009 Measuring alliance performance Main Image_Original

Measuring alliance performance

By Jim Bamford and David Ernst

Large companies often have dozens of alliances—and little idea how they are performing. Here’s how to evaluate them.

As corporations have evolved from command-and-control structures with sharply defined boundaries into loosely-knit organizations, corporate alliances1 have become central to many business models. Most large companies now have at least 30 alliances, and many have more than 100.

Yet despite the ubiquity of alliances—and the considerable assets and revenues they often involve—very few companies systematically track their performance, creating a substantial risk of negative surprises. We believe that every corporate manager, including CFOs, should be well tuned into the performance of their alliances. In our work with more than 500 companies around the world, we have found that fewer than one in four alliances have adequate performance metrics in place.2 As a result, alliances tend to be run by intuition and with incomplete information. Partners may not agree about the progress of their ventures and senior management can’t intervene quickly enough to correct problems. In many companies, 30 to 60 percent of alliances are underperforming—and three to five major deals are in desperate need of restructuring. Unfortunately, management does not really know where the problems lie or how it should best invest its time. To get a better grip on performance, companies must develop a more structured approach to evaluating the health of their alliances. Doing so is not a straightforward task.

The challenge

Why is alliance performance so hard to measure? The answer can be traced to three characteristics. First, alliances are by definition arrangements between separate companies, each employing its own reporting processes and systems and each pursuing its own goals for the alliance. This makes it hard to agree on a single measure of performance and creates incentives to conceal benefits and inflate costs.

A second critical characteristic is an operational interdependence that makes benefits and costs difficult to track. Most alliances receive some inputs from their parent companies (including raw materials, customer data, and administrative services) and in turn provide outputs to them, creating complicated transfer-pricing issues. Before Airbus Industrie was revamped in 2001, for instance, the four consortium members made aircraft sections and “sold” them to the joint venture, which then assembled and marketed the airplanes. Setting transfer prices was a challenge because of the partners’ sensitivity to sharing detailed cost data. In another case, two global technology companies agreed to jointly market a new product. This alliance involved more than 30 working teams whose 300-odd members spent between 20 percent and 60 percent of their time on the alliance. One executive admitted that he had no real idea how much the company had spent on the venture, so large were its hidden costs. Measuring benefits is no easier. Alliances often generate sales of related products for parent companies, which should also be taken into account in assessing performance and value. So should longer term benefits like opportunities for learning, access to new technologies and markets, and improved competitive positioning.

The third characteristic is the noncore position of alliances within the corporate portfolio. Because they are inside the corporation, not totally outside, they often do not receive the same level of management scrutiny as business units and internal initiatives; nor are they subject to the same level of market forces as standard customer or supplier relationships.

Measuring performance

To overcome these difficulties, companies must assess the performance of their alliances on three levels, each focusing on different aspects of the problem and prompting distinct managerial responses.

At the first level, every alliance should be individually assessed to establish how it is performing and whether the parent company needs to intervene. At the second level, a company should periodically search for performance patterns across the portfolio—a process that often leads to adjustments in the types of deals a company pursues and sometimes to additional investments in a drive to build alliance-related skills. At the third level, once a company better understands how its portfolio is performing, it can conduct a top-down review of overall strategy to ensure not only that its alliance portfolio is configured for optimal performance, but also that it has ranked new opportunities in a clear order of priority. The following excerpt focuses on the performance of individual alliances.3

Developing a detailed view of the economics of an alliance is indispensable to measuring its performance. This measurement should go well beyond the usual cash flow metrics to include transfer-pricing benefits, benefits outside the scope of the deal (for instance, sales of related products), the value of options created by the alliance, as well as start-up and ongoing management costs (Exhibit 1). This information is vital for managers to evaluate deals up front and to monitor their continuing performance.

For example, one company in the power industry calculated the embedded option value of a potential alliance to commercialize a critical new technology, evaluating the odds of different possible outcomes and the associated payoffs for each. Its calculation showed that for the given alliance the firm had a 10 percent chance of creating $1 billion in annual income within three years, a 20 percent chance of creating a modestly successful business producing $10 million to $30 million in annual income, a 60 percent chance of losing $10 million to $30 million, and a 10 percent chance of losing more than $200 million. This profile of potential option value was extremely sensitive to assumptions regarding technology and construction costs, leading the company to closely monitor the alliance’s early performance, while reserving the right to cut off funding in the event that technical progress slowed.

Likewise, one biotechnology company developed explicit values for the potential learning benefits from a planned development and marketing alliance with a large pharmaceutical company. This exercise provided the information that ultimately pushed the firm to choose one partner over another and led it to closely monitor the benefits of the alliance, eventually allowing the firm to “migrate” into new capabilities.

Having formed an alliance with a clear and integrated view of the economics in mind, a company must develop, within 30 days of the launch, a scorecard to track the venture’s performance. Partners must decide whether to share a single scorecard, to run separate scorecards, or to use a combination of the two. For a joint venture with its own P&L, a single scorecard is often possible. For most other alliances, the combination approach works best. Each partner can supplement a shared scorecard with additional metrics that track progress against goals that aren’t shared by the other partners. This approach also enables each partner to devise internal metrics that allow it to compare the performance of an alliance with the performance of business activities outside the alliance or to other, similar alliances.

It is essential, both at the alliance and the parent level, to take a balanced view of performance. To achieve such a balance, we have found it useful to include four dimensions of performance fitness: financial, strategic, operational, and relationship. Financial and strategic metrics show how the alliance is performing and whether it is meeting its goals—but may not provide enough insight into exactly what, if anything, isn’t going well. Operational and relationship metrics can help uncover the first signs of trouble and reveal the causes of problems. Together, the four dimensions of performance create an integrated picture that has proved invaluable to the relatively few companies, such as Siebel Systems, that have used them to measure the health of alliances (Exhibit 2).

Exhibit

1. Financial fitness: Metrics such as sales revenues, cash flow, net income, return on investment, and the expected net present value of an alliance measure its financial fitness. Most alliances should also monitor progress in reducing overlapping costs, achieving purchasing discounts, or increasing revenues. In addition, financial fitness can include partner-specific metrics such as transfer-pricing revenues and sales of related products by the parent companies. At one international oil industry joint venture, the partners tracked not only revenues and consolidation synergies on a quarterly basis, but also the costs of goods sold to and from the parents, as well as estimates of profitability on those parent-related transactions.

2. Strategic fitness: Nonfinancial metrics such as market share, new-product launches, and customer loyalty can help executives measure the strategic fitness of a deal; other metrics could, for example, track the competitive positioning and access to new customers or technologies resulting from it. Devising strategic metrics can take imagination. The international semiconductor research consortium SEMATECH, for instance, tracks the number of employees from member companies who are working on its research initiatives in order to assess whether it is transferring knowledge to its partners.

3. Operational fitness: The number of customers visited and staff members recruited, the quality of products and manufacturing throughput are examples of operational fitness metrics that call for explicit goals linked to the performance reviews and compensation of individuals. For example, executives at one health care company define operationally-fit alliances as those hitting 60 to 80 percent of their key operating milestones. Any figure higher than 80 percent indicates that the goals weren’t sufficiently ambitious.

4. Relationship fitness: Questions about the cultural fit and trust between partners, the speed and clarity of their decision making, the effectiveness of their interventions when problems arise, and the adequacy with which they define and deliver their contributions all fall under the heading of relationship fitness. To measure it, Siebel Systems developed a sophisticated partner-satisfaction survey, sent each quarter to key managers of alliance partners, that contains more than 80 questions about issues such as alliance management and partners’ loyalty to Siebel. The company uses this information to spot problems and to develop detailed action plans to address them.

The weight placed on each type of metric and the amount of detail it includes depend on the size and aims of the alliance. A consolidation joint venture whose main goal is to reduce costs, for instance, should focus heavily on financial and operational metrics. But managers of an alliance entering a new market expect negative financial returns in the early stages and should give more weight to strategic goals such as increasing market share and penetrating distribution channels. Smaller, short-term alliances might have simple scorecards with only four or five metrics; larger ventures with substantial assets or revenues deserve something more detailed.

Scorecard results provide important clues to what might be going wrong with an alliance, but uncovering the true problem often requires further investigation. For example, a large media company found that the hundreds of millions of dollars it had invested in alliances were at risk when close scrutiny revealed that five of its ten most important deals were losing money. In addition, two joint ventures with an international media company were found to have been troubled by flawed deal structures from the start. Further probing found that three unprofitable alliances could be renegotiated, saving $23 million a year, and that redefining each joint venture partner’s contributions and responsibilities could save another $45 million a year. Subsequently, the company established a corporate-level alliance unit to keep a critical eye on all of its ventures.


At a time when alliances are increasingly important, continuing to ignore their performance is simply not an option. Instead, managers should systematically measure the performance of each individual alliance to ensure that the maximum value is derived and management is able to intervene when a deal veers off track. Experience has proved that the effort pays substantial dividends.

About the author(s)

Jim Bamford is a consultant and David Ernst is a principal in McKinsey’s Washington, DC, office.

The authors would like to thank David Dorton, Tammy Halevy, C. Brent Hastie, and Eric A. Kutcher for their contribution to this article.

This article was first published in the Autumn 2002 issue of McKinsey on Finance. Visit McKinsey’s corporate finance site to view the full issue.