In the 1990s, conventional wisdom held that corporate ... · In the 1990s, conventional wisdom held that corporate leaders should leave divisions alone, ... Len Lauer, then president

92 HARVARD BUSINESS REVIEW

In the 1990s, conventional wisdom held that corporate leaders should leave divisions

alone, so long as they hit their numbers. That's changing. In uncertain markets,

CEOs need to provide strategic guidance more quickly and more often than they used to.

LEADFROM THE

CENTERHow to Manage Divisions Dynamicaiiy

by Michael E. Raynor andJoseph L Bower

No ONE NEEDS Convincing that in today'sturbulent,competitive business environments corporationsmust be fiexible and responsive. A host of forces are

arrayed against even the most prescient strategist: tech-nology, regulation, and globalization, to name only three.The question companies now face is not whether theyneed to be nimble and quick, but how.

Most of the advice on this score is remarkably consis-tent. Especially in large, complex, diversified companies,the prescription is "more decentralization"-at the limit,an almost complete devolution of decision-making author-ity to the operating divisions and those people closest toemerging technologies, competitors, and customers. Thispoint of view has been espoused so often and with suchconviction that one might even refer to it as the conven-tional wisdom.

MAY 2001 93

Lead f r o m t h e Cen te r : How to Manage D i v i s i o n s D y n a m i c a l l y

Like most conventional wisdom, however, this ap-proach does not always serve us well. We have found thatresponding effectively in uncertain markets often re-quires more-not less-direction from the center. Ourresearch into contemporary diversified companies sug-gests that, in industries undergoing rapid and difficult-to-predict change, corporate headquarters must play anactive role in defining the scope of division-level strategy.Furthermore, to compete effectively as a corporation, itoften falls to the CEO and a select staff to drive the tim-ing and nature of cooperative efforts between divisions.

The role of the corporate office in creating a strategi-cally flexible organization has cascading effects on how ex-ecutives manage other aspects of the company: whetherdivisions are clustered into groups, for example, and howcompensation is structured. The result is a set of man-agerial challenges fundamentally different from those ofdiversified companies in more stable and slowly changingenvironments. Through an examination of four corpora-tions-Sprint, WPP, Teradyne, and Viacom-we will ex-plore the defining managerial characteristics of what itmeans to manage divisions dynamically.

The Sprint StoryThe Sprint case illustrates the need for corporate-levelstrategic fiexibility-and is an example of its successful im-plementation. Now a $20 billion telecommunications cor-poration, Sprint began life in 1901 as a rural Kansas tele-phone company. For the next 75 years, the company grewin traditional ways, acquiring other local telephone com-panies and vertically integrating into the manufactureand distribution of telecommunications equipment. Be-tween 1976 and the mid-1990s, Sprint developed intoa major diversified telecom player: it built a nationwidefiber-optic network, created a substantial long-distancebusiness, entered the digital wireless business, took a mi-nority stake in an Internet service provider, invested inbroadband, and entered the Latin American, Western Eu-ropean, and Asian markets.

These businesses are at least nominally related, in thatthey all compete in the telecommunications industry.However, for much of its history. Sprint has had very fewopportunities to capture meaningful synergies betweenits various businesses. For example, when Sprint movedinto the long-distance market, it was unable to draw onexpertise or assets from the local division, largely for reg-ulatory reasons. Even ignoring regulation, the two busi-nesses had little in common: they were based on differ-ent technologies, served different markets, and operated

in different competitive contexts. That was the casefor many of Sprint's new business initiatives, whichfunctioned independently of one another despite theirsurface-level similarities.

As technological, regulatory, and market forces beganto change. Sprint looked for ways to exploit synergies be-tween its once autonomous divisions. Consider the evolv-ing relationship between the local telephone division andthe consumer long-distance division. In the wake of theTelecommunications Act of 1996, which, among otherthings, began to break down the barriers between thelocal and long-distance businesses. Sprint launched OneSprint, an initiative designed to expand the local phonebusiness by selling the fullest possible range of Sprint ser-vices. Practically speaking, this meant finding ways tocross-sell long-distance service to local customers. Begin-ning in 1998, the local telephone and the consumer long-distance divisions cooperated to create bundled products;for a flat fee. Sprint's local customers could purchase localloop access, specialized dialing features, and blocks oflong-distance time in configurations previously unavail-able from the long-distance division.

The results speak for themselves. According to Sprint'sexecutives, by the end of 1999 Sprint had achieved 27%consumer long-distance market share in areas in whichSprint was the local provider, compared with 7% marketshare in other areas. In fact, as the program gatheredsteam, during one 30-day period Sprint increased its mar-ket share by as much as 2%. Mike Fuller, CEO of the localbusiness, says his group became the long-distance divi-sion's best distribution channel.

In this case, the benefits of integration were clear, andcorporate headquarters did not have to twist arms to getthe divisions to cooperate. It simply set targets for thelocal division based on developing new sources of rev-enue. The long-distance division, meanwhile, operated onthe straightforward belief that any channel that prof-itably increases market share is a good channel.

In other cases, however. Sprint's corporate headquar-ters needed to take direct action to encourage coopera-tion among divisions. By doing so. Sprint was able to cre-ate and exploit corporate-level strategic flexibility.

Creating Flexibility

Michael E. Raynor is a director in the Toronto office ofDeloitte Consulting. Joseph L. Bower is the Donald K. DavidProfessor of Business Administration at Harvard BusinessSchool in Boston. This is his seventh article for HBR.

In creating strategic flexibility, a corporate office mustbalance the immediate need for divisional autonomywith the potential need for future cooperation. Withoutthis balance, divisions may act in ways that advance theircurrent competitiveness but undermine opportunitiesto collaborate in the future. The way in which Sprint ap-proached wireless Internet access is an excellent exampleof how to achieve balance.

Launched in 1998, Sprint PCS was an extremely suc-cessful new entrant into the wireless market, signing on


Lead from the Center: How to Manage Divisions Dynamically

4 million customers in its first full year of operation.When it decided to add wireless Internet service to its ex-isting wireless telephone service in 1999, PCS's executivesfaced a critical decision: which software interface shouldit use for customer access to tbe Internet. (For various rea-sons-including the small size ofthe handsets-estab-lished Internet browsers like Netscape Navigator andMicrosoft's Internet Explorer were

not options.) The key question washow much flexibility should PCSgive consumers: should they beable to customize their access soft-ware by, for example, choosing ahome page and other settings? Orshould Sprint hardwire those fea-tures? From the division's point ofview, it was a no-brainer: givecustomers as many choices as pos-sible. That would help the divisiongain market share and wouldprobably make the new serviceprofitable more quickly.

Corporate headquarters saw theproblem differently. Taking a longerterm, higher level perspective, theexecutives concluded thaf'owning"the interface between the cus-tomer and the network trumpedthe immediate appeal of growingmarket share quickly by appealingto customers' interest in openness. They foresaw a time inthe not-too-distant future when they'd want a commonplatform for various emerging products-narrowband,broadband, wireline, wireless, and so on. As a result, theyoverruled Sprint PCS's executives. Even though therewere no definite plans for integration. Sprint's corporateoffice created fiexibility by constraining the activities of itsfastest-growing division. Had Sprint given PCS completefreedom to pursue its own course -one that made perfectsense in isolation - it could have undermined the scope ofactivities that the larger organization might pursue in thefuture. Strategic constraints of the kind Sprint imposedhere are critical if a diversified corporation is to be gen-uinely fiexible.

The use of strategic constraints should not be confusedwith routine intervention by a corporate office. To createstrategic fiexibility, divisions must still enjoy consider-able operational autonomy and remain competitive asstand-alone businesses. Companies must walk a thin line -something Sprint's corporate executives seem to have ac-complished. In the PCS example, corporate executives im-posed constraints on the division, but in other cases, thecorporate office adopted a hands-off approach. Considerthe following example, in which a division made majoroperational changes with only minimal oversight.

HAD SPRINTGIVEN PCS

COMPLETE FREEDOM TO

PURSUE ITS OWN COURSE-

ONE THAT MADE PERFECT

SENSE IN ISOLATION-IT

COULD HAVE UNDERMINED

THE SCOPE OF ACTIVITIES

THATTHE LARGER

ORGANIZATION MIGHT

PURSUE IN THE FUTURE.

When the executives of Sprint's consumer long-distancedivision thought their "dime a minute" marketing plat-form was getting stale-partly because its longtimespokesperson, Candice Bergen, was no longer starring ina popular television show-they wanted to lower theirprices and make some other brand-altering decisions. LenLauer, then president ofthe division, and his team devel-

oped the "nickel nights" concept,

and Lauer ran it by Sprint's COO,Ron LeMay. The two executivesspent all of 30 minutes together,and then Lauer's team was freeto launch the new identity of theconsumer long-distance company-and cut the price of its core prod-uct in half. The only constraint im-posed was that the new programhad to meet rate-of-retum hurdlesestablished by the finance staff toensure that the company met itsEVA targets. Determining the rightmarket objectives- and the pricingtactics to meet those objectives-was the exclusive purview of theoperating division.

The kind of operational auton-omy that Sprint's consumer long-distance organization enjoyed inthis case demonstrates that a di-vision must be a successful stand-

alone business in its own right, rather than being de-pendent for success on future integration with otherdivisions. Corporate-level strategic fiexibility is not merelystaged integration, and the increasing interdependence ofoperating divisions is never a foregone conclusion. In-stead, divisional autonomy is combined with strategicconstraints to ensure that strong businesses - competitivein their own right - have the ability to integrate when andif the opportunity arises.

Exploiting FlexibilityThe goal, then, is to make sure the company is in a posi-tion to move nimbly when opportunities for integrationemerge. The way Sprint, for example, bundled its con-sumer long-distance and wireless services provides insightinto the demands that exploiting strategic fiexibilityplaces on the corporate office.

Recall that Sprint PCS had been launched in 1998 andthat its revenues and market share were growing ex-tremely quickly. For the consumer long-distance division,the PCS business represented an invaluable opportunityto sell Sprint long distance: by bundling long distancewith the popular wireless offering, the division could gainvaluable market share.

MAY 2001 95


The consumer long-distance marketing group was en-thusiastic about exploiting Sprint PCS's distribution chan-nels, but PCS's managers strongly resisted the idea. Onereason was that the sales process for PCS services is a longand expensive one: it can take upto 90 days and four visits beforea customer signs the contract Anintegration initiative that merelytacked on consumer long-distanceservices to existing PCS offeringswould run the risk of corruptingPCS's sales process, thereby un-dermining the very growth thatthe long-distance group sought toleverage.

It was a PCS product technolo-gist who found a possible way outof that impasse. He came up witha plan to integrate long-distanceand wireless services and increasethe attractiveness of the PCS of-fering-in a way that promisedto double the long-distance divi-sion's gross margin and improve customer value. His ap-proach, which was dubbed "block of time lite," or BOTL,was based on PCS's model of selling blocks of time -100,200, 500,1,000 minutes per month-for a set fee. Thesticking point in the typical sales process was that cus-tomers felt one plan didn't provide enough talk time, butthe next offering provided too much. The new BOTL planwould allow customers to use the minutes for eitherlong-distance or PCS services. This shift would mitigatecustomer concerns about over- or underbuying and letthe long-distance group charge ten cents (rather than thenow-standard five cents) a minute.

Whatever the theoretical attractiveness of this productoffering, however, PCS's executives still resisted it vigor-ously. Their division was growing fast, and their compen-sation was linked to meeting targets developed long be-fore this approach was proposed. Introducing productsthat might upset PCS's well-oiled and rapidly expandingdistribution systems was legitimately viewed as a riskyundertaking.

The corporate office saw the strategic value in the in-tegration and intervened to allay the PCS division's vari-ous concerns. PCS's executives were worried, for example,that the long-distance business's high level of churnwould increase their own churn. Tom Weigman, an exec-utive vice president who'd been president of the con-sumer long-distance division before assuming a corporaterole, tackled the chum debate at a very detailed level, de-veloping a plan to roll out the bundled service in severaltest markets. PCS's executives also argued thatintroducing the BOTL plan so late in the year (it wasslated for trial in the fourth quarter of 1999) would make

WHEN SPRINT BELIEVES THAT

A POTENTIAL SYNERGY IS

READY FOR EXPLOITATION, IT

BUILDS THAT BELIEF INTO THE

COMPENSATION STRUCTURE,

RESULTING IN DRAMATIC

SWINGS IN THE ALLOCATION

OF BONUS PAYMENTS.

PCS miss its operating targets. Al Kurtze, then an execu-tive vice president in Sprint's corporate office and for-merly the COO of Sprint PCS, stepped in and demon-strated how the new product tests could be rolled out

without endangering targets forthat fiscal year. As a result of thecorporate office's efforts, BOTLwas launched in test markets in1999. It took the detailed, top-down involvement of the corpo-rate executives to persuade divi-sion leaders that placing strategicconstraints on the fast-growingdivision would improve competi-tiveness in the long run.

Another aspect of exploitingflexibility is to have malleablecompensation structures that canreflect rapid changes in the de-gree of interdivisional coopera-tion. Sprint's management incen-tive program -which pays divisionpresidents annual cash bonuses of

up to 100% of their base salaries-certainly does that. Nodivision president has more than 65% percent of thebonus dependent on his or her division's performance.The remaining 35% to 50% depends on the performance ofother divisions, thereby providing an incentive to searchfor synergies. When Sprint believes that a potential syn-ergy is ready for exploitation, it builds that belief into thecompensation structure, resulting in dramatic swings inthe allocation of bonus payments. For example, bonuspayments tied to cross-selling initiatives can change by asmuch as 300% in a single year, accounting for 30% ofthetotal incentive bonus payment.

Dynamic ThinkingAbout DiversificationSprint exemplifies many of the defining characteristicsof strategic flexibility (see the exhibit "Characteristics ofCorporate-Level Strategic Flexibility). These characteris-tics lay the groundwork for a dynamic approach to man-aging corporate strategy at diversified corporations.

Any framework for thinking about corporate strategy isbuilt on an understanding of how divisions interact withone another and how they interact with the corporate of-fice. If divisions share valuable resources or capabilities,we take it as evidence of a strategy of related diversifica-tion. In such cases, the relationship between headquartersand divisions is structured to facilitate interdivisional co-operation. Divisions with related operations are often or-ganized in groups, headed by corporate-level executivesresponsible for managing synergies between divisionswithin their groups. Alternatively, if divisions compete for



resources or capabilities, we take that as evidence of un-related diversification. Division-level executives lead theorganizations as stand-alone businesses; the corporate of-fice typically exerts strong financial oversight but rarelyintervenes in divisional affairs.

What is remarkable about Sprint is that the companydoesn't conform to either model. Nor is it making thetransition from unrelated to related diversification (thoughthis is what an observer might assume). Instead, Sprint isdoing something new that calls into question several as-sumptions that have long been central to conventionalthinking about the management of diversified organiza-tions. Let's look at how several other companies are chal-lenging those assumptions and exploring new territory.

The first assumption is that diversification strat-egy must be either related or unrelated. But manycompanies are finding that some objectives arebest pursued with stand-alone divisions and oth-ers with cooperating divisions.

This is precisely the approach Martin Sorrellhas taken as CEO of WPP, the world's largest mar-keting services firm. Over the past 15 years, Sorrellhas created a collection of professional servicesfirms that includes three of the largest ad agen-cies in the world-J. Walter Thompson, Ogilvy &Mather, and Young & Rubicam-as well as pub-lic relations, corporate identity, market research,and specialty ad firms. His goal: to provide soup-to-nuts marketing services to large clients. Inthe pursuit of that goal, somewhat counterin-tuitively, Sorrell's structural moves have createdunrelated, freestanding divisions. To sustainstrong performance, a powerful finance and trea-sury staff has introduced the controls and incen-tive systems one expects to find in an unrelateddiversifier.

At the same time, Sorrell has taken steps to en-courage cooperation among stand-alone units.As at Sprint, incentive compensation and a stockprogram are managed to encourage cooperation.Most intriguing, perhaps, are WPP's"virtual com-panies." These organizations have essentially nostaff- a CEO and an assistant at most - and no of-fices. One, the Common Health, became the world'sbiggest health care marketing services firm with just asingle employee-the CEO. For each project, the CEOwould assemble a team from an ever-shifting alliance ofsubsidiaries, selecting the right capabilities to respond toeach client's needs.

Yet even the virtual companies, built on a model ofstrong interdivisional cooperation, have structures andprocesses that resemble those of an unrelated diversi-fier. The virtual companies have no income statement;profit and loss is calculated by project and allocated tothe participating subsidiaries. Compensation is based on

subsidiary performance and, to a significant extent, one'scareer develops in a given subsidiary.

WPP's approach demonstrates that it is possible to pur-sue varying degrees of relatedness among divisions. Thatis, some divisions are tightly linked, others operate in amore loosely allied way, while others are entirely stand-alone. Moreover, at WPP, the degree of relatedness waxesand wanes depending upon strategic circumstances. Re-latedness is sought only when the professionals withinfreestanding businesses believe that they can serve theirclients better by achieving it.

A second assumption of traditional thinking is thatonly business units-not the corporate office-createvalue for the company. But we've found the corporate

CHARACTERISTICS OF CORPORATE-LEVELSTRATEGIC FLEXIBILITY

Companies create flexibilily by balancing the competing demands ofdivisional autonomy in the present with divisional cooperation in thefuture.Theyexploitflexibility by moving nimbly when opportunitiesfor cooperation emerge.

PHASE DEFINING CHARACTERISTICS

CREATING • Build a portfolio of businesses between whichvaluable synergies are possible, now or in the future

- Ensure that divisions have sufficient autonomy todevelop business models and succeed as stand-aloneentities

• Impose strategic boundaries to ensure that futureintegration remains possible

' Reward search activity

ExPLOlTi NC • Promote integration opportunities when they areexpected to be profitable to the company as a whole

• Engage dimion-level resistance

• Change reward systems to reflect the importanceof interdivisional cooperation

office can create value by assembling a portfolio of assetsand capabilities that will drive competition in the future -and managing those assets in a flexible way so as not tohinder the ability of the divisions to compete now. Al-though the performance results show up in better com-petitiveness at the business-unit level, the seeds of thesuccess-the value creation - reside in the forward-look-ing actions of the corporate office.

Consider how the CEO of Teradyne, a manufacturer ofchip-testing equipment, created new value by fostering anemerging technology in the face of divisional resistance.Through the mid-1990s, Teradyne's primary product line

MAY 2001 97


consisted of highly engineered, very expensive testers soldto manufacturers of microprocessors. Each division hadseparate manufacturing facilities and a separate P&L;they competed for budget and capital and were compen-sated based on divisional performance. However, they allsold and serviced their products through a single globalorganization.

In the early 1990s, CEO Alex d'Arbeloff realized thatnew chip technology and advances in software might

TWO APPROACHES TO MANAGINGDIVERSIFIED COMPANIES

Conventional wisdom assumes that divisions are either related or unrelatedand that new value is created only at the business-unit level. New thinking,however, offers a dynamic approach to cooperation among divisions, throughdirect action by the corporate office.

STATIC THINKING DYNAMIC THINKING

Diversification strategy iseither related or unrelated;interdivisior}al relationshipsare static.

Diversification strategy can be a mixture ofrelated and unrelated elements; companies canpursue varying degrees of reiatedness betweendivisions. The degree of reiatedness waxes andwanes depending upon strategic circumstances.

Corporate value is createdexclusively at the business-unit level.

Corporate value is also created at the strategiclevel: portfolio structure can create shareholdervalue.

Corporate-divisional relation- Corporate-divisional relationships are signifi-ships are driven exclusively by cantly affected by corporate-level strategicbusiness-unit needs. considerations.

make a smaller, more flexible tester possible. When noneof the divisions would allocate the talent necessary to de-velop this potentially disruptive technology, d'Arbeloffintervened. He established a start up subsidiary that re-ported outside the regular control system to an internalboard of directors. As the product developed, he insistedthat the subsidiary make regular briefings to the other di-visions to share knowledge of the new technology. Whennew customers (minicontroller manufacturers) embracedthe tester-turning it into a hit-the sales force ralliedaround what had been an ugly duckling. Existing divi-sions recognized the need and opportunity to cooperatewith the new division and began to adapt the new chipand software technology.

If the corporate office can generate value by creatinga forward-looking portfolio, it follows that headquartersneeds to infiuence division-level decisions, especiallythose that might lead to interdivisional cooperation.Stated simply, there is a clear role for selective top-down,detail-oriented corporate management. And so crumblesthe third and final assumption about diversification -that

corporate-divisional relationships are driven exclusivelyby the needs of the divisions.

Global media giant Viacom illustrates the benefits ofan activist corporate office. In 1998, Paramount, Viacom'smovie studio, was negotiating European distributionagreements with the Kirsch Group, the leading Germanmedia conglomerate. With a $1 billion offer on the table.Paramount was anxious to close the deal. But MTV andNickelodeon (both part of Viacom) also wanted to estab-

lish positions in the German market.

Viacom's president encouraged the divi-sions to develop a collective position, anddivision managers spent several monthsat the end of 1998 attempting, but failing,to do so. Early in 1999, Sumner Redstone,Viacom's chairman and majority share-owner, intervened directly. He traveledto Europe, met with many of the keyplayers in the German market, and de-veloped an auction for the Paramountfilms. Several media companies, includ-ing Kirsch, bid on the distribution rights.When the contract was finally signed, itwas for $2 billion-and it included accessfor both MTV and Nickelodeon.

Like the executives at Sprint, WPP, andTeradyne, Redstone intervened basedon his unique corporate-level under-standing of the value of synergies be-tween divisions and the singular powerof the corporate office to effect integra-tion quickly. When he took action, he

was careful not to undermine the princi-ples of decentralized operating authority

that governed the divisions, which the company contin-ued to value. However, Redstone, like all these executives,understood that long-term strategic integration requiresselective, top-down intervention by the corporate officefrom time to time.

The differences between the more static view of diver-sification and the dynamic approach described here aresummarized in the exhibit "TWo Approaches to Manag-ing Diversified Companies." This comparison revealstbat, where traditional thinking was comparatively rigid,the new thinking is aligned with the dynamic needs ofcorporate-level strategic fiexihiiity.

Preparing for Dynamic MovesThe companies that will benefit most from a dynamic ap-proach to corporate strategy are those operating in highlyuncertain competitive environments in which, despitethe uncertainty, the need to make significant portfolio-level investments remains. It is no accident that the com-panies profiled here come from just such environments.



MANAGING DIVISIONS INSTABLE VERSUS UNCERTAIN MARKETS

The need for strategic flexibility in some

companies does not render traditional

frameworks for thinking about manag-

ing diversified companies invalid in

all competitive contexts. Those that will

benefit the most from strategic flexibility

are companies operating in uncertain

markets. But many well-managed and

respected diversified corporations will

continue to conform to traditional

approaches for the simple reason that

they remain entirely appropriate.

Consider, for example, AlliedSignal.

Acquired by Honeywell in 2000, which

in turn was acquired by GE in 2001,

AlliedSignal was a creature of the diversi-

fication wave of the late 1970S and early

1980s. Under the leadership of Larry

Bossidy from 1991 to 2000, AlliedSignal

was a paragon of the multibuslness en-

terprise, and there was very little about

the organization that suggested the kind

of strategic flexibility that characterizes

Sprint, WPF̂ Teradyne, and Viacom.

For instance, AlliedSignal made exten-

sive use of a group structure; corporate-

level executives were responsible for

managing synergies between divisions

withintheir groups. Consequently, the

company had an aerospace group, a plas-

tics group,a chemicals group,and soon.

Although grouping divisions in this way

improves the information-processing ca-

pacity of the organization as a whole, it

can also restrict the interactions between

divisions that are not in the same group.

The imposition of a hierarchical struc-

ture serves to filter information, embed

patterns of communication in reporting

relationships, and make it more difficult

for managers to see and act upon oppor-

tunities for cooperation between operat-

ing units. In a sense, then, a traditional

group structure serves to hardwire the

nature of the possible links within a di-

versified corporation.

This was a small price to pay for

AlliedSignal, since there was very little

uncertainty or variability in the nature

of interdivisional relationships. Simply

put, it was unlikely that the Prestone

antifreeze division in the consumer

products group would have much to

offer the engine systems division in the

aerospace group, so the increased effi-

ciency of the group structure more than

outweighed the cost of any lost flexibility.

At Sprint, by contrast, rapid changes

in the telecommunications industry con-

tinue to make it difficult to predict where

and when key linkages might emerge.

There are potentially compelling argu-

ments to be made for grouping long

distance with local service or bundling

wireless access with long-distance service

or integrating local service with wireless

access, and so on. With so many possibili-

ties and such uncertainty. Sprint cannot

afford to sacrifice its flexibility in execu-

tion, and so the company has avoided im-

posing a group structure. Consequently,

all of Sprint's operating divisions report

directly to the corporate office.

Compensation systems also reinforce

either traditional or flexible diversifica-

tion strategies. The compensation systems

at AlliedSignal reflected the stability of

the relationships between its divisions.

The compensation of managers whose

divisions were operationally linked was

tied to their joint performance. Compen-

sation structures for operating units that

had no material links with other operating

units contained no interunit compo-

nents. That approach is in stark contrast

to Sprint's compensation system, in

which every division's compensation

is tied to every other unit regardless of

whether links actually exist.

When it came to exploiting synergies

at AlliedSignal, corporate executives typi-

cally exerted pressure on the divisions,

but operating managers chose how to

respond to that pressure. In other words,

in keeping with traditional management

thinking, the corporate office took advan-

tage of its broader view to press for more

synergies where appropriate, but it re-

spected the more detailed knowledge

of those closest to the work and the cus-

tomers. Similarly, strategy formulation at

AlliedSignal required the corporate office

to provide financial targets but relied on

operating managers to develop specific

initiatives to meet those targets.

At Sprint, both of those processes

work quite differently. Operating divi-

sions are subject to strategic constraints

as part oftheir strategy formulation

process. This serves to create fiexibility.

But,toexp/o/t that flexibility, division

managers must take more than passing

notice of the detailed input provided by

corporate executives.

COMPARING TRADITIONAL AND FLEXIBLECORPORATE MANAGEMENT SYSTEMS

TRADITIONAL SYSTEMS FOR ASTABLE BUSINESS ENVIRONMENT

Extensive use of group structure

Stable compensation systems linkedto existing interdivisional synergies

Integration efforts typically identified bysenior management but designed andimplemented by operating management

The corporate office guides the formulationof strategy by imposing financial constraints

FLEXIBLE SYSTEMS FOR A RAPIDLYCHANGING BUSINESS ENVIRONMENT

Little or no use of group structure

Rapidly changing compensation systemsthat reward cooperative behavior, even ifthere are no current linkages

Integration efforts identified anddesigned to a large degree by corporatemanagement over the active resistanceof operating management

The corporate office guides the formulationof strategy by imposing strategic constraints

MAY 2001 99


At Sprint, competing successfully in the wireless businessmeant moving quickly to secure spectrum licenses andinvesting billions in infrastructure. If WPP was to includethe marquee advertising brands in its stable, it had to ac-quire those businesses when they became available. AndTeradyne's new business development efforts had to pro-ceed at least as quickly as the underlying technologyevolved.

In each case, portfolio diversification preceded the in-tegration efforts, which had always been a part of theoriginal strategic intent. Forced todiversify before meaningful syn-ergies can be captured, flexible cor-porations must maintain the com-petitiveness of each individualbusiness. To do so, they rely on thestrict financial controls and the di-visional autonomy typically seenin more static, unrelated diversi-fied companies. At the same time,if future integration is to remain apossibility, the corporate office

must impose strategic constraints,lest complete independence leadto the pursuit of division-level strategies that underminepossible future synergies. Coping with the disconnect be-tween expanding the corporation's portfolio and inte-grating divisions in pursuit of synergy requires the adop-tion of four management tactics.

Combine strict financial controls with a flexiblestructure. The powerful financial controls and planningassociated with traditional diversified companies must bein place to ensure that individual businesses maintainstrong performance results. However, a conventionalgroup structure is likely to be inappropriate. Group struc-tures typically make it more difficult for the corporateoffice to identify and assess the full range of possiblecross-divisional opportunities. Moreover, interdivisionalcooperation is usually hardwired by the structure, Hm-ited to divisions in the same group. Group structures alsotend to make it more difficult for corporate line leader-ship-not financial or planning staff-to communicateregularly with operating business managers about strate-gic questions. A more fiexible structure, coupled withtight financial controls, best serves a dynamic corporatestrategy. Compensation structures, too, must be flexibleenough to support frequent and significant changes instrategic priorities.

Be a player. If corporate leaders are going to contrib-ute to the substance of strategy, they need to be informed.That means being out in the market and in touch withcustomers, regulators, competitors. Wall Street analysts,and even academics. At Sprint, CEO Bill Esrey was a keyplayer in Washington as the telecommunications debateevolved. At Teradyne, Alex d'Arbeloff kept himself in-

CORPORATE AND DIVISION

EXECUTIVES SHOULD HAVE

FREQUENT CONVERSATIONS

THAT ARE NOT CLUTTERED UP

WITH OPERATIONAL ISSUES.

formed about cutting edge technology through an activerole in the venture capital community, extensive volun-teer activities at MIT, and regular attendance at industrymeetings. Sumner Redstone is notorious in his companyfor talking to anyone and everyone about the issues he be-lieves are important.

Have a lean but powerful corporate office. For allthe corporate-level activity, the corporate office sup-porting a dynamic strategy should be relatively small.With the exception of finance and human resources, most

staff positions should remain inthe divisions. The executives in thecorporate office are there to helpthe CEO; they usually haveworked closely with the CEO in thepast and gained his or her trust.Their importance lies in theirjudgment, not in their formalroles. At Sprint, the two corporateexecutives charged with driving in-terdivisional cooperation had ex-tensive operating experience at the

company. Each was supported bya staff of two or three people at

most. These few executives formed a kitchen cabinetrather than a structured corporate staff with assigned re-sponsibilities.

Spend time on strategy. Divisional executives can'twaste time when they are with the CEO. Monthly or quar-terly reviews of operations are necessary, but the focus ofthe executive meetings has to be the strategic opportuni-ties that markets offer, regardless of divisional lines. In ad-dition to formally scheduled meetings, corporate and di-vision executives should have frequent conversations thatare not cluttered up with operational issues. Operatingexecutives at Sprint and Viacom speak of being on thephone with their CEOs often. Martin Sorrell uses e-mailto stay in constant touch with his operating executives-and most anyone else who writes him.

Creating a truly dynamic corporate strategy goes far be-yond merely attempting to combine various existing ap-proaches. It is not enough for the corporate office to be byturns directive and standoffish. Nor should it attempt tobe both simultaneously. Rather, dynamic corporate strat-egy is something fundamentally different that brings withit a host of new management challenges. As the forcesof change impinge ever more sharply on an increasingrange of industries, we expect that more and more di-versified companies will benefit from thinking-and act-ing-in ways that create and exploit corporate-level stra-tegic fiexibility. v

Reprint R0105FTo order reprints, see the last page of Executive Summaries.


Documents

In the 1990s, conventional wisdom held that corporate ... · In the 1990s, conventional wisdom held that corporate leaders should leave divisions alone, ... Len Lauer, then president