Performance Measurement

Outlook Special Edition, October 1999

Executives have long believed that alliances succeed on the power of intangibles—qualities such as trust, culture and communication. At the same time, some executives have also realized that each alliance is different and thus needs its own measures of performance.

These two assumptions—neither incorrect—have led executives to conclude that alliance performance is hard to measure and impossible to benchmark. Indeed, our data show that only 51 percent of alliances use formal performance measures, and of those that do, just 20 percent of executives involved believe the measures to be sufficient. All told, barely 10 percent of alliances have meaningful measures of performance.

To be sure, alliances are hard to measure—but for different reasons than most executives cite. In alliances, for instance, many resources and deliverables have no attached market price. Partners are forced to value such elusive resources as contacts, know-how, managerial advice, market positions and other intangibles to create a true picture of performance.

Meanwhile, many benefits often accumulate outside the alliance. For example, parent companies commonly apply the gains of an alliance to nonalliance activities, often without disclosure to the partner. For instance, in one oil exploration joint venture in the Gulf of Mexico, one partner uses the business as a training ground for its new engineers. Once the engineers gain certain skills, they are sent overseas to work in one of that corporate parent's other businesses—where the joint venture partner has no involvement.

Likewise, alliance performance is hard to measure because companies have an incentive to withhold information about the value of the alliance, lest their partners want to share in, or be compensated for, this value.

Finally, there is the lack of convention. In most cases, alliances have not received detailed attention from accountants or other auditors who could impose a more standard view on performance assessment.

Despite the difficulties, alliance performance can and must be measured. After all, performance measurement is a critical tool for managing any entity; it would be dangerous to presume that alliance performance is an exception.

Although the economics of alliances is a new and imperfect science and not every alliance will be measured in the same way, we believe that universal principles of good alliance measurement exist and that companies should follow them. These practices include developing a balanced scorecard, building a dollar defense and accounting for surplus value.

A Balanced Scorecard
The nature of alliances—multifaceted and fluid—makes it important for companies to monitor performance from a variety of perspectives. Our client work shows that companies can gain substantial focus by applying a "balanced scorecard" technique to alliances. This is an approach that has been used successfully in cross-functional teams and other complex, internal activities where many different facets have to form a coherent mosaic.

To understand how this approach can work for alliances, consider how one energy company created a balanced scorecard to measure its alliance performance. Management established 10 to 15 performance measures for each alliance. The number was large enough to incorporate the alliances' diverse interests but limited enough for busy executives to easily monitor. Each set of performance measures provided a balanced combination of perspectives: financial and strategic measures, short-term and long term measures, and measures that monitored processes and end results. At least one measure was used to focus on customer satisfaction.

Under the balanced scorecard concept, we also suggest three specific techniques that are critical to meaningful performance measurement:

1. Measure the Effect on Share Price. Because it is so difficult to pin down, one performance measure that few companies use is the effect of an alliance on company share price. And yet, linking alliances to share price is the ultimate measure because changes in market capitalization are the most universal measure of corporate performance. We believe that all public companies should include the direct effect of the announcement of an alliance on shareholder value in the set of performance measures developed for each alliance. As a step in this direction, we, along with professors from Harvard and Yale, have developed the Partnership Value Assessment, a statistical tool that isolates the effect of alliance announcements on company share price (See "Linking Performance to Share Price").

2. Elevate a Few Measures. No matter which measures are used for a balanced scorecard, it is important for companies to elevate two or three of them to the status of red flags that quickly indicate when there are critical problems. The selected measures become the poster children for the alliance, focal points for all those who have a stake in alliance performance, including the alliance staff, parent company boards and the media.

The power of the simple message conveyed by just a few selected measures can be seen in the results achieved by one best practitioner, SEMATECH, a semiconductor research alliance. For the last five years, it has focused its member firms on two performance measures: member firm return on investment and individual project performance. By providing the discipline of focusing on a few measures (and the internal infrastructure to measure them), the alliance has increased its cohesiveness, flexibility and ability to respond to market conditions. Proof lies in SEMATECH's ability to have survived the withdrawal of government funding and to have attracted a series of new international member firms.

3. Link to Individual Incentives. Performance measures do not get much attention without a direct link to reward systems. Devising appropriate rewards for alliance performance generally is a straightforward exercise once the measures are selected. For optimal effectiveness, companies should ensure that at least a few executives are evaluated on the basis of the full balanced scorecard of between 10 and 15 performance measures.

One interesting practice has emerged in the pharmaceutical industry. On occasion, the corporate parents have allocated a small pool of funds to the alliance steering committee, which, in turn, used this money to reward the alliance managers (managers from each parent who are focused full-time on the alliance). This motivates the managers to strive for excellence on the full set of performance measures, helps align their individual incentives and increases teamwork.

A Dollar Defense
Successful alliances create a "dollar defense"—an indisputable, stable source of financial value that allows the alliance to pursue strategic benefits without living under the constant threat of venture dissolution. A dollar defense also sends a strong message to the financial markets that the alliance continues to create significant value. Such sources of value can include relatively impervious cost or revenue enhancers like new sales, purchasing discounts and reduced overhead costs that likely will survive competitive onslaughts as long as the alliance exists.

Surplus Value
Parent companies need to address the issue of surplus value—that is, the value generated by alliances that often goes unrecognized because it goes unnoticed or because only one partner sees it. One common example is the halo effect a small partner (say, a biotechnology company or an emerging market airline) receives from an alliance with a large multinational. In effect, the alliance is an endorsement of the small company, often raising investor enthusiasm, business contacts and overall market opportunities.

How can companies incorporate this kind of surplus value into the performance-measurement process? A few simple steps can help. During the alliance negotiations, companies should draw up at least three sets of performance measures—one for the alliance and one for each parent. This will show that much of the potential value of the alliance will be captured separately by the parents. At a minimum, this will broaden the discussion about the appropriate performance measures and division of the profits.

Meanwhile, recalling that alliances are ultimately about self-interest, companies should try to maximize their own surplus value. But remember, the name of the game is enlightened self-interest: This value should come at no expense to the partner, since value derived from using know-how against a partner may destroy the foundation of cooperation.

Clearly, alliances present real challenges in measuring performance—firms must account for different corporate interests, work in short time frames and deal with nonfinancial benefits and surplus value. Difficult as it may be, however, it is simply not possible to have consistent success in alliances without overcoming these issues.

Linking Performance to Share Price

Linking alliances to share price is the ultimate measure of their performance (see story). The Partnership Value Assessment measurement method strips out all external factors that affect share price to focus solely on the effect of an alliance's debut. It measures the activity of the parent companies' shares in the days immediately before and after the alliance is announced.

This tool was developed by Accenture and is based on an analysis of work done by professors Tarun Khanna of Harvard Business School and Bharat Anand of Yale School of Management. It included analysis of the historical data covering nearly 2,000 alliances (1,106 licensing arrangements and 870 joint ventures) over a four-year span.

The research demonstrated systematically what some executives know intuitively: that an individual firm's alliance portfolio can have tremendous implications for shareholder value, both positive and negative.

The PVA methodology comprises five basic steps:

1. Alliance participants' "ordinary returns" are estimated, using prior-year historical stock price growth (proxied by using S&P 500 returns and the stock's beta and alpha).

2. Stock price changes surrounding the announcement date of the alliance are measured to determine the firm's "abnormal returns" that could be attributed to the announcement.

3. The ordinary returns are then subtracted from the abnormal returns to arrive at the firm's "excess returns" from the alliance.

4. These excess returns are then used in conjunction with the firm's total market value of equity to calculate the total shareholder value (in dollars).

5. All the participating firms' profiles and excess returns are then aggregated and disaggregated along a number of dimensions to analyze the value-creating effect of various industry factors and alliance features.

To form and maintain successful alliances these days, companies will need to discard the myths we have questioned here and embrace the new realities. In many cases, this will require new skills, structures and perspectives that most companies, especially industry incumbents, do not yet possess. As with any rapidly evolving situation in the marketplace, someone will seize these new opportunities. If traditional companies do not, entirely new participants—many with deep alliance experience and unhindered by old myths—certainly will.

Back to Contents

To Top