Do You Have a Well-Designed Organization

TOOL KIT

Do You Have a

Well-DesignedOrganization?

Creating a new organizational structure

is one ofthe toughest-and most

politically explosive-challenges

that an executive faces. Here are

nine tests to guide the way.

by Michael Goold and Andrew Campbell

FOR MOST COMPANIES, Organiza-tion design is neither a sciencenor an art; it's an oxymoron. Or-

ganizational structures rarely resultfrom systematic, methodical planning.Rather, they evolve over time, in fits andstarts, shaped more by politics than bypolicies. The haphazard nature of theresulting structures is a source of con-stant frustration to senior executives.Strategic initiatives stall or go astray be-cause responsibilities are fragmented orunclear. Turf wars torpedo collabora-tion and knowledge sharing. Promisingopportunities die for lack of manage-rial attention. Overly complex struc-tures, such as matrix organizations, col-lapse because of lack of clarity aboutresponsibilities.

Most executives can sense when theirorganizations are not working well, butfew know how to correct the situation.A comprehensive redesign is just too

intimidating. For one thing, it's im-mensely complicated, involving an end-less stream of trade-offs and variables.For another, it's divisive, frequently dis-integrating into personality conflicts andpower plays. So when organization de-sign problems arise, managers often fo-cus on the most glaring flaws and, in theprocess, make the overall structure evenmore unwieldy and even less strategic.

What's been lacking is a practicalframework to guide executives throughthe complexities of organization design.That's what we aim to provide in thisarticle. We have reviewed the principlesof good design, studied the structures ofdozens of companies, large and small,and observed how executives go aboutmaking design decisions. We have en-capsulated our findings into nine testsof organization design, which can beused either to evaluate an existing struc-ture or to create a new one. The first four

MARCH 2002 117

TOOL KIT • Do You Have a Well-Designed Organization?

tests are what we call "fit" tests. Theyprovide an initial screen for design alter-natives, revealing whether the struc-tures support the company's strategy,talent pool, and situation. The next fiveare "good design" tests. They can helpa company refine a prospective designby addressing potential problem areas,including the balance between empow-erment and control. This set of tests helpsyou establish the right amount of hier-archy, control, and process-enough forthe design to work smoothly but not somuch as to dampen initiative, fiexibility,and networking.

Many of the tests, and their underly-ing principles, will sound famiMar. Theirpower stems not from their innovative-ness -we're not trying to promote a newtheory of business organization - butfrom their rigor and completeness. To-gether, they provide a company's man-agement with a structured approachfor analyzing all the key variables of or-ganizational success. Individual designdecisions will still be difficult, often re-quiring subjective judgments and hardtrade-offs, but using the framework willhelp make the debate more rational,shifting it away from issues of person-ality and toward issues of strategy andeffectiveness.

Getting the Fit Right

The Market AdvantageTest Does your designdirect sufficient manage-ment attention to yoursources of competitiveadvantage in each market?

In formulating a strategy, a companyhas to ask itself two fundamental ques-tions: Which markets should we com-pete in, and how will we gain an advan-tage over competitors in those markets?It may seem obvious that these ques-tions should also drive the company'sorganization design, but many struc-tures end up impeding market strategyrather than furthering it Some distrib-ute responsibilities in ways that distractthe management team's attention fromtarget customers. Others create divisions

among units that make it difficult forthem to operate in ways that providethe company with a competitive edge.The penalties of such misalignmentscan be enormous.

The first and most fundamental testof a design, therefore, is whether it fitsyour company's market strategy. Youshould begin by defiining your targetmarket segments. The definitions willvary depending on which part of yourorganization is being evaluated. If GE,for example, were designing its overallcorporate organization, it would usebroad definitions such as "aircraft en-gines" or "broadcasting." But if it werelooking only at the design of its financialservices unit, it would use much nar-rower definitions, probably combiningparticular service lines with particulargeographic markets: "aircraft leasing inEurope," for instance, or "receivablesfinancing in Mexico." There should beno dispute about the relevant marketsegments; if there is, you need to dosome fr'esh strategy thinking before youproceed with the design effort.

. Next, determine whether the designdirects enough attention to each marketsegment. Here's our rule of thumb: If asingle unit is dedicated to a single seg-ment, the segment is receiving suffi-cient attention. If no unit has respon-sibility for the segment, the design isfatally flawed and needs to be revamped.Often, the analysis is not so clear; a unitmay have responsibility for a numberof segments. (This is often the case withsmall, but rapidly growing, market seg-ments.) Yoy'will need to evaluate suchsituations carefully, making judgmentsabout whether the division of responsi-bilities will allow sufficient attention tobe focused on the segment.

It's also important to determinewhether the design supports your keysources of advantage (speedy introduc-tion of products, for example, or low-cost manufacturing) and related oper-

ating initiatives (product launches, fac-tory automation). List these sources andinitiatives, and check how the designaddresses them. In a perfect world, youwould have a single unit, or department,dedicated to each source and initiative.In reality, however, market advantagesoften require coordination across units.For instance, your source of advantagein one segment may be superior new-product development. To achieve thatadvantage, the business unit responsi-ble for the segment may need to collab-orate with a central research function.Or your advantage may be an economyof scale in manufacturing that requirescoordinated production across numer-ous business units.

Because collaboration across units isalways more difficult to manage thancollaboration within units, any sourceof advantage that requires cross-borderlinks - particularly complex ones - shouldbe a cause for concern. You'll need tobe confident that the design will enablethe unit managers to give sufficient at-tention to maintaining the links. Somecompromises may remain; they'll befurther analyzed by the good-designtests below.

The Parenting AdvantageTest Does your design helpthe corporate parent addvalue to the organization?

Just as parents play varying roles infamilies, corporate headquarters playvarying roles in different companies.The focus of this test is to make sure theorganizational design is tailored to sup-port these roles. First, explicitly defineand list your company's "parentingpropositions" -the corporate-level activ-ities that provide real value to the over-all company. The propositions mightinvolve narrow tasks-for example, man-aging government relations-or broad

Michael Goold and Andrew Campbell are directors at the Ashridge Strategic Man-agement Centre in London. They are the authors o/Deslgnlng Effective Organizations:How to Create Structured Networks (Jossey-Bass, 2002) and The CollaborativeEnterprise: Why Links Between Business Units Often Fail and How to Make ThemWork (Perseus, 1999) as well as several HBR articles.

118 HARVARD BUSINESS REVIEW

Do You Have a Well-Designed Organization? • TOOL KIT

coordination roles, such as maintainingstrong research capabilities across allunits. Or they might entail specific initia-tives, such as implementing a company-wide ERP system. (See the exhibit "HowParents Create Value.")

Next, determine whether the designgives sufficient attention to these value-adding tasks and initiatives. If, for exam-ple, one of the parent's key roles is en-couraging iaiowledge sharing among aparticular group of units, it's importantto ask whether there is a manager in theparent unit focused on that task. You'llalso need to look hard at the organiza-tional links among those units. If theunits are located in different divisions, itmay make sense to change the design sothat they become members of the samedivision, making collaboration mucheasier. Sometimes, this test will highlightdifficult trade-offs that need to be made.If one of the parenting propositions is tospurhigh-speed innovation, for instance,you wiU need to decide whether it makesmore sense to centralize R&D in a cor-porate unit or disperse it in the businessunits, which are closer to the market.

The parenting advantage test canhelp companies see more clearly the

organizational implications of theirstrategies, as agriculture giant Cargillrecently discovered. One of the mostimportant parenting propositions ofCargiU's headquarters was encouraging

a greater focus on broad customer solu-tions rather than on individual prod-ucts. When top management viewed theorganization in this light, it saw that cer-tain fundamental changes were needed.

How Parents Create Value

To be effective, parent units need to think through the ways in whichthey can create value or add value to the rest of the organization.We call these sources of added value "parenting propositions," and,in general, they fall into five categories:

Select PcopositionsThe parent unit creates value by acquiring units or people for less thanthey are worth or disposing of activities for more than they are worth.

Build PropositionsThe parent unit helps units expand their size and scope of activityby, for example, helping with globalization or product extensions.

Stretch PropositionsThe parent unit helps units improve costs, quality, or profitability by,for instance, setting stretch targets or providing benchmarks.

Link PropositionsThe parent unit helps units work together in ways they would finddifficult if left to themselves. For example, it might centralize activitiesor alter incentives.

Leverage PropositionsThe parent unit finds ways to exploit a central resource, such as a brand,a relationship, a skill, or a patent, in new markets or new businesses.

MARCH 2002 119


Cargill created new, more market-focused business units, and it groupedthem together into broad "platforms"with management teams that could pro-mote a coordinated approach to cus-tomer relationships and solutions. TheFood Applications platform, for exam-ple, brought together all of CargiU'sbusinesses that sold products to foodmanufacturers; the businesses that dealtwith fanners became the Farmgate plat-form. The exercise enabled Cargill tomore clearly define its parenting prop-ositions and create an organization thatsupported them.

The People Test.Does your design reflect thestrengths, weaknesses, andmotivations of your people?

When an organization doesn't workright, executives are often quick toblame "people problems." But that'swrongheaded. If an organization is notsuited to the skills and attitudes of itsmembers, the problem lies with the de-sign, not the people. For this test, firstlook at your key players-the membersof the top management team and otherindividuals deemed critical to the busi-ness. For each, ask whether the designprovides the appropriate responsibili-ties and reporting relationships andwins their commitment. If, for example,your CEO is a marketing type and thedesign focuses her attention on perfor-mance management, you've got a prob-lem. If your CFO is a hands-on, detail-oriented guy and your design has thetop finance manager in each businessunit reporting to the unit head insteadof to him, you're setting yourself up forbig conflicts.

Now look at the pivotal jobs in the de-sign-the positions that will need to bestaffed by highly talented people if theorganization is to work well. Typically,these will include the heads of all keybusiness units and the managers of allfunctions involved in critical cross-unitrelationships. Do you have outstandingpeople to staff these jobs today? Do youhave the career paths and development

initiatives needed to create and retainnew talent for tomorrow? If you had tofind replacements outside, would yoube able to attract and hire them? A de-sign that cannot be staffed with compe-tent managers should be abandoned.

If you're creating a new structure, youals(.> need to look at the losers-the em-ployees who will forfeit status or powerin the revamped organization. Ail re-designs create losers, and losers can turncynical and resistant, becoming road-blocks to change. You need to make twodifficult judgments. First, determinewhich of the losers are influential. Thendecide how to deal with them-eitherbuying their support through addedcompensation or neutralizing their in-fluence by changing their roles or lettingthem go.

i The Feasibility Test.Have you taken account ofall the constraints that mayimpede the implementationof your design?

All companies have constraints on theirability to act. Some constraints, such aslaws, are external. Others, such as infor-mation systems, are internal. Becausethey can impede or even block certainorganizational changes, such constraintsneed to be identified and assessed earlyin any design effort. Look, in particular,at four categories:

• Government regulations are rarelya dominant influence on organizationalstructure, but they can preclude certaindesign elements. In some countries, forinstance, it is impossible to do businesswithout setting up a separate joint ven-ture with a local partner. And in the util-ities industry, regulators often insist onkeeping regulated and nonregulatedbusiness activities in separate units.Companies with international opera-tions need to be particularly cognizantof legal issues.

• The interests of a company's stake-holders should be considered carefully.Large shareholders often need to signoff on major organizational moves, andstock markets may impose rules-such

as restrictions on minority investors-that have organizational implications.In some companies, other stakeholders,such as trade unions, will demand avoice in decisions.

• A company's information systemsmay prevent certain organizationalchanges. You may, for example, want tomove from a country-based to a product-based structure, but if your systems areunable to report performance by prod-uct, you will either have to retool them,at considerable cost, or seek anotherdesign option.

• Corporate cultures can limit thefeasibility of design choices. Executivesfrequently discover this in postmergerintegration situations, as was the caseat DaimlerChrysler. Try to identify theroot causes of cultural constraints. Forexample, if your company boasts ofa strong performance culture but haspoor interunit coordination, the causemay be an incentive system that pro-vides no rewards for cooperation. If yourredesign relies on cross-unit processes,you'll have to change the incentive sys-tem for it to work.

Refining the Design

The Specialist Cultures Test.Does your design protectunits that need distinctcultures?

In most companies, there are certainunits that should maintain distinct cul-tures. They need to think and work inways that are different from the pre-vailing organizational norms. Examplesmight include new-product develop-ment teams, e-business groups, or func-tional service units. In evaluating thestrength of an organization design, youneed to make sure that such "specialistcultures" are sufficiently insulated fromthe rest of the organization.

Once you've identified the specialistcultures in your company, ask yourselfwhether any of them is in danger ofbeing dominated. Look, in particular, atsister units and the parent unit to whichthe specialist culture unit reports. Ifthese other units share a culture that's


Do You Have a Weil-Designed Organization? • TOOL KIT

different from the specialist culture, youcan assume that the specialist culture isat risk of being "contaminated" by thedominant culture. Let's say that a bigchemical company has housed a spe-cialty chemicals unit in a division withmany bulk chemicals units, and thedivisional headquarters is staffed mainlyby bulk chemicals executives. That's apotential problem to which a solutionmust be found. Dow Chemical, for ex-ample, realized that its specialty chem-icals units were in danger of being dom-inated by the bulk chemicals cultures ofDow's large integrated sites. So whenthe company reorganized, it groupedall the specialty chemicals units in a newdivision, separating them from the bulkchemicals units.

When you find a specialist culturethat is at risk, first look for ways to pro-tect it without changing the basic struc-ture. You might, for example, put a high-ranking corporate executive in charge ofthe unit, giving it the standing neces-sary to resist external influences. Or youmight grant the unit greater autonomy,freeing it, say, from corporate human-resources policies. Or you might try tosolve the problem through communi-cation-educating the rest of the com-pany about the unique goals and re-quirements of the specialist unit. If youcan't come up with protective measures,you'll need to alter the design, possiblyby rethinking the way you organizeyour business units.

The Difficult-Links Test.Does your design providecoordination soiutionsforthe unit-to-unit links tliatare likeiy to be probiematic?

However you define your units, somecollaboration among them is likely tobe necessary. (See the exhibit "HowUnits Connect.") We have found that thelarge majority of these links are besthandled through self-managed net-working among the units. Wheneverpossible, top management should leavethis up to the units rather than imposetop-down coordination processes. But

this may not be sufficient for links thatpose particular challenges. Such "diffi-cult links," as we call them, arise formany reasons. Managers in differentunits may not perceive the benefits ofcollaborating, they may have no incen-tives to work together, or they may sim-ply lack the skills and resources to makethe necessary cooperation happen. {Fora fuller discussion, see "DesperatelySeeking Synergy,"by Michael Goold andAndrew Campbell, HBR September-October 1998.) Whatever the cause, dif-ficult links call for specially designedsolutions, such as a clearly defined arbi-tration process for resolving disputes.

One example of a difficult link is col-laboration on pricing decisions betweenproduct-based units and customer-basedunits. In Citibank's Corporate BankingGroup, for example, customer-basedunits, defined around industry group-ings, must work together with productspecialists in areas such as cash man-agement, foreign exchange, and struc-

tured finance. The customer units maywant to discount certain products to re-inforce customer relationships, whilethe product units' motivation is to pre-serve product-specific margins and prof-itability. A conflict of this sort can onlybe resolved by designing detailed coor-dination solutions. For example, a pro-cess for reaching agreement on pricingdecisions, specifying who has ultimateauthority, can be established. Alterna-tively, incentives and performance mea-sures can be amended to align the in-terests of managers in the product andcustomer units more closely,or a group-level manager can be given authority toarbitrate and resolve disputes.

Refinements to the basic structuremay not always be sufficient to resolve adifficult link. A typical example is whenthe link will lead to hard-to-compensatelosses for one or more of the units (as ina facilities rationalization program).Substantial redesign, such as bringingthe units involved into the same division

How Units Connect

Some of the biggest organizational challenges involve coordinationamong units. In evaluating a design, you need to pay dose heed tounit-to-unit links, which take six basic forms:

Shared Know-How LinksSharing best practices, leveraging expertise in functional areas, poolingknowledge about how to succeed in specific geographic regions, orsharing product or market know-how

Shared Tangible Resources LinksCreating economies of scale and eliminating duplicated effort throughthe sharing of physical assets (such as an R&D lab) or resources (suchas people)

Pooled Negotiating Power LinksGenerating economies of scale or other benefits through commonpurchases or joint negotiation with stakeholders (such as customers,governments, and universities)

Coordinated Strategies LinksAligning the strategies of two or more units by, for example, coordinat-ing responses to a new competitor

Vertical Integration LinksCoordinating the flow of products or services from one unit to anotherin order to reduce inventory costs, enhance product development,increase capacity utilization, or improve market access

New-Business Creation LinksFashioning new businesses by combining know-how from differentunits through teams, internal joint ventures, or other alliances

MARCH 2002 121

Harvard Business School ExecutiveEducation offers a wide varietyof programs to allovkf exceptionalindividuals and organizations tothrive in today's global businessenvironment. Each year, we offerover 40 programs that range induration from a few days to10 weeks, including:

trategyJune 16-22, 2002

> Managing Brand MeaningJune 23-27. 2002

> leading Product DevelopmentJuly 7-12, 2002

> Leading Professional Service FirmsJuly 7-13, 2002

> Finance for Senior ExecutivesJuly 14-20, 2002

> Strategic Perspectives inNonprofit ManagementJuly 14-20. 2002

> Delivering Information ServicesJuly 21-August 2, 2002

> Driving Corporate Performance:From Scorekeeping to StrategyJuly 21-26, 2002

> Making Corporate BoardsMore EffectiveJuly 21-24, 2002

> Building Competitive AdvantageThrough OperationsJuly 28-August 2, 2002

FOR MORE INFORMATIONE-maiit [email protected]: 1-800-HBS-5577, ext. 4093(outside the U.S., call+1-617-495-6SS5, ext. 4093)Or visit our Web site at

For a window into theintellectual capital of HBS and

other expertise from around the world,V I S I T >

or setting up a new dedicated unit,may then be needed. IBM, for example,found it necessary to set up a GlobalServices Division in order to achieve suf-ficient integration in the provision ofservices to its customers.

Taken together, the specialist culturestest and the difficult-links test can guidetop management's judgments abouthow narrowly or broadly to define unitresponsibilities. They allow the design tostrike a suitable balance between spe-cialization and coordination.

The Redundant-HierarchyTest. Does your design havetoo many parent levels andunits?

Big companies can have many parentunits, including corporate headquartersand various groups that oversee line di-visions, corporate functions, and geo-graphic regions. Some of these unitsmay be very small, consisting of just aline manager and a finance executive.Others may be large and complex, en-compassing many staff members. Inevaluating an organization design, ourbasic presumption is in favor of decen-tralizing decisions to frontline units andretaining decisions at upper levels onlyif those levels can add value. Therefore,it's important to determine whethereach parent level is needed and, if so,whether it has the resources necessaryto do its job.

First, identify each level and unit inyour corporate hierarchy above the op-erating units. Then ask yourself whethereach has dear and distinct parentingpropositions (as described in the parent-ing advantage test). If a level's proposi-tions echo those of the level above orbelow it, one of the levels may be re-dundant, and you should think hardabout removing it. To cover the inevi-table costs and drawbacks of an extralayer, we believe it needs to be able toimprove the performance of the unitsreporting to it by at least io%. That ruleof thumb can make it easier to deter-mine whether levels are worth keeping.It's a powerful argument against having

many layers: If there are, say, three lay-ers above the business units, the totalparenting added value needs to be atleast 30% to justify the layers!

You now need to determine whetherevery level with compelling parentingpropositions has access to the skills andresources it needs. Let's say, for example,that you have a division responsible forall businesses in Europe, and its parent-ing proposition is to coordinate man-ufacturing and customer service andintegrate back-office functions acrosscountries. To succeed, the division willrequire deep functional expertise inmanufacturing, major-account manage-ment, and information systems. If itlacks those skills, you will need to de-velop them or alter some roles and re-sponsibilities. For instance, the coordi-nation of information systems may needto be pushed down to the country-levelbusiness units, which could together or-ganize a pan-European council of sys-tems managers.

This test does not require exhaustiveanalysis or fine-grained judgments. Theobject is to spot major problems, notminor ones. But even when conductedat a high level, the redundant-hierarchytest can provide powerful insights. Manycompanies that have used it have foundways to cut out layers of management,shrink corporate and divisional func-tions, and refocus corporate and divi-sional managers so that they add greatervalue to the company.

iThe Accountability Test.Does your design supporteffective controls?

hbsworkingknowledge.hbsedu

In decentralized organizations, account-ability for performance is important.The purpose of this test is to ensure thatevery unit has appropriate controls overits performance-controls that suit itsresponsibilities, are economical to im-plement, and motivate managers.

In assessing accountability, focusparticularly on two common sources ofproblems. First, look at any units withshared responsibilities, especially if theircollaboration is mandatory. A company

HARVARD BUSINESS REVIEW

H A R V A R D B U S I N E S S S C H O O L

with both global business units and na-tional operating companies, for exam-ple, may require global product heads toreach consensus with operating com-pany executives on issues such as majorinvestments and profit targets. Or busi-ness units may be required to use theservices provided by corporate IT or HRdepartments. Whenever shared respon-sibilities are imposed by the corporateparent in this way, it dampens account-ability. It becomes easy for units toblame performance problems on oneanother and difficult for senior execu-tives to determine who's at fault

Second, focus on any units whose per-formance is difficult to measure, eitherbecause there are no objective outcomesfor comparison or because of the ex-pense of collecting performance data.A corporate unit doing basic research isa typical example of a unit for which it'shard, if not impossible, to come up withclear-cut, easy-to-collect performancemeasures.

Where possible, remedies should beput in place for units with blurred re-sponsibilities or fuzzy measures. Clearerperformance measures should be devel-oped, and greater clarity should be pro-vided for overlapping responsibilities.Often, however, full solutions are notpossible. In such cases, parent managerswill have no choice but to rely on sub-jective judgments about performance.This usually makes the control processunsatisfactory unless the parent man-agers have a deep operating knowledgeof the units they oversee. If such man-agers are unavailable, you may need tomodify the design.

f The Flexibility Test. Doesyour design facilitate thedevelopment of new strategiesand provide the flexibilityrequired to adapt to change?

A well-designed organization is fiexi-ble for the future as well as fit for thepresent. It provides ways for a companyto pursue innovation and allows foradaptability to changing circumstances.Of course, it takes more than a good

MARCH 2002

t'll come back to you.DEFINING EXPERIENCE: "Being challenged to effectively

interact with—and win the respect of—a group ofpeople from amazingly diverse global cultures."

FIRST ENCOUNTERED: Spring 1998, HBSAdvanced Management Program

LATER USED: "On a daily basis—helping a $100 billion

multinational organization create responsible social programs

in the 84 countri

VP Global

E X E C U T I V E E D U C A T I O N

Time ana "transformation" that occurs during

Some call it a "defining experience" in their careers. Rigorous intensity.Passionate faculty. State-of-the-art facilities. Exceptional colleagues. Alladd up to a singular ability to bolster confidence, refine capabilities, andprovide ac t ionab le learn ing. These programs inspire important newperspectives and provide the opportunity to establish lasting connectionswith colleagues from around the world. What's more, a relationship with HBSbecomes a lifelong career asset—with perennial access to the latest businessideas and special opportunities for alumni. Visit our Web site to learn more.


organization design to ensure innova-tion and flexibility - it also requires flex-ible minds, deep talent, and robust re-

Getting the Bugs Out

when you identify a problem with

your design, first look for ways to fix

it without substantially altering it. If

that doesn't work, you'll have to make

fundamental changes or even reject

the design. Here's a step-by-step

process for resolving problems:

Steps not involving major design change

Modi fy without changing the units.

• Refine the allocation of responsibili-ties (for example, clarify powers andresponsibilities).

• Refine reporting relationships andprocesses (for example, define newparenting propositions).

• Refine lateral relationships and pro-cesses (for example, define coordina-tion mechanisms).

• Refine accountabilities (for example,define more appropriate performancemeasures).

Rede f i ne skill requirementsand incentives.• Modify criteria for selecting people.

• Redefine skill development needs.

• Develop incentives.

Shape informal context.

• Clarify the leadership styie needed.

• Define normsof behavior, values, orsocial context.

Steps Involving major design change

M a k e substantial changes in theunits.

• Make major adjustments to unitboundaries.

• Change unit roles (for example, turnfunctional units into business unitsor shared services).

• Introduce new units or merge units.

Change the structure.

• Change reporting lines.

• Create new divisions.

sources. This test is therefore aimed notat determining whether the companyhas all the capabilities required to inno-vate and adapt, but whether there willbe any major organizational roadblocksalong the path to the future.

Start by assembling a group of man-agers from across lthe company whohave deep knowledge of products andmarkets. Ask them to create a list oftenor so future opportunities, including acouple of off-the-wall ones, that are notanticipated in the current strategy butthat are representative of the oppor-tunities the company is likely to en-counter in the future. Now look at yourdesign and see whether it would sup-port or impede the pursuit of each op-portunity. This is, of course, a tough task.It's not easy to predict how groups ofpeople will act in the face of hypotheti-cal challenges. But it is possible to get arough sense of whether the organiza-tion will be supportive, neutral, or ob-structive and to think of changes thatwill reduce any obstructions. For exam-ple, in the mid-1990s, Emap, the maga-zine publisher, listed new media as oneof its future opportunities. Recognizingthat its magazine-focused business unitswould he reluctant to cannibalize theirexisting franchises, top managementcreated business development functionsat the division level to ensure that theseopportunities were not overlooked.

If you find that your design could ob-struct the pursuit of opportunities, ex-plore possible modifications to it. Youmight, for instance, set aside seed capi-tal for new products or revamp yourstrategic planning process to force man-agers to spend time thinking about po-tential new businesses. Keep in mind,however, that the opportunities arespeculative. You don't want to makechanges that will render your organiza-tion less able to fulfill its current busi-ness imperatives.

Follow a similar approach in examin-ing flexibility. First, identify five or tenmajor organizational changes that maybe required over the next three to fiveyears. Then, identify any parts of theorganization that would be resistant tothe changes. Pockets of resistance can

often be found in the top managementlayer, with its rigid personal loyaltiesand entrenched fiefdoms. It can alsooccur because of tightly integrated setsofunits,with complex relationships andpolicies that are difficult to untangle. Ifpush came to shove, would you be ableto mitigate these obstacles-by, for ex-ample, redistributing top managers' re-sponsibilities or moving to more inde-pendent and self-contained businessunits? If you find that the risk of inertiais too high, you'll need to think aboutaltering the design.

An Iterative ProcessOnce you've gone through all the tests,you'll probably have made a number ofminor, and perhaps some major, changesto your design. (For a summary of theproblem-solving process, see the exhibit"Getting the Bugs Out.") You'll want torun through the tests again to ensurethat the changes made to pass one testhaven't caused the design to fail any ofthe others. Let's say, for example, that acompany finds a difficult link betweenits business units and one of its sharedcorporate functions. To fix the problem,it decides that the function should berun as an informal joint venture oper-ated by all the business units. But whilegoing back through the tests, the com-pany realizes that the business unitsshare a single culture that's very differ-ent from the one in the corporate func-tion. As a result, the design now fails thespecialist cultures test. The companyneeds to find another way to solve thedifficult-link problem.

The iterative nature ofthe tests is oneof their great strengths. Organizationaldecisions are inevitably complex, andtweaking one area may produce unan-ticipated consequences in a very differ-ent area. To get the best design, youneed to take the broad view, workingstep-by-step through the myriad trade-offs. It's not an easy process, but it is amanageable one. And the alternative -waiting for design flaws to turn into dis-asters - is far worse. v

Reprint R0203KTo order reprints, see tbe last pageof Executive Summaries.


Documents

Do You Have a Well-Designed Organization