What to Read Next
The French food giant Groupe Danone, long a leader in the Chinese market for beverages and food products, has recently seen its position in this enormous market deteriorate drastically. The reason: Danone’s strategic partnership with Hangzhou Wahaha Group Co. Ltd. is breaking up. Wahaha became the dominant player in the Chinese bottled water and other nonalcoholic beverage market through its 1996 alliance with Danone. But by 2007, Wahaha was blaming Danone for setting up competing joint ventures with other local companies, such as Robust, Aquarius, Mengniu Dairy and Bright Dairy & Food, while Danone was suing Wahaha for using the brand outside the scope of their joint ventures. Wahaha retaliated by dragging several Danone officials to court for conflict of interest because of their simultaneous membership on the boards of the Wahaha-Danone joint venture and other competing joint ventures Danone had in China. As a result, the relationship further deteriorated, and over 30 lawsuits were eventually filed on three different continents. By the end of 2009, a settlement was reached in which Danone pulled out of the alliance, which had accounted for a dominant share of the French company’s sales in China of almost US$3 billion — about 10% of its total worldwide sales.
Danone’s bungled approach to the formation of corporate alliances probably resulted in the destruction of several billion dollars’ worth of market capitalization. Our study of how companies make decisions on the formation of alliances shows that this sort of dysfunctional behavior is all too common. Most companies now maintain an alliance portfolio comprising multiple simultaneous alliances with different partners.1 In the global air transportation industry, for example, most airlines maintain broad portfolios of code-sharing alliances with other carriers, which allow them to significantly extend their route networks by offering services to their partners’ destinations. In 1994, the average number of alliances per airline company was only four. By 2008, however, the picture had changed dramatically: The average alliance portfolio size across the industry had increased to 12, with some airlines engaging simultaneously in as many as 30 or 40 alliances.2 Despite this proliferation of corporate collaborations, research reveals a troublesome pattern. When a company adds a new alliance to its portfolio, it tends to focus on how much value the alliance will create as a stand-alone transaction but ignore the fact that the composition of its entire alliance portfolio is an important determinant of the value that will come from a new alliance. In other words, an alliance opportunity that promises to create value from a stand-alone perspective may not necessarily be value-creating from an alliance portfolio perspective. The formation of the new alliance may even be an overall value-destroying move.
The Leading Question
How can companies form strategic alliances that create value on a stand-alone basis and at the alliance portfolio level?
- Use an alliance business case framework that takes into account the costs and benefits on both levels.
- Empower an individual or a department to oversee alliance formation decisions.
- Implement an integrated and codified decision process.
By studying the global air transportation industry, we found concrete evidence of this proposition. Formations of alliances that create synergies with other alliances in an existing alliance portfolio have a more positive effect on companies’ stock prices than alliances with little or no synergy-creating potential. We also found that the stock market penalizes companies that enter into alliances that create conflict in the form of market overlap with existing alliance partners. Structural incentives at many companies often encourage a process that results in actions that may benefit one business unit while hurting the corporate whole.
About the Research
We conducted two types of research. First, we conducted qualitative research and interviewed executives involved in alliance decisions in globally operating companies from industries in which alliance portfolios are an important strategic device and an essential part of business strategy. More specifically, we interviewed executives from the global air transportation sector (American Airlines, Air Canada, Air France, Delta Air Lines, Deutsche Lufthansa and others), packaged goods (Danone), aerospace and defense (EADS, SAFRAN), financial services (Banco Santander, Banco Bilbao Vizcaya Argentaria) and telecommunications (Telefónica de España, Ningbo Bird).
Second, we conducted quantitative research on companies in the global air transportation industry to better understand how alliance portfolios affect the value the companies derive from individual alliance formations. We examined 24 publicly traded, internationally operating airlines from 19 countries and their alliance portfolios as well as 259 formations of code-sharing alliances of these companies over a five-year period. We applied event study methodology and tracked the abnormal stock market returns following the announcements of the code-sharing alliances. To examine how the alliance portfolios of these 24 companies contribute to the explanation of the abnormal stock market return following individual alliance formations, we operationalized various alliance portfolio level measures and applied multivariate statistical techniques to analyze the impact on company value.
Observation of this phenomenon raises three questions. Why do companies behave so thoughtlessly when forming alliances? What are the impediments to a more strategic approach to configuring effective alliance portfolios? And, finally, what’s the way out — that is, what systems and processes should organizations adopt that will enable them to optimize the value of their alliance portfolios?
Understanding the counterproductive incentives that skew alliance formation allows us to propose a new decision-making process that we believe will encourage companies to shift from the currently prevalent ad hoc approach to a smarter approach. In particular, companies should manage their alliances not as stand-alone arrangements but much more strategically, paying far more attention to how their various partnerships interact with one another.
Pitfalls of Alliance Portfolio Expansion
Alliance portfolios often result from a “sedimentary” accumulation process. That is, companies engage in multiple alliances over time, and all these partnerships accumulate haphazardly. Most alliances — even far-reaching partnerships that profoundly affect companies’ overall performance — are initiated on an operational level as ad hoc responses to local business issues. In contrast, broad alliances promoted at the corporate level rarely translate into any actual business development.
As a result, one part of an organization, such as a business unit, will enter into partnerships that serve its own parochial interests, often without realizing or without regard for the impact on other parts of the organization or the company as a whole. It is not surprising that alliance formation decisions are often local matters and that coordination across units in the organization tends to be limited. Because business problems, and thus alliance formation opportunities, often are located in different units within an organization, problem owners are rarely the same individual. The issues associated with creating an alliance tend to affect different managers, who may interact with one another rarely, if ever. From an alliance portfolio perspective, such bottom-up alliance formation decision making is especially problematic when decision makers focus exclusively on criteria relevant to their local business problems and ignore how the new alliance fits into the company’s alliance portfolio. Companies implement patchwork solutions that address problems for parts of a company but may actually create new troubles for other parts. The net result is often a failure to create value for the company overall — or, worse, a destruction of value overall.
Explanations for such silo thinking are manifold. Business-unit managers tend to have clear performance targets that are intrinsically linked to the success of their own units; this naturally leads to the prioritization of local needs over broader corporate needs. Academic research has also shown that managers frequently behave opportunistically and use alliance formations as ways to improve their own freedom of action.3 Indeed, because alliance management is shared with a partner and because interpersonal relationships are often crucial in this process, it is difficult for corporate-level management to interfere. As a result, business-unit managers find themselves with a great deal of autonomy in alliance affairs.
Consider, for example, the decision of Construcciones Aeronauticas SA or CASA, a Spanish aerospace and defense company, to partner with McDonnell Douglas Corp. in the 1980s. This move originated at the plant level: CASA had idle manufacturing capacity that needed to be utilized. By becoming a partner in McDonnell Douglas’s MD-80 commercial airliner project, CASA was able to create activity for one of its factories in Seville, Spain. The alliance with McDonnell Douglas thus solved a local business problem (unused manufacturing capacity). From an alliance portfolio perspective, however, the move was problematic. Since 1971, CASA had been collaborating with Airbus — a McDonnell Douglas rival and the manufacturer of the A319 aircraft, a direct competitor of the MD-80. As a result of its partnership with McDonnell Douglas, CASA’s standing with Airbus suffered and it was not able to increase its share in new Airbus projects as it had hoped. Because the Seville plant was not involved in the Airbus collaboration, its managers lacked the perspective to fully understand and evaluate the consequences that would follow from teaming up with McDonnell Douglas.
Nor is that the only instance of a new alliance formation that appeared to be value-creating when viewed as a stand-alone transaction but turned out to be value-destroying from an alliance portfolio perspective. In the 1990s, the Spanish telecommunications company Telefónica de España maintained a key alliance with Unisource NV, a consortium that included telecommunications providers Koninklijke KPN N.V. of the Netherlands; Stockholm, Sweden-based Telia; Swiss Telecom; and AT&T World Partners. In the late 1990s, one of Telefónica’s business units, Telefónica Internacional or TISA, engaged in an alliance with Concert, a venture between British Telecom and MCI, in order to promote business in Latin America. While TISA managers focused on the value that Concert could add to Telefónica, they ignored the alliance’s value-destroying effect from the portfolio-level perspective. Shortly after the TISA-Concert alliance was formed, Unisource partners raised concerns that Concert was becoming AT&T World Partners’ main competitor and asked Telefónica to leave the Unisource alliance. As part of its exit, Telefónica paid more than 14 billion pesetas to Unisource (about US$94 million).
Making Alliances Fit
How can companies create alliances that not only are valuable at the local business-unit level but also work together to comprise a coherent alliance portfolio — a portfolio that is worth more than the sum of its parts? One part of the answer lies in having a corporate-level department that coordinates all alliance-related activity across a company’s multiple units.4 Such a department — what we call an alliance function — is crucial in ensuring the overall effectiveness of the alliance portfolio. Even though ideas for alliances have to come from low down in the organization, there must be an overall alliance portfolio compatibility check in order to ensure that value is created not only on the local, and thus individual, alliance level but also on the alliance portfolio level.
A dedicated alliance function should keep track of the alliance portfolio as it evolves over time and manage the balance between local problem solving and overall alliance portfolio effectiveness. The reality, however, is that such functions are often performed by one of many internal corporate consulting units. Such units are run as cost centers, and their managers typically have little power over operational decision makers. Conflict between such a central function and the business unit that initiates particular alliances is often unavoidable due to diverging performance measures and interests. Although a dedicated alliance function is measured against overall alliance portfolio effectiveness and how well it enforces the company’s overall alliance policy, individual alliances often tie into business-unit performance metrics for which the alliance initiator is responsible. The buck essentially stops at the business unit, not at the corporate staff level where the portfolio is coordinated.
Based on our research, we put forward a framework that will help managers to systematically make alliance formation decisions by considering interests both on the local and the alliance portfolio levels. The following three elements can help managers address this challenge: (1) an alliance business case framework that takes into account costs and benefits on the individual level as well as the alliance portfolio level of analysis, (2) an integrated and codified decision process involving managers on the business as well as the corporate levels, and (3) clearly defined roles and responsibilities for all actors involved in decision making.
Holistic Cost-Benefit Analysis
By engaging in strategic alliances, companies incur both costs and benefits.5 Rationally behaving companies will enter into an alliance only when it creates value — that is, when its expected benefits exceed its costs.6 The construction of a business case for each alliance formation is therefore essential.7
Typically, alliance business cases tend to take into account only those benefits and costs that occur for a particular alliance, in isolation from existing ones. But such a myopic cost-benefit evaluation can be misleading. An alliance viewed on a stand-alone basis may appear to create value — but when the benefits and costs on the alliance portfolio level of analysis are taken into account, the picture can be quite different. To construct a robust business case, therefore, companies need to identify and quantify a broad range of benefits and costs both at the level of the individual alliance and at the portfolio level.
Costs and Benefits: Individual Alliance Viewed on a stand-alone basis, alliances can help companies do four things:
- Achieve economies of scale by pooling similar assets, knowledge or skills. For example, Volkswagen AG and Renault S.A. teamed up to jointly develop and produce automatic gearboxes because the market for automatic cars in Europe is small and neither company had large enough volume to achieve efficient production volumes.
- Obtain access to a partner’s complementary assets, knowledge and skills. For example, Volkswagen teamed up with Shanghai Automotive Industry Corp. to create Shanghai VW, which leveraged VW’s products and technology into China, where SAIC had local knowledge and access to manufacturing and distribution assets.
- Obtain access to new skills. Similarly, General Motors Co. and Toyota Motor Corp. formed the NUMMI alliance so that GM could learn about Toyota’s lean manufacturing process and Toyota could take a cue from GM on how to deal with U.S. labor unions.
- Reduce competition in the market and increase market power. Rather than competing individually, DaimlerChrysler Aerospace AG (DASA), British Aerospace, CASA and Aerospatiale Matra formed the Airbus alliance, offering a jointly developed and manufactured product. The companies thus benefited from a less crowded playing field within the European market for commercial aircraft.
Still working from this stand-alone perspective, we can see that alliances also create costs for their partners. Most apparent are start-up costs, along with ongoing coordination costs. More subtle are the costs that stem from unexpected leakage, where one partner in the alliance acts opportunistically and appropriates skills or knowledge from the other. There are numerous examples of alliances in which one partner appropriates skills and knowledge from the other partner in an opportunistic fashion. Such poaching has afflicted even the seemingly genteel business of flower breeding. Meilland International, headquartered in France and one of the world’s leading developers of rose hybrids, found that one of its former Chinese licensees was growing and marketing several of its patented varieties on a very large scale without paying any royalties.8
But it is not sufficient to focus only on the costs and benefits at the level of individual alliances: The alliance business case needs to be pushed further, to its impact at the level of a company’s entire alliance portfolio.
Costs and Benefits: Alliance Portfolio The benefits that a new alliance can create on the alliance portfolio level mainly stem from ways in which the new alliance and the existing ones can enhance each other. Two types of synergy are:
- Sharing or recombining know-how. Assets, knowledge and skills contributed or developed by a new alliance may be used in another, ongoing alliance. A new alliance may help a company develop a technology that may be of use in an existing alliance. Similarly, assets and knowledge contributed or developed by a new alliance may be combined with those associated with an existing alliance to create a new service or product. For example, California-based on-demand service provider Salesforce.com inc. has a portfolio of alliances that allows the company to combine physical and intellectual assets from various alliance partners to offer a more comprehensive service to its customers. Salesforce.com teamed up with Google Inc. to offer a joint product called Salesforce for Google Apps, a business productivity application that allows “business professionals to communicate, collaborate and work together in real time over the Web.”9 In order to offer its customers more computing power and allow them to host public Web sites and company intranets over its Force.com platform, Salesforce.com also engaged in an alliance with Amazon.com Inc. The online retailer, which maintains a vast network of servers, makes excess server capacity available to Saleforce.com. Thus, Salesforce.com is able to leverage both alliances and offer a broader and more comprehensive service to its customers.
- Reinforcing existing coalitions. The airline industry provides a good example of this type of synergy. As a consequence of merging with Air France, KLM joined the Skyteam alliance that had previously been formed by Air France and Delta Air Lines Inc., while retaining its own long-standing relationship with Northwest Airlines Corp. Northwest eventually decided to follow its long-time partner KLM into Skyteam, rather than partner with other airlines, which strengthened the existing coalition. (Some years after this move, Northwest merged with Delta.)
The costs on the alliance portfolio level that a new alliance can create mainly stem from conflict resolution in existing alliances. These costs are related to re-establishing trust and good will with an existing alliance partner and can arise when a new alliance interferes with the relationship between a company and its existing partners. Such a disturbance may occur when the new alliance represents a competitive threat toward the existing partner — perhaps through imitating the other partner’s technology or offering similar products or services.
When a newly formed alliance overlaps in product or market scope with an existing partner’s business operations, the focal company may not only incur increased conflict resolution costs but in the worst case may also have to bear the consequences associated with dissolving the pre-existing alliance. An existing partner may decide to terminate an alliance altogether as a reaction to the formation of a new alliance and the resulting conflict. Such termination can result in the loss of valuable resources and particular revenue streams and therefore create a cost, as in the unhappy story of Danone and Wahaha.
Other examples can be found in the airline industry. British Airways Plc’s decision to engage in an alliance with American Airlines Inc. directly led to the termination by US Airways Group Inc. of its existing code-share and frequent flyer partnership with British Airways. USAir likely felt threatened by the new links that BA was forging with American. Although BA called USAir’s move “disappointing and puzzling,”10 USAir may have suspected that BA would systematically favor its relationship with American, a larger and more powerful airline. BA might, for example, direct most of its flights toward American’s hub cities or optimize its connections with American at USAir’s expense. Another example is Singapore Airlines Limited, which entered into a multipartner alliance with Delta and Swissair involving equity swaps. Later, however, Singapore Airlines entered into a code-sharing and joint marketing agreement with Deutsche Lufthansa AG, leading to the termination of its alliance with Delta and Swissair.
The second source of tension between new partnerships and existing ones becomes salient when the focal company incurs costs associated with redundancy and value cannibalization as a result of a new alliance. Overlap in scope between a new alliance and an existing one can diminish either’s benefits. It’s true that redundancy is sometimes created deliberately to pursue various simultaneous options and thus spread risks, especially in research and development alliances to develop competing technologies. Nevertheless, companies that pursue such a strategy will incur redundancy costs because they will have to staff more people on the alliances than they would if they only had one alliance.
Toward Smarter Alliance Formation
Sound alliance formation decisions require an integrated decision process that involves actors on the business level as well as the corporate level in order to ensure a systematic assessment of the proposed alliance as well as the alliance portfolio level benefits and costs. The first phase of such a decision process should take an individual alliance perspective and the second phase an alliance portfolio perspective. Companies should assess alliance opportunities according to a rigorous set of criteria. By systematizing what is at present done on a mostly ad hoc basis, companies should be better able to ascertain that the partnership they are contemplating will do more good than harm.
ALLIANCE ASSESSMENT DECISION PROCESS
The other critical step that companies should take to make sense of their alliance formation is to clarify exactly who in the organization plays what role in the process. Alliance formation decision making in companies with alliance portfolios is a complex undertaking that requires the involvement of a broad range of actors on both the business-unit and corporate levels.
To ensure effective alliance formation decision making, the exact roles and responsibilities of each individual player or committee in the process needs to be formalized, codified and communicated throughout the organization. One commonly used method of business process analysis, RACI analysis, tags the various actors in an organization as having one or more critical roles. Based on our research, we have developed a RACI diagram that precisely defines the various actors in the alliance formation process and assigns to each actor one or more RACI tags.
ALLIANCE DECISIONS: WHO’S RESPONSIBLE FOR WHAT?
A clearly defined decision process and explicitly codified roles and responsibilities could have helped companies such as Telefónica to see the bigger picture and avoid falling into the trap of forming alliances that appear to be value-creating deals from the business-unit perspective but create negative intra-alliance portfolio dynamics that destroy value overall. More specifically, it would have forced Telefónica’s TISA unit to flush out the benefits as well as downsides of the new partnership and forced it to involve other parts of the organization in the decision. Through clear roles and responsibilities, the corporate alliance function could have put the potential outcomes in the context of the company’s overall strategy.
There is no doubt that alliance portfolios have become a popular and important strategic device for companies seeking both to exploit opportunities and to neutralize threats. But surprisingly, many decisions pertaining to the formation of alliances are made on an operational level, without much regard for the impact on the company’s overall alliance portfolio. By following a rigorous process that includes an integrated alliance business case and clear roles and responsibilities in decision making, companies can ensure that their alliance portfolios do not grow in a haphazard fashion and that their alliances are worth more as a whole than as the sum of their individual parts.
As noted by corporate management scholar D.A. Levinthal, “in complex decision problems, the discovery of the optimum is an extremely difficult task.”11 Such decisions are even more difficult if they span multiple organizational units within a company or if outside parties are involved. Alliance formation — which by definition entails bringing together multiple business units and inevitably requires interaction with independent partner companies — thus presents numerous possibilities for mismanagement. In fact, viewed in that light, alliance formation emerges as one of the most fraught and perilous moves an organization can make.
The tools described above are essentially a set of mechanisms that help companies address three common issues with alliance formation decisions from an alliance portfolio perspective: breaking down silo thinking, overcoming information asymmetries and balancing local and global needs across the organization. By sitting down around one table, decision makers from different units can better understand each others’ perspectives, realize how a new alliance they are considering might be detrimental elsewhere in the organization and bargain with one another to arrive at a mutually acceptable solution. Also, by putting the responsibility for alliances in a central position, this approach will increase the status, influence and power that function has. As a result, companies will find it easier to arrive at decisions in which overall corporate objectives prevail over local, business-unit interests. Finally, working out the business case for an alliance will force all parties involved in the process to quantify, or at least estimate, the benefits and the downsides of the partnership being considered and allow for a somewhat more objective evaluation of the various implications of forming that particular alliance, both on the level of each individually affected business unit and on the overall corporate level.
1. See, for example, S. Parise and A. Casher, “Alliance Portfolios: Designing and Managing Your Network of Business-Partner Relationships,” Academy of Management Executive 17, no. 4 (2003): 25-39; W.H. Hoffmann, “Strategies for Managing a Portfolio of Alliances,” Strategic Management Journal 28, no. 8 (2007): 827-856; and U. Wassmer, “Alliance Portfolios: A Review and Research Agenda,” Journal of Management 36, no. 1 (2010): 141-171.
2. “Airline Business Alliance Survey,” Airline Business, 1994-2008.
3. J.J. Reuer and R. Ragozzino, “Agency Hazards and Alliance Portfolios,” Strategic Management Journal 27, no. 1 (2006): 27-43.
4. P. Kale, J.H. Dyer and H. Singh, “Alliance Capability, Stock Market Response and Long-Term Alliance Success: The Role of the Alliance Function,” Strategic Management Journal 23, no. 8 (2002): 747-767; and P. Kale, J.H. Dyer and H. Singh, “Value Creation and Success in Strategic Alliances: Alliancing Skills and the Role of Alliance Structure and Systems,” European Management Journal 19, no. 5 (2001): 463-471.
5. P.J. Buckley and M. Casson, “A Theory of Cooperation in International Business,” in “Cooperative Strategies in International Business,” ed. F.J. Contractor and P. Lorange (Lexington, Massachusetts: Lexington Books, 1988): 31-53; A. Madhok and S.B. Tallman, “Resources, Transactions and Rents: Managing Value Through Interfirm Collaborative Relationships,” Organization Science 9, no. 3 (1998): 326-339; S.H. Park and D. Zhou, “Firm Heterogeneity and Competitive Dynamics in Alliance Formation,” Academy of Management Review 30, no. 3 (2005): 531-554; and S. White and S.S. Lui, “Distinguishing Costs of Cooperation and Control in Alliances,” Strategic Management Journal 26, no. 10 (2005): 913-932.
6. Buckley and Casson, “Theory of Cooperation”; J. Koh and N. Venkatraman, “Joint Venture Formations and Stock Market Reactions: An Assessment in the Information Technology Sector,” Academy of Management Journal 34, no. 4 (1991): 869-892; and Madhok and Tallman, “Resources, Transactions and Rents.”
7. J.H. Dyer, P. Kale and H. Singh, “How to Make Strategic Alliances Work,” MIT Sloan Management Review 42, no. 4 (summer 2001): 37-43.
8. P. Dussauge, “Alliances, Joint-Ventures and Chinese Multinationals,” in “Chinese Multinationals,” ed. J.P. Larçon (Hackensack, New Jersey: World Scientific Publishing Company, 2008).
10. K. Schwartz, “USAir Ends Code Share with British Airways,” The Associated Press, Oct. 24, 1996.
11. D.A. Levinthal, “Organizational Adaptation and Environmental Selection: Interrelated Processes of Change,” Organization Science 2, no. 1 (1991): 140-145.