The Performance Measurement Manifesto

Revolutions begin long before they are officially declared. For several years, senior executives in a broad range of industries have been rethinking how to measure the performance of their businesses. They have recognized that new strategies and competitive realities demand new measurement systems. Now they are deeply engaged in defining and developing those systems for their companies. At the heart of this revolution lies a radical decision: to shift from treating financial figures as the foundation for performance measurement to treating them as one among a broader set of measures. Put like this, it hardly sounds revolutionary. Many managers can honestly claim that they—and their companies—have tracked quality, market share, and other nonfinancial measures for years. Tracking these measures is one thing. But giving them equal (or even greater) status in determining strategy, promotions, bonuses, and other rewards is another. Until that happens, to quote Ray Stata, the CEO of Analog Devices, “When conflicts arise, financial considerations win out.”1 The ranks of companies enlisting in this revolution are rising daily. Senior managers at one large, high-tech manufacturer recently took direct responsibility for adding customer satisfaction, quality, market share, and human resources to their formal measurement system. The impetus was their realization that the company’s existing system, which was largely financial, undercut its strategy, which focused on customer service. At a smaller manufacturer, the catalyst was a leveraged recapitalization that gave the CEO the opportunity formally to reorder the company’s priorities. On the new list, earnings per share dropped to last place, preceded by customer satisfaction, cash flow, manufacturing effectiveness, and innovation (in that order). On the old list, earnings per share stood first and almost alone. In both companies, the CEOs believe they have initiated a sea change in how their managers think about business performance and in the decisions they make. Executives at other companies engaged in comparable efforts feel the same—rightly. What gets measured gets attention, particularly when rewards are tied to the measures. Grafting new measures onto an old accounting-driven performance system or making slight adjustments in existing incentives accomplishes little. Enhanced competitiveness depends on starting from scratch and asking: “Given our strategy, what are the most important measures of performance?” “How do these measures relate to one another?” “What measures truly predict long-term financial success in our businesses?” Dissatisfaction with using financial measures to evaluate business performance is nothing new. As far back as 1951, Ralph Cordiner, the CEO to of General Electric, commissioned a high-level task force to identify key corporate performance measures. (The categories the task force singled out were timeless and comprehensive: in addition to profitability, the list included market share, productivity, employee attitudes, public responsibility, and the balance between short-and long-term goals.) But the current wave of discontent is not just more of the same. One important difference is the intensity and nature of the criticism directed at traditional accounting systems. During the past few years, academics and practitioners have begun to demonstrate that accrual-based performance measures are at best obsolete—and more often harmful.2 Diversity in products, markets, and business units puts a big strain on rules and theories developed for smaller, less complex organizations. More dangerously, the numbers these systems generate often fail to support the investments in new technologies and markets that are essential for successful performance in global markets. Such criticisms reinforce concern about the pernicious effects of short-term thinking on the competitiveness of U.S. companies. Opinions on the causes of this mind-set differ. Some blame the investment community, which presses relentlessly for rising quarterly earnings. Others cite senior managers themselves, charging that their typically short tenure fosters shortsightedness. The important point is that the mind-set exists. Ask almost any senior manager and you will hear about some company’s failure to make capital investments or pursue long-term strategic objectives that would imperil quarterly earnings targets. Moreover, to the extent that managers do focus on reported quarterly earnings—and thereby reinforce the investment community’s short-term perspective and expectations—they have a strong incentive to manipulate the figures they report. The extent and severity of such gaming is hard to document. But few in management deny that it goes on or that managers’ willingness to play the earnings game calls into question the very measures the market focuses on to determine stock prices. For this reason, many managers, analysts, and financial economists have begun to focus on cash flow in the belief that it reflects a company’s economic condition more accurately than its reported earnings do.3 Finally, many managers worry that income-based financial figures are better at measuring the consequences of yesterday’s decisions than they are at indicating tomorrow’s performance. Events of the past decade substantiate this concern. During the 1980s, many executives saw their companies’ strong financial records deteriorate because of unnoticed declines in quality or customer satisfaction or because global competitors ate into their market share. Even managers who have not been hurt feel the need for preventive action. A senior executive at one of the large money-center banks, for example, grew increasingly uneasy about the European part of his business, its strong financials notwithstanding. To address that concern, he has nominated several new measures (including customer satisfaction, customers’ perceptions of the bank’s stature and professionalism, and market share) to serve as leading indicators of the business’s performance. Discontent turns into rebellion when people see an alternative worth fighting for. During the 1980s, many managers found such an alternative in the quality movement. Leading manufacturers and service providers alike have come to see quality as a strategic weapon in their competitive battles. As a result, they have committed substantial resources to developing measures such as defect rates, response time, delivery commitments, and the like to evaluate the performance of their products, services, and operations. In addition to pressure from global competitors, a major impetus for these efforts has been the growth of the Total Quality Movement and related programs such as the Malcolm Baldrige National Quality Award. (Before a company can even apply for a Baldrige Award, it must devise criteria to measure the performance of its entire operation—not just its products—in minute detail.) Another impetus, getting stronger by the day, comes from large manufacturers who are more and more likely to impose rigid quality requirements on their suppliers. Whatever the stimulus, the result is the same: quality measures represent the most positive step taken to date in broadening the basis of business performance measurement. Another step in the same direction comes from embryonic efforts to generate measures of customer satisfaction. What quality was for the 1980s, customer satisfaction will be for the 1990s. Work on this class of measures is the highest priority at the two manufacturing companies discussed earlier. It is equally critical at another high-tech company that recently created a customer satisfaction department reporting directly to the CEO. In each case, management’s interest in developing new performance measures was triggered by strategies emphasizing customer service. As competition continues to stiffen, strategies that focus on quality will evolve naturally into strategies based on customer service. Indeed, this is already happening at many leading companies. Attention to customer satisfaction, which measures the quality of customer service, is a logical next step in the development of quality measures. Companies will continue to measure quality on the basis of internally generated indexes (such as defect rates) that are presumed to relate to customer satisfaction. But they will also begin to evaluate their performance by collecting data directly from customers for more direct measures like customer retention rates, market share, and perceived value of goods and services. Just as quality-related metrics have made the performance measurement revolution more real, so has the development of competitive benchmarking.4 First, benchmarking gives managers a methodology that can be applied to any measure, financial or non-financial, but that emphasizes nonfinancial metrics. Second (and less obvious), it has a transforming effect on managerial mind-sets and perspectives. Benchmarking involves identifying competitors and/or companies in other industries that exemplify best practice in some activity, function, or process and then comparing one’s own performance to theirs. This externally oriented approach makes people aware of improvements that are orders of magnitude beyond what they would have thought possible. In contrast, internal yardsticks that measure current performance in relation to prior period results, current budget, or the results of other units within the company rarely have such an eye-opening effect. Moreover, these internally focused comparisons have the disadvantage of breeding complacency through a false sense of security and of stirring up more energy for intramural rivalry than for competition in the marketplace. Finally, information technology has played a critical role in making a performance measurement revolution possible. Thanks to dramatically improved price-performance ratios in hardware and to breakthroughs in software and database technology, organizations can generate, disseminate, analyze, and store more information from more sources, for more people, more quickly and cheaply than was conceivable even a few years back. The potential of new technologies, such as hand-held computers for employees in the field and executive information systems for senior managers, is only beginning to be explored. Overall, the range of measurement options that are economically feasible has radically increased. Veterans know it is easier to preach revolution than to practice it. Even the most favorable climate can create only the potential for revolutionary change. Making it happen requires conviction, careful preparation, perseverance, and a decided taste for ambiguity. As yet, there are no clear-cut answers or predetermined processes for managers who wish to change their measurement systems. Based on the experience of companies engaged in this revolution, I can identify five areas of activity that sooner or later need to be addressed: developing an information architecture; putting the technology in place to support this architecture; aligning incentives with the new system; drawing on outside resources; and designing a process to ensure that the other four activities occur. Developing a new information architecture must be the first activity on any revolutionary agenda. Information architecture is an umbrella term for the categories of information needed to manage a company’s businesses, the methods the company uses to generate this information, and the rules regulating its flow. In most companies, the accounting system implicitly defines the information architecture. Other performance measures are likely to be informal—records that operating managers keep for themselves, for instance—and they are rarely integrated into the corporate-driven financial system. The design for a new corporate information architecture begins with the data that management needs to pursue the company’s strategy. This may sound like a truism, but a surprising number of companies describe their strategies in terms of customer service, innovation, or the quality and capabilities of their people, yet do little to measure these variables. Even time—the newest strategic variable—remains largely underdeveloped in terms of which time-based metrics are most important and how best to measure them. As part of this identification process, management needs to articulate a new corporate grammar and define its own special vocabulary—the basic terms that will need to be common and relatively invariant across all the company’s businesses. Some of these terms (like sales and costs) will be familiar. Others, however, will reflect new strategic priorities and ways to think about measuring performance. For example, both a large money-center bank and a multidivisional, high-technology manufacturer introduced the use of cross-company customer identification numbers so they could readily track such simple and useful information as the total amount of business the company did with any one customer. It sounds elementary and it is—as soon as you start to look at the entire measurement system from scratch. Uniformity can be carried too far. Different businesses with different strategies require different information for decision making and performance measurement. But this should not obscure the equally obvious fact that every company needs to have at least a few critical terms in common. Today few large companies do. Years of acquisitions and divestitures, technological limitations, and at times, a lack of management discipline have all left most big organizations with a complicated hodgepodge of definitions and variables—and with the bottom line their only common denominator. Developing a coherent, companywide grammar is particularly important in light of an ever-more stringent competitive environment. For many companies, ongoing structural reorganizations are a fact of life. The high-technology company described above has reorganized itself 24 times in the past 4 years (in addition to a number of divisional and functional restructurings) to keep pace with changes in its markets and technologies. Rather than bewail the situation, managers relish it and see their capacity for fast adaptation as an important competitive advantage. A common grammar also enhances management’s ability to break apart and recombine product lines and market segments to form new business units. At a major merchant bank, for example, the organization is so fluid that one senior executive likens it to a collection of hunting packs that form to pursue business opportunities and then disband as the market windows on those opportunities close. The faster the company can assemble information for newly formed groups, the greater the odds of success. So this executive (who calls himself the czar of information) has been made responsible for developing standard definitions for key information categories. How a company generates the performance data it needs is the second piece of its information architecture. Not surprisingly, methods for measuring financial performance are the most sophisticated and the most deeply entrenched. Accountants have been refining these methods ever since double-entry bookkeeping was invented in the fifteenth century. Today their codifications are enforced by a vast institutional infrastructure made up of professional educators, public accounting firms, and regulatory bodies. In contrast, efforts to measure market share, quality, innovation, human resources, and customer satisfaction have been much more modest. Data for tracking these measures are generated less often: quarterly, annual, or even biannual bases are common. Responsibility for them typically rests with a specific function. (Strategic planning measures market share, for example, while engineering measures innovation and so on.) They rarely become part of the periodic reports general managers receive. Placing these new measures on an equal footing with financial data takes significant resources. One approach is to assign a senior executive to each of the measures and hold him or her responsible for developing its methodologies. Typically, these executives come from the function that is most experienced in dealing with the particular measure. But they work with a multifunctional task force to ensure that managers throughout the company will understand the resulting measures and find them useful. Another, less common, approach is to create a new function focused on one measure and then to expand its mandate over time. A unit responsible for customer satisfaction might subsequently take on market share, for example, or the company’s performance in human resources. Unlike a company’s grammar, which should be fairly stable, methods for taking new performance measures should evolve as the company’s expertise increases. Historical comparability may suffer in the process, but this is a minor loss. What matters is how a company is doing compared with its current competitors, not with its own past. The last component of a corporate information architecture is the set of rules that governs the flow of information. Who is responsible for how measures are taken? Who actually generates the data? Who receives and analyzes them? Who is responsible for changing the rules? Because information is an important source of power, the way a company answers these questions matters deeply. How open or closed a company is affects how individuals and groups work together, as well as the relative influence people and parts of the company have on its strategic direction and management. Some companies make information available on a very limited basis. At others, any individual can request information from another unit as long as he or she can show why it is needed. Similarly, in some companies the CEO still determines who gets what information—not a very practical alternative in today’s world. More often what happens is that those who possess information decide with whom they will share it. Advances in information technology such as powerful workstations, open architectures, and relational databases vastly increase the options for how information can flow. It may be centralized at the top, so that senior executives can make even more decisions than they have in the past. Or it may be distributed to increase the decision-making responsibilities of people at every level. The advantages of making information widely available are obvious, though this also raises important questions that need to be addressed about the data’s integrity and security. In principle, however, this portion of the information architecture ought to be the most flexible of the three, so that the company’s information flows continue to change as the conditions it faces do. Determining the hardware, software, and telecommunications technology a company needs to generate its new measurement information is the second activity in the performance revolution. This task is hard enough in its own right, given the many choices available. But too often managers make it even harder by going directly to a technology architecture without stopping first to think through their information needs. This was the case at a high-tech manufacturing company that was growing more and more frustrated with its information systems planning committee. Then the CEO realized that he and the other senior managers had not determined the measures they wanted before setting up the committee. Equipped with that information, the committee found it relatively easy to choose the right technology. Once the information architecture and supporting technology are in place, the next step is to align the new system with the company’s incentives—to reward people in proportion to their performance on the measures that management has said truly matter. This is easier said than done. In many companies, the compensation system limits the amount and range of the salary increases, bonuses, and stock options that management can award. In companies that practice pay-for-performance, compensation and other rewards are often tied fairly mechanically to a few key financial measures such as profitability and return on investment. Convincing managers that a newly implemented system is really going to be followed can be a hard sell. The president of one service company let each of his division general managers design the performance measures that were most appropriate for his or her particular business. Even so, the managers still felt the bottom line was all that would matter when it came to promotions and pay. The difficulty of aligning incentives to performance is heightened by the fact that formulas for tying the two together are rarely effective. Formulas have the advantage of looking objective, and they spare managers the unpleasantness of having to conduct truly frank performance appraisals. But if the formula is simple and focuses on a few key variables, it inevitably leaves some important measures out. Conversely, if the formula is complex and factors in all the variables that require attention, people are likely to find it confusing and may start to play games with the numbers. Moreover, the relative importance of the variables is certain to change more often—and faster—than the whole incentive system can change. For these reasons, I favor linking incentives strongly to performance but leaving managers free to determine their subordinates’ rewards on the basis of all the relevant information, qualitative as well as quantitative. Then it is up to the manager to explain candidly to subordinates why they received what they did. For most managers, this will also entail learning to conduct effective performance appraisals, an indirect—and invaluable—benefit of overhauling the measurement system. Outside parties such as industry and trade associations, third-party data vendors, information technology companies, consulting firms, and public accounting firms must also become part of the performance measurement revolution. Their incentive: important business opportunities. Industry and trade associations can play a very helpful role in identifying key performance measures, researching methodologies for taking these measures, and supplying comparative statistics to their members—so can third-party data vendors. Competitors are more likely to supply information to a neutral party (which can disguise it and make it available to all its members or customers) than to one another. And customers are more likely to provide information to a single data vendor than to each of their suppliers separately. Consulting firms and information technology vendors also have important roles to play in forwarding the revolution. Firms that specialize in strategy formulation, for example, often have well-developed methods for assessing market share and other performance metrics that clients could be trained to use. Similarly, firms that focus on strategy implementation have a wealth of experience designing systems of various kinds for particular functions such as manufacturing and human resources. While many of these firms are likely to remain specialized, and thus require coordination by their clients, others will surely expand their capabilities to address all the pieces of the revolution within a client company. Much the same thing is apt to happen among vendors of information technology. In addition to helping companies develop the technological architecture they need, some companies will see opportunities to move into a full range of services that use the hardware as a technology platform. IBM and DEC are already moving in this direction, impelled in part by the fact that dramatic gains in price-performance ratios make it harder and harder to make money selling “boxes.” Finally, public accounting firms have what may be the single most critical role in this revolution. On one hand, they could inhibit its progress in the belief that their vested interest in the existing system is too great to risk. On the other hand, all the large firms have substantial consulting practices, and the revolution represents a tremendous business opportunity for them. Companies will need a great deal of help developing new measures, validating them, and certifying them for external use. Accounting firms also have an opportunity to develop measurement methods that will be common to an industry or across industries. While this should not be overdone, one reason financial measures carry such weight is that they are assumed to be a uniform metric, comparable across divisions and companies, and thus a valid basis for resource allocation decisions. In practice, of course, these measures are not comparable (despite the millions of hours invested in efforts to make them so) because companies use different accounting conventions. Given that fact, it is easy to see why developing additional measures that senior managers—and the investment community—can use will be a massive undertaking. Indeed, the power of research analysts and investors generally is one of the reasons accounting firms have such a crucial role to play. Although evidence exists that investors are showing more interest in metrics such as market share and cash flow, many managers and analysts identify the investment community as the chief impediment to revolution.5 Until investors treat other measures as seriously as financial data, they argue, limits will always exist on how seriously those measures are taken inside companies. GE’s experience with its measurement task force supports their argument. According to a knowledgeable senior executive, the 1951 effort had only a modest effect because the measures believed to determine the company’s stock price, to which incentives were tied, were all financial: earnings per share, return on equity, return on investment, return on sales, and earnings growth rate. He believed that once the financial markets valued other measures, progress within companies would accelerate. Investors, of course, see the problem from a different perspective. They question whether managers would be willing to publish anything more than the financial information required by the SEC lest they reveal too much to their competitors. Ultimately, a regulatory body like the SEC could untie this Gordian knot by recommending (and eventually requiring) public companies to provide nonfinancial measures in their reports. (This is, after all, how financial standards became so omnipotent and why so many millions of hours have been invested in their development.) But I suspect competitive pressure will prove a more immediate force for change. As soon as one leading company can demonstrate the long-term advantage of its superior performance on quality or innovation or any other nonfinancial measure, it will change the rules for all its rivals forever. And with so many serious competitors tracking—and enhancing—these measures, that is only a matter of time. Designing a process to ensure that all these things happen is the last aspect of the revolution. To overcome conservative forces outside the company and from within (including line and staff managers at every level, in every function), someone has to take the lead. Ultimately, this means the CEO. If the CEO is not committed, the revolution will flounder, no matter how much enthusiasm exists throughout the organization. But the CEO cannot make it happen. Developing an information architecture and its accompanying technology, aligning incentives, working with outside parties—all this requires many people and a lot of work, much of it far less interesting than plotting strategy. Moreover, the design of the process must take account of the integrative nature of the task: people in different businesses and functions including strategic planning, engineering, manufacturing, marketing and sales, human resources, and finance will all have something to contribute. The work of external players will have to be integrated with the company’s own efforts. Organizationally, two critical choices exist. One is who the point person will be. Assigning this role to the CEO or president ensures its proper symbolic visibility. Delegating it to a high-level line or staff executive and making it a big piece of his or her assignment may be a more effective way to guarantee that enough senior management time will be devoted to the project. The other choice is which function or group will do most of the work and coordinate the company’s efforts. The CEO of one high-tech company gave this responsibility to the finance function because he felt they should have the opportunity to broaden their perspective and measurement skills. He also thought it would be easier to use an existing group experienced in performance measurement. The president of an apparel company made a different choice. To avoid the financial bias embedded in the company’s existing management information systems, he wanted someone to start from scratch and design a system with customer service at its core. As a result, he is planning to combine the information systems department with customer service to create a new function to be headed by a new person, recruited from the outside. What is most effective for a given company will depend on its history, culture, and management style. But every company should make the effort to attack the problem with new principles. Some past practices may still be useful, but everything should be strenuously challenged. Otherwise, the effort will yield incremental changes at best. Open-mindedness about the structures and processes that will be most effective, now and in the future, is equally important. I know of a few companies that are experimenting with combining the information systems and human resource departments. These experiments have entailed a certain amount of culture shock for professionals from both functions, but such radical rethinking is what revolution is all about. Finally, recognize that once begun, this is a revolution that never ends. We are not simply talking about changing the basis of performance measurement from financial statistics to something else. We are talking about a new philosophy of performance measurement that regards it as an ongoing, evolving process. And just as igniting the revolution will take special effort, so will maintaining its momentum—and reaping the rewards in the years ahead. 1. Ray Stata, “Organizational Learning—The Key to Management Innovation,” Sloan Management Review, Spring 1989, pp. 63–74. 2. Donald A. Curtis, “The Modern Accounting System,” Financial Executive, January–February 1985, pp. 81–93; and H. Thomas Johnson and Robert S. Kaplan, Relevance Lost (Boston: Harvard Business School Press, 1987). 3. Yuji Ijiri, “Cash Flow Accounting and Its Structure,” Journal of Accounting, Auditing, and Finance, Summer 1978, pp. 331–348. 4. Robert C. Camp, Benchmarking (Milwaukee, Wisconsin: ASQS Quality Press, 1989). 5. “Investors: Look at Firms’ Market Share,” Wall Street Journal, February 26, 1990, pp. C1–2.

The Performance Measurement Manifesto

Partner Center