MAKE VERSUS BUY

Strategic
By assessing the relative costs and risks of making or buying, companies can leverage their skills and resources for increased protability James Brian Quinn Frederick G Hilmer

approaches, when properly combined, allow managers to leverage their companies skills and resources well beyond levels available with other strategies:
WO NEW STRATEGIC

Concentrate the rms own resources on a set of core competencies where it can achieve denable preeminence and provide unique value for customers.1 Strategically outsource other activities including many traditionally considered integral to any company for which the rm has neither a critical strategic need nor special capabilities.2 The benets of successfully combining the two approaches are signicant. Managers leverage their companys resources in four ways. First, they maximize returns on internal resources by concentrating investments and energies on what the enterprise does best. Second, well-developed core competencies provide formidable barriers against present and future competitors that seek to expand into the companys areas of interest, thus facilitating and protecting the strategic advantages of market share. Third, perhaps the greatest leverage of all is the full utilization of external suppliers investments, innovations, and specialized professional capabilities that

James Brian Quinn is Buchanan Professor of Management at the Amos Tuck School of Business Administration, Dartmouth College. Frederick G. Hilmer is Dean of the Australian Graduate School of Management, University of New South Wales. This article is reprinted by permission of the publisher from the Sloan Management Review, Summer 1994. Copyright 1994 Sloan Management Review Association. All rights reserved.
1

For notes, see page 70.

48

THE McKINSEY QUARTERLY 1995 NUMBER 1

outsourcing
would be prohibitively expensive or even impossible to duplicate internally. Fourth, in rapidly changing marketplaces and technological situations, this joint strategy decreases risks, shortens cycle times, lowers investments, and creates better responsiveness to customer needs. Two examples from our studies of Australian and US companies illustrate our point: Nike, Inc. is the largest supplier of athletic shoes in the world. Yet it outsources 100 percent of its shoe production and manufactures only key technical components of its Nike Air system. Athletic footwear is technology- and fashion-intensive, requiring high exibility at both the production and marketing levels. Nike creates maximum value by concentrating on preproduction (research and development) and postproduction activities (marketing, distribution, and sales), linked together by perhaps the best marketing information system in the industry. Using a carefully developed, on-site expatriate program to coordinate its foreign-based suppliers, Nike even outsourced the advertising component of its marketing program to Wieden & Kennedy, whose creative eforts drove Nike to the top of the product recognition scale. Nike grew at a compounded 20 percent growth rate and earned a 31 percent ROE for its shareholders through most of the past decade. Knowing it could not be the best at making chips, boxes, monitors, cables, keyboards, and the like for its explosively successful Apple II, Apple Computer outsourced 70 percent of its manufacturing costs and components. Instead of building internal bureaucracies where it had no unique skills, Apple outsourced critical items like design (to Frogdesign), printers (to Tokyo Electric), and even key elements of marketing (to Regis McKenna, which achieved a $100 million image for Apple when it had only a few employees and about $1 million to spend).

Reprinted from the Sloan Management Review

THE McKINSEY QUARTERLY 1995 NUMBER 1

49

STRATEGIC OUTSOURCING

Apple focused its internal resources on its own Apple DOS (disk operating system) and the supporting macro sotware to give Apple products their unique look and feel. Its open architecture policy stimulated independent developers to write the much-needed sotware that gave the Apple IIs customers uniquely high functionality. Apple thus avoided unnecessary investments, beneted from its vendors R&D and technical expertise, kept itself How can managers combine exible to adopt new technologies as they core competency concepts became available, and leveraged its limited and strategic outsourcing for capital resources to a huge extent. Operating maximum efectiveness? with an extremely at organization, Apple enjoyed three times the capital turnover and the highest market value versus xed investment ratio among major computer producers throughout the 1980s.3 How can managers combine core competency concepts and strategic outsourcing for maximum efectiveness? To achieve benets like Nikes or Apples requires careful attention to several dicult issues, each of which we discuss in turn: 1. What exactly is a core competency? Unfortunately, most of the literature on this subject is tautological core equals key or critical or fundamental. How can managers analytically select and develop the core competencies that will provide the rms uniqueness, competitive edge, and basis of value creation for the future? 2. Granting that the competencies dening the rm and its essential reasons for existence should be kept in-house, should all else be outsourced? In most cases, common sense and theory suggest a clear no. How then can managers determine strategically, rather than in a shortterm or ad hoc fashion, which activities to maintain internally and which to outsource? 3. How can managers assess the relative risks and benets of outsourcing in particular situations? And how can they contain critical risks especially the potential loss of crucial skills or control over the companys future directions when outsourcing is desirable?

Core competency strategies

The basic ideas behind core competencies and strategic outsourcing have been well supported by research extending over a twenty-year period.4 In 1974, Rumelt noted that neither of the then-favored strategies unrelated diversication or vertical integration yielded consistently high returns.5 Since then, other carefully structured research has indicated the efectiveness

50

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

of disaggregation strategies in many industries.6 Noting the failures of many conglomerates in the 1960s and 1970s, both nancial theorists and investors began to support more focused company concepts. Generally this meant sticking to your knitting by cutting back to fewer product lines. Unfortunately, this also meant a concomitant increase in the systematic risk these narrower markets represented. However, some analysts noticed that many highly successful Japanese and American companies had very wide product lines, yet were neither conglomerates nor truly vertically integrated.7 Japanese companies, like Sony, Mitsubishi, Matsushita, or Yamaha, had extremely diverse product oferings, as did 3M or Hewlett-Packard in the United States. Yet they were not conglomerates in the normal sense. They were termed related conglomerates, redeploying certain key skills from market to market.8 At the same time, these companies also contracted out signicant support activities. Although frequently considered vertically integrated, the Japanese auto industry, for example, was structured around mother companies that primarily performed design and assembly, with a number of independent suppliers and alliance partners without ownership bonds to the mother companies feeding into them.9 Many other Japanese hi-tech companies, particularly the more innovative ones like Sony and Honda, used comparable strategies leveraging a few core skills against multiple markets through extensive outsourcing. The term core competency strategies was later used to describe these and other less diversied strategies developed around a central set of corporate skills.10 However, there has been little theory or consistency in the literature about what core really means. Consequently, many executives have Managers need to think in been understandably confused about the terms of the specic skills the topic. They need not be if they think in company must have to create terms of the specic skills the company has unique value for customers or must have to create unique value for customers. However, their analyses must go well beyond looking at traditional product or functional strategies to the fundamentals of what the company can do better than anyone else.11 For example, ater some dicult times, it was easy enough for a beer company like Fosters to decide that it should not be in the nance, forest products, and pastoral businesses into which it had diversied. It has now divested these peripheral businesses and is concentrating on beer. However, even within this concept, Fosters true competencies are in brewing and marketing beer. Many of its distribution, transportation, and can production activities, for example, might actually be more efectively contracted out.

THE McKINSEY QUARTERLY 1995 NUMBER 1

51

STRATEGIC OUTSOURCING

Within individual functions like production, Fosters could further extend its competitive advantage by outsourcing selected activities such as maintenance or computing where it has no unique capabilities.

The essence of core competencies

What then is really core? And why? The concept requires that managers think much more carefully about which of the rms activities really do or could create unique value and which activities managers could more efectively buy externally. Careful study of both successful and unsuccessful corporate examples suggests that efective core competencies are: 1. Skill or knowledge sets, not products or functions. Executives need to look beyond the companys products to the intellectual skills or management systems that actually create a maintainable competitive edge. Products, even those with valuable legal protection, can be too easily backengineered, duplicated, or replaced by substitutes. Nor is a competency typically one of the traditional functions such as production, engineering, sales, or nance, around which organizations were formed in the past. Instead, competencies tend to be sets of skills that cut across traditional functions. This interaction allows the organization consistently to perform an activity better than functional competitors and continually to improve on the activity as markets, technology, and competition evolve. Competencies thus involve activities such as product or service design, technology creation, customer service, or logistics that tend to be based on knowledge rather than on ownership of assets or intellectual property per se. Knowledge-based activities Competencies involve activities generate most of the value in services and that tend to be based on manufacturing.

knowledge rather than on ownership of assets or intellectual property per se

In services, which account for 79 percent of all jobs and 76 percent of all value-added in the United States, intellectual inputs create virtually all of the value-added. Banking, nancial services, advertising, consulting, accounting, retailing, wholesaling, education, entertainment, communications, and health care are clear examples. In manufacturing, knowledge-based activities like R&D, product design, process design, logistics, marketing research, marketing, advertising, distribution, and customer service also dominate the value-added chain of most companies (see Exhibit 1). 2. Flexible, long-term platforms capable of adaptation or evolution. Too many companies try to focus on the narrow areas where they currently excel, usually on some product-oriented skills. The real challenge is to

52

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

consciously build dominating skills in areas that the customer will continue to value over time, as Motorola is doing with its focus on superior quality, portable communications. The uniqueness of Toys R Us lies in its powerful information and distribution systems for toys, and that of State Street Boston in its advanced information and management systems for large custodial accounts. Problems occur when managers choose to concentrate too narrowly on products (as computer companies did on hardware) or too inexibly on formats and skills that no longer match customer needs (as FotoMat and numerous department stores did). Flexible skill sets and constant, conscious reassessment of trends are hallmarks of successful core competency strategies.

Exhibit 1

Make or buy?
When to outsource Corporate staff services Regulatory affairs Public relations

Basic research

Maintenance

Data center

Accounting

Personnel

Example Potential provider Internal

Low ($)

Potential cost reduction

Cost Value

Finance

Legal

Potential value gain Benchmark

High ($)

AMEX Andersen Banc One DEC EDS IBM

When to insource Value chain services Plant engineering Applied research Market research Manufacturing Quality control Product design Process design

Warehousing

Distribution

Advertising

Marketing

Logistics

3. Limited in number. Most companies target two or three (not one Example and rarely more than ve) activities Potential Potential Potential in the value chain most critical to provider cost reduction value gain Cost Value future success. For example, 3M Internal concentrates on four critical techLow ($) High ($) Company A nologies in great depth and supports Benchmark Company B these with a peerless innovation Company C system. As work becomes more complex, and the opportunities to excel in many detailed activities proliferate, managers nd they cannot be best at every activity in the value chain. As they go beyond three to ve activities or skill sets, they are unable to match the performance of their more focused competitors or suppliers. Each skill set requires intensity and management dedication that cannot tolerate dilution. It is hard to imagine Microsots top managers taking their enthusiasm and skills in sotware into, say, chip design or even large-scale training in sotware usage. And if they did, what would be the cost of their loss of attention on sotware development?

THE McKINSEY QUARTERLY 1995 NUMBER 1

53

Service

Repair

Sales

STRATEGIC OUTSOURCING

4. Unique sources of leverage in the value chain. Efective strategies seek out places where there are market imperfections or knowledge gaps that the company is uniquely qualied to ll and where investments in intellectual resources can be highly leveraged. Raychem and Intel concentrate on depth in design and on highly specialized test-feedback systems supporting carefully selected knowledge-based products not on volume production of standardized products to jump over the experience curve advantages of their larger competitors. Morgan Stanley, through its TAPS system, and Bear Stearns, through its integrated bond-trading programs, have developed in-depth knowledge bases creating unique intellectual advantages and protability in their highly competitive markets. 5. Areas where the company can dominate. Companies consistently make more money than their competitors only if they can perform some activities which are important to customers more efectively than anyone else. True focus in strategy means the capacity to bring more power to bear on a selected sector than any competitor can. Once, this meant owning and managing all the elements in the value chain supporting a specic product or service in a selected market position. Today, however, some outside supplier, by specializing in the specic skills and technologies underlying a single element in the value chain, can become more procient at that activity than virtually any company spreading its eforts over the True focus in strategy means whole value chain. In essence, each company is in competition with all potential suppliers of each activity in its value chain. Hence, it must benchmark its selected core competencies against all other potential suppliers of that activity and continue to build these core capabilities until it is demonstrably best. Thus the basic nature of strategic analysis changes from an industry analysis perspective to a horizontal analysis of capabilities across all potential providers of an activity, regardless of which industry the provider might be in (see Exhibit 1).

the capacity to bring more power to bear on a selected sector than any competitor can

6. Elements important to customers in the long run. At least one of the rms core competencies should normally relate directly to understanding and serving its customers that is, the right half of the value chain in Exhibit 1. Hi-tech companies with the worlds best state-of-the-art technology oten fail when they ignore this caveat. On the other hand, Merck matches its superb basic research with a prescription drug marketing knowhow that is equally outstanding. By aggressively analyzing its customers value chains, a company can oten identify where it can specialize and provide an activity at lower cost or more

54

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

efectively to the customer. Such analyses have created whole new industries, like the specialized mortgage broker, syndication, secondary market, transaction-processing, escrow, title search, and insurance businesses that have now taken over these risks and functions for banks and have disaggregated the entire mortgage industry. 7. Embedded in the organizations systems. Maintainable competencies cannot depend on one or two talented stars such as Steven Jobs and Stephen Wozniak at Apple or Herbert Boyer and Arthur D. Riggs at Genentech whose departure could destroy a companys success. Instead, the rm must convert these competencies into a corporate reputation or culture that outlives the stars. Especially when a strategy is heavily dependent on creativity, personal dedication, and initiative or on attracting top-ight professionals, core competencies must be captured within the companys systems broadly dened to include its values, organization structures, and management systems. Such competencies might include recruiting (McKinsey, Goldman Sachs), training (McDonalds, Disney), marketing (Procter & Gamble, Hallmark), innovation (Sony, 3M), motivation systems (ServiceMaster), or control of remote and diverse operating sites within a common framework and philosophy (Exxon, CRA, Inc.). These systems are oten at the heart of consistent superior performance; in many cases, a rms systems become its core competencies.12

Preeminence: The key strategic barrier

For its selected core competencies, the company must ensure that it maintains absolute preeminence. It may also need to surround these core competencies with defensive positions, both upstream and downstream. In some cases, it may have to perform some activities where it is not bestFor its selected core in-world, just to keep existing or competencies, the company potential competitors from learning, must ensure that it maintains taking over, eroding, or bypassing absolute preeminence elements of its special competencies. In fact, managers should consciously develop these core competencies to block competitors strategically and avoid outsourcing them or giving suppliers access to the critical knowledge bases or skills that underpin them. Honda, for example, does all its engine R&D in-house and makes all the critical parts for its small motor design core competency in closely controlled facilities in Japan. It will consider outsourcing any other noncritical elements in its products, but builds a careful strategic block around this most essential element for all its businesses.13

THE McKINSEY QUARTERLY 1995 NUMBER 1

55

STRATEGIC OUTSOURCING

Most important, as a companys preeminence in selected elds grows, its knowledge-based core competencies become ever harder to overtake. Knowledge bases tend to grow exponentially in value with investment and experience. Intellectual leadership tends to attract the most talented people, who then work on and solve the most interesting problems. The combination in Intellectual leadership tends to turn creates higher returns and attracts the attract the most talented people, next round of outstanding talent. In addition who then work on and solve to the examples we have already cited, the most interesting problems organizations as diverse as Bechtel, AT&T Bell Labs, Microsot, Boeing, Intel, Merck, Genentech, McKinsey, Arthur Andersen, Sony, Nike, Nintendo, Bankers Trust, and Mayo Clinic have found this to be true. Some executives regard core activities as those the company is continuously engaged in, while peripheral activities are those that are intermittent and therefore can be outsourced. From a strategic outsourcing viewpoint, however, core competencies are the activities that ofer long-term competitive advantage and thus must be rigidly controlled and protected. Peripheral activities are those not critical to the companys competitive edge.

Strategic outsourcing
If supplier markets were totally reliable and ecient, rational companies would outsource everything except those special activities in which they could achieve a unique competitive edge, that is, their core competencies. Unfortunately, most supplier markets are imperfect and do entail some risks for both buyer and seller with respect to price, quality, time, or other key dimensions. Moreover, outsourcing entails unique Exhibit 2 transaction costs searching, contracting, controlCompetitive advantage versus strategic vulnerability ling, and recontracting that at times may exceed the transaction costs of having the activity directly Strategic under managements in-house control. control
Potential for competitive edge (produce internally)

To address these diculties, managers must answer three key questions about any activity considered for outsourcing. First, what is the potential for obtaining competitive advantage in this activity, Low control taking account of transaction costs? Second, what needed (buy off the shelf) is the potential vulnerability that could arise Degree of strategic vulnerability from market failure if the activity is outsourced? Conceptually, these two factors can be arrayed in a simple matrix (see Exhibit 2). Third, what can we do to alleviate our vulnerability by structuring arrangements with suppliers to aford appropriate controls yet provide for necessary exibilities in demand?
Moderate control needed (special venture or contract arrangements)

56

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

The two extremes in Exhibit 2 are relatively straightforward. When the potential for both competitive edge and strategic vulnerability is high, the company needs a high degree of control, usually entailing production internally or through joint ownership arrangements or tight long-term contracts (explicit or implicit). Marks & Spencer, for example, is famous for its network of tied suppliers, which create the unique brands and styles that underpin the retailers value reputation. Spot suppliers would be too unreliable and unlikely to meet the demanding standards that are Marks & Spencers unique consumer franchise. Hence, close control of product quality, design, technology, and equipment through contracts and even nancial support is essential. The opposite case is perhaps oce cleaning, where little competitive edge is usually possible and there is an active and deep market of supplier rms. In between, there is a continuous range of activities requiring diferent degrees of control and strategic exibility. At each intervening point, the question is not just whether to make or buy, but how to implement a desired balance between independence and incentives for the supplier versus control and security for the buyer. Most companies will benet by extending outsourcing rst in less critical areas, or in parts of activities, like payroll, rather than all of accounting. As they gain experience, they may increase prot opportunities greatly by outsourcing more The question is how to critical activities to noncompeting rms that implement a balance between can perform them more efectively. In a few cases, more complex alliances with competitors may be essential to garner specialized skills that cannot be obtained in other ways. At each level, the company must isolate and rigorously control strategically critical relationships between its suppliers and its customers.

independence and incentives for the supplier versus control and security for the buyer

Competitive edge
The key strategic issue in insourcing versus outsourcing is whether a company can achieve a maintainable competitive edge by performing an activity internally usually cheaper, better, in a more timely fashion, or with some unique capability on a continuing basis. If one or more of these dimensions is critical to the customer and if the company can perform that function uniquely well, the activity should be kept in-house. Many companies unfortunately assume that because they have historically performed an activity internally, or because it seems integral to their business, the activity should be insourced. However, on closer investigation

THE McKINSEY QUARTERLY 1995 NUMBER 1

57

STRATEGIC OUTSOURCING

and with careful benchmarking, a companys internal capabilities may turn out to be signicantly below those of best-in-world suppliers. Ford Motor Company, for example, found that many of its internal suppliers quality practices and costs were nowhere near those of external suppliers when it began its famous best in class worldwide benchmarking studies on 400 subassemblies for the new Taurus-Sable line. A New York bank with extensive worldwide operations investigated why its Federal Express costs were soaring and found that its internal mail department took two days more than Federal Express to get a letter or package from the third oor to the We thought we were the best fortieth oor of its building.

in the world, but we found we were not even up to the worst of the benchmarking cases

In interviews about benchmarking with top operating managers in both service and manufacturing companies, we frequently encountered some paraphrase of We thought we were the best in the world at many activities. But when we benchmarked against the best external suppliers, we found we were not even up to the worst of the benchmarking cases.

Transaction costs
In all calculations, analysts must include internal transaction costs as well as those associated with external sourcing. If the company is to produce the item or service internally on a long-term basis, it must back up its decision with continuing R&D, personnel development, and infrastructure investments that at least match those of the best external supplier; otherwise, it will lose its competitive edge over time. Managers oten tend to overlook such backup costs, as well as the losses from laggard innovation and unresponsiveness of internal groups that know they have a One of the great gains of guaranteed market. Finally, there are the headquarters and support costs of constantly managing the insourced activity. One of the great gains of outsourcing is the decrease in executive time spent managing peripheral activities freeing top management to focus more on the core of its business. Various studies have shown that when these internal transaction costs are thoroughly analyzed, they can be extremely high.14 Since it is easier to identify the explicit transaction costs of dealing with external suppliers, these generally tend to be included in analyses. Harder-to-identify internal transaction costs, however, are oten not included, thus biasing results.

outsourcing is freeing top management to focus more on the core of its business

58

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

Vulnerability
When there are many suppliers (with adequate but not dominating scale) and mature market standards and terms, a potential buyer is unlikely to be more ecient than the best available supplier. If, on the other hand, there is not sucient depth in the market, overly powerful suppliers can hold the company ransom. Conversely, if the number of suppliers is limited or individual suppliers are too weak, they may be unable to supply innovative products or services as well as a much larger buyer could by performing the activity in-house. While the activity or product might not be one of its core competencies, the company might nevertheless benet by producing internally rather than undertaking the training, investment, and codesign expenses necessary to bring weak suppliers up to needed performance levels. Another form of vulnerability is the lack of information available in the marketplace or from individual suppliers; for example, a supplier may secretly expect labor disruptions or raw material problems, but hide these concerns until it is too late for the customer to go elsewhere. A related problem occurs Sometimes the whole when a supplier has unique information structure of information in capabilities: for example, large wholesalers an industry will militate for or retailers, market research rms, sotware or against outsourcing companies, or legal specialists may have information or fact-gathering systems that would be impossible for the buyer or any other single supplier to reproduce eciently. Such suppliers may be able to charge what are essentially monopoly prices, but purchasing from them could still be less costly than reproducing the service internally. In other cases, there may be many capable suppliers (for example, in R&D or sotware), but the costs of adequately monitoring progress on the suppliers premises might make outsourcing prohibitive. Sometimes the whole structure of information in an industry will militate for or against outsourcing. Computing, for example, was largely kept inhouse in its early years because the information available to a buyer of computing services and its ability to make judgments about such services were very diferent for the buying company (which knew very little) than for the supplier (which had excellent information). Many buyers lacked the competency either to assess or to monitor sellers, and feared loss of vital information. A company can outsource computing more easily today, in part because buyers computer, technical management, and sotware knowhow are sucient to make informed judgments about external suppliers. In addition to information anomalies, Stuckey and White note three types of asset specicity that commonly create market imperfections, calling

THE McKINSEY QUARTERLY 1995 NUMBER 1

59

STRATEGIC OUTSOURCING

for controlled sourcing solutions rather than relying on ecient markets.15 These are: (1) site specicity, where sellers have located costly xed assets in close proximity to the buyer, thus minimizing transport and inventory costs for a single supplier; (2) technical specicity, where one or both parties must invest in equipment that can be used only by the parties in conjunction with each other and has low value in alternative uses; and (3) human capital specicity, where employees must develop in-depth skills that are specic to a particular buyer or customer relationship. Stuckey and White explain the outsourcing implications of information and specicity problems in the case of a bauxite mine and an alumina rener. Reneries are usually located close to mines because of the high cost of transporting bauxite, relative to its value. Reneries in turn are tuned to process the narrow set of physical properties associated with the particular mines bauxite. Diferent and highly specialized skills and assets are needed for rening versus mining. Access to information further compounds problems; if an independent mine expects a strike, it is unlikely to share that information with its customers, unless there are strong incentives. As a result, the aluminum industry has moved toward vertical integration or strong bilateral joint ventures, as opposed to open outsourcing of bauxite supplies despite the apparent presence of a commodity product and many suppliers and sellers. In this case, issues of both competitive advantage and potential market failure dictate a higher degree of sourcing control.

Degree of sourcing control

In deciding on a sourcing strategy for a particular segment of their business, managers have a wide range of control options (see Exhibits 3 and 4 for the most basic). Where there is high potential both for vulnerability and for competitive edge, tight control is indicated (as in the bauxite case). At the opposite end is, say, oce cleaning. Between these extremes are opportunities for developing special incentives or more complex oversight contracts to balance intermediate levels of vulnerability against more moderate prospects for competitive edge. Nikes multi-tier strategy ofers an interesting example (see boxed insert on page 62).
Exhibit 3

Range of outsourcing options

Make versus Buy

Self

Partnership

Long term

Short term

Controlled

Not controlled

The practice and law of strategic alliances are rapidly developing new ways to deal with common control issues by establishing specied procedures that permit direct involvement in limited stages of a partners activities,

60

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

Exhibit 4

Potential contract relationships

Full ownership Partial ownership Control need Joint development Retainer Long-term contract Call option Short-term contract Flexibility need

without incurring either the costs of ownership arrangements or the loss of control inherent in arms-length transactions. As they work their way through the available options, managers should ask themselves a series of strategic questions: 1. Do we really want to produce the goods or service internally in the long run? If we do, are we willing to make the back-up investments necessary to sustain a best-in-world position? Is it critical to defending our core competency? If not

2. Can we license technology or buy knowhow that will let us be best on a continuing basis? If not 3. Can we buy the item as an of-the-shelf product or service from a bestin-world supplier? Is this a viable long-term option as volume and complexity grow? If not 4. Can we establish a joint development project with a knowledgeable supplier that will ultimately give us the capability to be best at this activity? If not 5. Can we enter into a long-term development or purchase agreement that gives us a secure source of supply and a proprietary interest in knowledge or other property of vital interest to us and the supplier? If not 6. Can we acquire and manage a best-in-world supplier to advantage? If not, can we set up a joint venture or partnership that avoids the shortcomings we see in each of the above? If so 7. Can we establish controls and incentives that reduce total transaction costs below those of producing internally?

Flexibility versus control

Within this framework, there is a constant tradeof between exibility and control. One of the main purposes of outsourcing is to have the supplier assume certain classes of investment and risk, such as demand variability. To optimize costs, the buying company may want to maintain its internal capacity at relatively constant levels despite highly uctuating sales demands. Under these circumstances, it needs a surge strategy.

THE McKINSEY QUARTERLY 1995 NUMBER 1

61

NIKES MULTI-TIER PARTNER STRATEGY

Nike, Inc. has a three-tier strategy of production partners. Developed partners produce Nikes latest and most expensive statement products, which can absorb higher production costs. These companies usually produce lower volumes, codevelop products, and coinvest in new technologies. Volume producers are above-average in size (making 70,000 to 85,000 units a day, compared to 20,000 to 25,000 units for developed partners). They generally produce a specic type of footwear (say, basketball shoes) and are vertically integrated. Nike does no development work with them because each company may produce for seven or eight other buyers to keep up its volume. Although Nike tries very hard to stabilize volumes for its developed partners, volume partners are expected to handle most surges in volume themselves. Developing sources are attractive primarily because of their low labor costs and their capacity to diversify assembly locations. All produce exclusively for Nike, which has a strong tutelage program to develop them into higher-level suppliers. To help both parties, Nike tries to link developing sources (through joint ventures) with its developed partners. The latter assist by providing training, helping to nance some operations, and moving some of their own labor-intensive activities to these units. In addition to its rst tier of assembly units, Nike supports a complex network of secondtier suppliers for materials, components, and subassemblies. In Nikes third tier, some of the more specialized and technical components for example, patented features like the Nike Airsole are made in supplier companies that Nike totally controls. Nike manages its external suppliers through its expatriate program. Nike expatriates become permanent personnel in each factory producing Nike footwear, functioning as liaisons with corporate R&D, headquarters, and worldwide quality assurance and product development efforts. Nikes own R&D group participates in design, its corporately owned component factories give it a second window on production, and its corporate offices constantly scan new suppliers for better techniques.

McDonalds, for example, with $8 billion in sales and 10.1 percent growth per year, needs to call in part-time and casual workers to handle extensive daily variations yet also be able to select its future permanent or managerial personnel from these people. IBM has had the opposite problem; since its core demand has been declining, the company has had to lay of employees. Yet it needs surge capacity for: (1) quick access to some former employees basic skills; (2) available production capacity without the costs of supporting facilities full time; and (3) the ability to exploit strong outside parties specialized capabilities through temporary consortia for example, in applications sotware, microprocessors, network development, or factory automation. Strategically, McDonalds has created a pool of people available on call options, while IBM through spinouts of factories with baseload commitments to IBM, guaranteed consulting employment for key people, exible joint ventures, and strategic alliances has created put options to handle surge needs as it downsizes and tries to turn around its business.

62

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

There is a full spectrum of outsourcing arrangements, depending on the companys control and exibility needs (see Exhibit 4). The issue is less whether to make or buy an activity than it is how to structure internal versus external sourcing on an optimal basis. Companies are outsourcing much more of what used to be considered either integral elements of their value chains or necessary staf activities. Because of greater complexity, higher specialization, and new technological capabilities, outside suppliers can now perform many such activities at lower cost and with higher value-added than Outside suppliers can now a fully integrated company can. In some cases, new production technologies have moved manufacturing economies of scale toward the supplier. In others, service technologies have lowered transaction costs substantially, making it possible to specify, transport, store, and coordinate inputs from external sources so inexpensively that the balance of benets has shited from insourcing to outsourcing. In certain specialized niches, outside companies have grown to such size and sophistication that they have developed economies of scale, scope, and knowledge intensity so formidable that neither smaller nor more integrated producers can efectively compete with them (for example, ADP Services in payroll, and ServiceMaster in maintenance). To the extent that knowledge of a specic activity is more important than knowledge of the end product itself, specialized suppliers can oten produce higher value-added at lower cost for that activity than almost any integrated company.

perform many activities at lower cost and with higher value-added than a fully integrated company can

Strategic benets versus risks

Too oten companies look at outsourcing as a means to lower only shortterm direct costs. However, through strategic outsourcing, companies can lower their long-term capital investments and leverage their key competencies signicantly, as Apple and Nike have done. They can also force many types of risk and unwanted management problems onto suppliers. Gallo, the largest producer and distributor of wines in the United States, outsources most of its grapes, pushing the risks of weather, land prices, and labor problems onto its suppliers. Argyle Diamonds, one of the worlds largest diamond producers, outsources virtually all aspects of its operation except the crucial steps of separation and sorting of diamonds. It contracts all its huge earth-moving operations (to avoid capital and labor risks), its housing and food services for workers (to avoid confrontations on nonoperating issues), and much of its distribution (to De Beers to protect prices, to nance inventories, and to avoid the

THE McKINSEY QUARTERLY 1995 NUMBER 1

63

STRATEGIC OUTSOURCING

complications of worldwide distribution). By outsourcing to best-inclass suppliers in each case, it further ensures the quality and image of its operations.

Important strategic benets

Strategically, outsourcing can provide the buyer with greater exibility, especially in the purchase of rapidly developing new technologies, fashion goods, or the myriad components of complex systems. It reduces the companys design-cycle times as multiple best-in-class suppliers work simultaneously on individual components of a system. Each supplier can both have more personnel depth and sophisticated technical knowledge about its specic area and support more specialized facilities for higher quality than the coordinating (buyer) company could possibly achieve alone. In the same vein, strategic outsourcing spreads the companys risk for component and technology developments across a number of suppliers. The company does not have to undertake the full failure risks of all component R&D programs or invest in and constantly update production capabilities for each component system. Further, the buyer is not limited to its own innovative capabilities; it can tap into a stream of new product and process ideas and quality improvement potentials it could not possibly Outsourcing has become a generate itself. In the worlds advanced economies, increased auence has forced much greater attention on new product ideas, quality details, and customization. Because small specialized suppliers oten ofer greater responsiveness, and new technologies have reduced the size needed to achieve economies of scale, the average size of industrial rms has decreased since the late 1960s, and subcontracting constitutes an ever greater portion of most producers costs.16 Outsourcing has become a major strategy to leverage internal technical capabilities and to tap the rapid response and innovative capabilities of small enterprises. Richard Leibhaber, chief strategy and technology ocer at MCI, commented: MCI constantly seeks to grow by nding and developing associations with small companies having interesting services they can hang onto the MCI network. Although we employ only about 1,000 professional technical personnel internally, 19,000 such personnel work directly for us through contracts Now we do about 60 percent of our sotware development work internally, but we manage [in detail] the other 40 percent in contractors hands. We do all the specication, process rating, operational procedures, and system testing inside our company. We design the system. We control the process. Then we let others do what they can do best.17

major strategy to tap the rapid response and innovative capabilities of small enterprises

64

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

Boston Consulting Group, which has studied more than a hundred major companies doing extensive outsourcing, has concluded that most Western companies outsource primarily to save on overhead or short-term costs.18 The result is a piecemeal approach that results in patches of overcapacity scattered at random throughout the companys operation [These Western companies outsource companies] end up with large primarily to save on overhead numbers of subcontractors, which or short-term costs are more costly to manage than inhouse operations that are individually less ecient.19 Worse still, the buying companies, by not providing adequate monitoring and technical backup, oten lose their grip on key competencies they may need in the future. The Japanese, by contrast, outsource primarily to improve the eciency and quality of their own processes, focus on a very few sources, build close interdependent relationships, and hold on tightly to high value-added activities that are crucial to quality. For those systems they contract out, Japanese companies advise closely on manufacturing and cooperate in process and product R&D on the suppliers premises.

Strategic risks
Outsourcing complete or partial activities creates great opportunities but also new types of risks. Managements main strategic concerns are: 1. Loss of critical skills or developing the wrong skills. Unfortunately, many US companies outsourced manufacture of what at the time seemed to be only minor components, like semiconductor chips or bicycle frames, and taught suppliers how to build them to needed quality standards. Later, these The Japanese outsource companies found their suppliers were unable primarily to improve the or unwilling to supply them as required. By eciency and quality of then, they had lost the skills they needed to their own processes reenter manufacture and could not prevent their suppliers from either assisting competitors or entering downstream markets on their own. In some cases, by outsourcing a key component, companies lost their own strategic exibility to introduce new designs when they wanted, rather than when vendors permitted a change. Few manufacturers, for example, can aford to design a new laser-printer engine to obtain a slight timing edge, rather than wait for Canon (with its 84 percent market share in such engines) to move.20 The activity share that Canon has in the design function for this eld, or that Matsushita now has in drives for CD players, gives each company such an overwhelming

THE McKINSEY QUARTERLY 1995 NUMBER 1

65

STRATEGIC OUTSOURCING

competitive advantage that it can control other characteristics of its industry, limiting the strategic options of others. The dual strategy framework we propose prevents such bypassing and helps managers explicitly address both needed long-term competencies and strategic vulnerabilities before embarking on outsourcing. 2. Loss of cross-functional skills. The interactions among skilled people in diferent functional activities oten develop unexpected new insights or solutions. Companies fear outsourcing will make such cross-functional serendipity less likely. However, if the company consciously ensures that its remaining employees interact constantly and closely with its outsourced experts, its employees knowledge base can be much higher than if production were in-house, and the creativity benets can be even greater. Companies using Texas Instruments or Intels design capabilities, for example, can have these suppliers designers with their much greater technical expertise and access to support technologies codesign chips in-house with their own design teams during development. In such circumstances, the buyers employees are in contact with much more skilled integrated Outsourcing can oten increase circuit people than the rm could possibly total innovation potential possess itself. In a number of industries studied, twothirds of all innovation occurred at the customer/supplier interface.21 By denition, outsourcing increases the buyers own participation (as the customer) in such interfaces. Consequently, if managed properly, the practice can oten increase total innovation potential substantially through leveraging multiple supplier relationships.

substantially through leveraging multiple supplier relationships

However, having outsourced expertise at many diferent locations may make close cross-functional teamwork more dicult. To guard against this, in entering long-term outsourcing relationships that may involve future innovation, many managers specify that an outsource partners personnel may be seconded to the buying company for special development projects. They then ensure that close personal relationships are developed between the supplier companies and their own technical personnel at the bench and operating levels. Contractual prearrangements are usually necessary to ensure that critical personnel from the outsource partner are available when needed. But the buying company must be close enough to its partner to evaluate and name the specic people it wants; otherwise the provision may be useless when it is needed.

66

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

3. Loss of control over a supplier. Real problems can occur when the suppliers priorities do not match the buyers. The most successful outsourcers nd it absolutely essential to have both close personal contact and rapport at the oor level and political clout and understanding with the suppliers top management. For this reason, Nike both has full-time production expatriates on its suppliers premises and frequently brings the suppliers top people to its Beaverton, Oregon headquarters to exchange details about future capabilities and prospects. When conicts occur, both the suppliers CEO and key operating personnel can be pressured directly to break Some suppliers, ater building the logjam. Even then, serious diculties can occur if the buyer does not have sucient market power relative to the seller. Some buying companies go to the extreme of owning key pieces of equipment the seller uses to make the components they are purchasing. If priorities conict too badly, the buyer can remove its equipment and shut down the sellers whole line. These buyers say that such arrangements ensure we can get the sellers attention when we need it.

up their expertise with the buyers support, will attempt to bypass the buyer directly in the marketplace

Nevertheless, unless the buyers core competency is a true block to the marketplace, some suppliers, ater building up their expertise with the buyers support, will attempt to bypass the buyer directly in the marketplace, as Giant Manufacturing of Taiwan, a supplier of bicycle frames, did to Schwinn. Alternatively, the seller may learn as much as possible from the buyer and its engineering groups and then attempt to resell this knowledge in diferent product congurations to the buyers competitors, as Toshiba did with submarine propeller technology. Careful denition, limitation, and implementation of means to remedy such external conicts are critical in any but the most routine outsourcing arrangements. Companies that outsource extensively have generally found satisfactory legal and operational ways to deal with the problem and are oten willing to share useful techniques with those outside their industry.

Careful implementation of means to remedy external conicts are critical in any but the most routine outsourcing arrangements

New management approaches

Most large companies have very sophisticated techniques for the traditional purchasing of parts, subassemblies, supplies, equipment, construction, or standard services. And models from the natural resources, real estate/ construction, and nance/insurance industries where joint ventures have

THE McKINSEY QUARTERLY 1995 NUMBER 1

67

STRATEGIC OUTSOURCING

been common for years can provide useful guides for more complex partnering relationships. In addition to seeking out these experiences, the main managerial adjustments for most companies are those needed for coping with the increased scale, diversity, and service-oriented nature of the activities potentially outsourced. These center on: (1) a much more professional and highly trained purchasing and contract management group (as compared with the lowly purchasing groups of the past); and (2) a greatly enhanced logistics information system (to track and evaluate vendors, coordinate transportation activities, and manage service transactions and materials movements from the vendors hands to the customers). There is vast electronic document interchange (EDI) and materials requirements planning (MRP) literature on such logistics manageTo manage more extended ment for products and components. Now, similar concepts are needed for the management of knowledge and servicebased activities. A number of companies are establishing direct computer connections between their service suppliers and the top managers controlling that function, as Apple has done with its consulting and public relations groups. It is increasingly easy to implement sotware interfaces that allow continuous electronic monitoring and executive interactions for the design, nancial, advertising, public relations, R&D, real estate, or personnel-search activities, as well as manufacturing, being performed at remote locations.

outsourcing efectively, contracting and logistics activities need to be elevated to corporate strategic levels

To manage more extended outsourcing efectively, contracting and logistics activities generally need to be elevated to corporate strategic levels. So should the development of the much more sophisticated knowledge-based systems needed to capture and analyze essential details about the companys own internal processes and those of vendors. Most companies systems need improving in this respect, but some interesting developments are under way. A consortium headed by Digital Equipment, Ford, Texas Instruments, US West, Carnegie Group, and Alcorp is developing a next-generation sotware tool called Initiative for Managing Knowledge Assets (IMKA). Built around the proven knowledge-based systems of Digital and TI, IMKA uses many features found in object-oriented programs that allow collection and comparison of knowledge from a variety of diferent end nodes, inside and outside the company. For example, it can compare each supplier plant throughout the world for its capability to meet a given subsystems circuit performance, manufacturing, design, interconnection, weight, cost, or power requirement parameters for any of 100,000 diferent objects against

68

THE McKINSEY QUARTERLY 1995 NUMBER 1

STRATEGIC OUTSOURCING

the capabilities of other possible supplier plants. In milliseconds, it can coordinate and check the production dynamics necessary to ensure that parts are produced in optimum locations at lowest cost and to specication throughout the world. Many companies fear they may not be able to maintain sucient knowledge internally to manage their specialist suppliers, a real problem if not attacked systematically. Most successful outsourcers upgrade both their top management talent When they move aggressively, and their information systems for this purmany companies have found pose. When they move aggressively, many that they actually improve their companies have found that they actually knowledge bases through improve their knowledge bases through strategic outsourcing strategic outsourcing, as Ford did with its best-in-class program. By actively riding circuit on the best outside suppliers and experts, they obtain more stimulation and insights than any insider group could possibly ofer, unless that group represented the core competency of the company. Further, when executives continuously interact with the very best talent in the world, not just the best in the next oce, they are considerably more likely to stay at the top of their professions. As a side benet of outsourcing, they can pressure internal supply groups to compete with the best external companies, question subordinates more knowledgeably, and keep internal groups more competitive.

Most companies can substantially leverage their resources through strategic outsourcing by: (1) developing a few well-selected core competencies of signicance to customers and in which the company can be best-in-world; (2) focusing investment and management attention on them; and (3) strategically outsourcing many other activities where it cannot or need not be best. There are always some inherent risks in outsourcing, but there are also risks and costs associated with insourcing. When approached within a genuinely strategic framework, using the variety of outsourcing options available, and analyzing the strategic issues developed here, companies can overcome many of the costs and risks. When intelligently combined, core competency and extensive outsourcing strategies provide improved returns on capital, lowered risk, greater exibility, and better responsiveness to customer needs at lower cost.

THE McKINSEY QUARTERLY 1995 NUMBER 1

69

STRATEGIC OUTSOURCING

NOTES
The authors gratefully acknowledge the research support of McKinsey & Company; NovaCare, Inc.; Marsh & McLennan; William M. Mercer Companies; Arthur Andersen & Co.; and American Express on various aspects of this project. 1 2 3 J. B. Quinn, T. L. Doorley, and P. C. Paquette, Technology in services: Rethinking strategic focus, Sloan Management Review, Winter 1990, pp. 7987. J. B. Quinn, Leveraging knowledge and service-based strategies through outsourcing, in Intelligent Enterprise, Free Press, New York, 1992, pp. 7197. M. Moritz, The Little Kingdom: The Private Story of Apple Computer, Morrow, New York, 1984; and W. Davidson, Apple Computer Inc., University of Virginia, Charlottesville, Darden Research Foundation, Case UVA-BP219, 1984. R. Coase, The nature of the rm, Economica, November 1937, pp. 386405; and O. Williamson, Markets and Hierarchies, Analysis and Antitrust Implications, Free Press, New York, 1975. R. Rumelt, Strategy, Structure and Economic Performance, Harvard University Press, Cambridge, Massachusetts, 1974. R. DAveni and A. Illinich, Complex patterns of vertical integration in the forest products industry, Academy of Management Journal, Vol. 35, 1992, pp. 596625; P. Y. Batteyri, The concept of impartition policies: A diferent approach to vertical integration strategies, Strategic Management Journal, Vol. 9, 1988, pp. 50720. G. J. Maloney, The choice of organizational form, Strategic Management Journal, Vol. 3, 1992, pp. 55984; R. Miles and C. Snow, Organizations, new concepts and new forms, California Management Review, Spring 1986, pp. 6273. Rumelt, Strategy, Structure and Economic Performance. W. Davidson, The Amazing Race, Winning the Technorivalry with Japan, John Wiley, New York, 1983. C. Prahalad and G. Hamel, The core competence of the corporation, Harvard Business Review, MayJune 1990, pp. 7991. J. B. Quinn, T. L. Doorley, and P. C. Paquette, Beyond products: Service-based strategies, Harvard Business Review, MarchApril 1990, pp. 5868. D. Turner and M. Crawford, Managing current and future competitive performance: The role of competence, University of New South Wales, Australian Graduate School of Management, Center for Corporate Change, Kensington, 1992. H. Mintzberg and J. B. Quinn, Honda Motor Co., The Strategy Process, Prentice Hall, Englewood Clifs, New Jersey, 1993, pp. 14055. R. DAveni and D. Ravenscrat, Economics of integration versus bureaucracy costs: Does vertical integration improve performance? Academy of Management Journal, Vol. 37, Number 5, 1994, pp. 11671206; and H. Mintzberg, The Nature of Managerial Work, Harper & Row, New York, 1973. J. Stuckey and D. White, When and when not to vertically integrate, Sloan Management Review, Spring 1993, pp. 7183. The incredible shrinking company, The Economist, 15 December 1990, pp. 656; and Costing the factory of the future, The Economist, 3 March 1990, pp. 612. Interview with J. B. Quinn, March 1992. Manufacturing: The ins and outs of outing, The Economist, 31 August 1991, pp. 54, 56. M. F. Blaxill and T. M. Hout, The fallacy of the overhead quick x, Harvard Business Review, JulyAugust 1991, pp. 93101. R. Reich, Who is us?, Harvard Business Review, JanuaryFebruary 1990, pp. 5364. E. von Hippel, The Sources of Innovation, Oxford University Press, New York, 1988.

5 6

8 9 10 11 12

13 14

15 16 17 18 19 20 21

70

THE McKINSEY QUARTERLY 1995 NUMBER 1