, see:
J.F. Rayport and J. Sviokia, "Managing in the
Marketspace," Harvard Business Review, volume 72,
November-December 1994, pp. 141-150.
• 6 See J. Hagel III and A.G. Armstrong, Net.Gain:
Expanding Markets through Virtual Communities
(Boston: Harvard Business School Press, 19971
• 7. When we refer to customer, we include both
customers and end-consumers
• 8. N. Negroponte, Being D/pifa/[New York. Alfred
A. Knopf, 1995).
• 9. See, for instance:
J.B. Quinn, J. Baruch, and K.A. Zien, Innovation
Explosion (New York: Free Press 1997); see also:
A.J. Slywotsky and D Morrison, The Profit Zone
Sloan Maoagemenl Review
Fall 1998
Veokattaman • Henderson

16. (New York: Times Business, 1998).
• 10. C, Baldwin and K. Clark, "Managing in an Age
of ModLlarily," Harvard Business Review, volume 75,
September-October 1997, pp. 84-93.
• 11. B.J. Pine III, Mass CustomuationiBo^tan:
Harvard Business School Press, 1993).
• 12. M. Dell, "The Power of Virtual integration,"
Harvard Business Review, voiume 76, March-April
199B, pp. 73-84.
• 13. Hagel and Armstrong (1997), p. 37.
• 14 For an overview, see Hagel and Armstrofig
11997).
• 15 In this paper, we use the lerms competencies
48 and capabilities interchangeably. The main distinc-
tion in our view is that Prahalad and Hamel use the
term competencies in their paper more narrowlv to
refer to technology-based competencies. See:
C.K. Prahalad and G. Hamel, "The Core Competence
of the Corporation," Harvard Business Hevieiv. vol-
ume 68, May-June 1990, pp. 79-91.
Fora broader view, see:
G. Stalk, P. Evans, and L E. Shulman, "Competing on
Capabilities: The New Rules of Corporate Strategy,"
Harvard Business Review, March-April 1992, pp. b7-
69. We adopt a broader definition when using the
two terms interchangeably.
• 16. W. Davidow and M. Malone, Itie Virtual
Corporar/on (New York: Harper Collins, 1992!.
• 17. S. Goldman, P. Nagel, and K. Preiss, Agile
Competitors and Virtual Organiiations (New York:
Van Nostrand Reinhold. 1995).
• 18 Quinn (1992)
• 19. See Quinn (19921: and
R Venkatesan, "Strategic Sourcing: To Make or Not
to Make," Harvard Business Review, volume 70,
November-Qecember 1992, pp. 98-107.
• 20 See, for instance:
J. Lewis, The Connected Corporatior) (New York: Free
Press, 1995);
Guinn 11992): and
B. Gomes-Casseres, Ttie Alliance Revolution
(Cambridge. Harvard University Press, 1996)
• 21. See. for instance
Venkatesan (19921
• 22. Baldwin and Clark 11997).
• 23. Negroponte (19951.
• 24. For more details, see.
H V^. Chesbrough and D.J Teece, "When Is Virtual
Virtuous' Qrgani/ing for Innovation," Harvard
Business Review, volume 74, January-Febmary 1995.
pp 65-73
• 2b- See J. Short and N. Venkatraman, "Beyond
Business Process Redesign: Redefining Baxter's
Business Network," Sloan Management Review, vol-
ume 34, Fall 1992, pp. 65-73.
• 26. G2 Research.
• 27 See N Venkalraman, "IT-Enabled Business
Transformation: From Automation to Business Scope
Redefinition," Sloan Management Beview. volume
35, Winler 1994, pp. 73-87
• 28 See A. Bfandenburger and B. Nalebuff, Co-
opelition{Uevj York. Currency Quubleday, 19961
• 29. P Drucker, 'The Coininij uf the New
Organization," Harvard Business Review, volume 66.
January-Februarv 1988, pp 44-63
• 30. See, for instance:
T. Stewart, Intellectual Capital{^e York: Currency-
Doubleday, 1997).
• 3l.0uinn|1992|.p 102
• 32. N. Venkatraman and E. Christiaanse.
"Electronic Channels for Expertise Exploitation' An
Empirical Test of the Airline-Travel Agency
Relationships," Academy ot Management Best Paper
Proceedings. 1996
• 33. See J.B. Ouinii, R Anderson, and S. Finkelstein,
'Managing Piofessicnal Intellect Making the Must
of the Best," Harvard Business Review, volume 74.
March-April 1996, pp 71-80
• 34 See T Davenport and L. Prusak, Working
Knowledge How Organiiatioiis Manage What They
^nuwlBoston: Harvard Business School Press, 199B).
• 35. See D. Bell, D. Bobrow, 0 Raiman, and M
Shirley, Dynamic Documents and Situated
Processes Building on Local Knowledge in Field
Service." in T Wakayama. S Kannapn, C M Khoong,
S Ndvdthe, andJ Yales. n^^. Information and
Process Inlegrdlion in Enterprises: Rethinking
Documents HaNe. Massachusetts Kluwer
Academic Publishers, 19971.
• 35, Field research at Boston LJniversity School of
Management, Systems Research Center, 1997.
• 37 J Browne, speech, "Science, Technology, and
Responsibility" (London. Royal Society, 28 October
199/1
• 38. J Wolfensohn. 1996 annual meeting speech.
Available at
miAv.worldbankorg/html/extdr/extme/jdwamsaB.htm.
• 39 See G R Sullivan and M Harper. Hope Is Not a
Method What Business Leaders Can iearn troni the
AiinyiUtiw York Bruddway Buuks. 1995).
• 4(J Field resedrch at Bustun University Schoul of
Management. Systems Research Center
• 41. Boston University School of Management,
research on knowledge management,
• 42 See M. Amram and N, Kulatilaka, Real
Options Managing Strategic Investments in an
Uncertain l/Vo/WIBoston Harvard Business School
Press. 1998). and
N. Kuldtilaka and N Venkatraman. "Are You
Preparing to Compete in the New Economy? Use a
Real Options Navigator" (Boston. Boston University
School of Managenient, working paper, August 199B|.
Reprint 4013
Venkatraman • Hendeison Sloan Management Review
Fall 19gB