A decision framework provides the logical connections between decisions about a resource (for example, financial capital, customers, or talent) and the strategic success of the organization. This is true in HR, as we show in subsequent chapters that describe such connections in various domains of HR. It is also true in other more familiar decision sciences such as finance and marketing. It is instructive to compare HR to these other disciplines. Figure 1-1 shows how a decision framework for talent and HR, which Boudreau and Ramstad called “talentship,” has a parallel structure to decision frameworks for finance and marketing.

Reprinted by permission of Harvard Business School Press, from Beyond HR: The New Science of Human Capital by John Boudreau and Peter M. Ramstad. Boston, MA, 2007, pp. 31.

Figure 1-1. Finance, Marketing, and Talentship Decision Frameworks.

Finance is a decision science for the resource of money, marketing is the decision science for the resource of customers, and talentship is the decision science for the resource of talent. In all three decision sciences, the elements combine to show how one factor interacts with others to produce value.

To illustrate the logic of such a framework, consider marketing as an example. Investments in marketing produce a product, promotion, price, and placement mix, which creates responses in certain customer segments, which in turn creates changes in the lifetime profits from those customers. Similarly, with regard to talent decisions, efficiency describes the connection between investments in people and the talent-related programs and practices they produce (such as cost per training hour). Effectiveness describes the connection between the programs/practices and the changes in the talent quality or organizational characteristics (such as whether trainees increase their skill or their interactions with others in the organization). Impact describes the connection between the changes in talent/organization elements and the strategic success of the organization (such as whether increased skill actually enhances the organizational processes or initiatives that are most vital to strategic success). The chapters in this book show how to measure not just HR efficiency, but also elements of effectiveness and impact. In addition, each chapter provides a logical framework for the measures, to enhance decision making and organizational change.

In a well-developed decision science, the measures and data are deployed through management systems, they are used by leaders who understand the principles, and they are supported by professionals who add insight and expertise. In stark contrast, HR data, information, and measurement face a paradox today. There is increasing sophistication in technology, data availability, and the capacity to report and disseminate HR information, but investments in HR data systems, scorecards, and integrated enterprise resource systems fail to create the strategic insights needed to drive organizational effectiveness. HR measures exist mostly in areas where the accounting systems require information to control labor costs or to monitor functional activity. Efficiency gets a lot of attention, but effectiveness and impact are often unmeasured. In short, many organizations are “hitting a wall” in HR measurement.

Type “HR measurement” into a search engine and you will get more than 900,000 results. Scorecards, summits, dashboards, data mines, data warehouses, and audits abound. The array of HR measurement technologies is daunting. The paradox is that even when HR measurement systems are well implemented, organizations typically hit a “wall.” Despite ever more comprehensive databases, and ever more sophisticated HR data analysis and reporting, HR measures only rarely drive true strategic change.^[3]

Figure 1-2 shows how, over time, the HR profession has become more elegant and sophisticated, yet the trend line doesn’t seem to be leading to the desired result. Victory is typically declared when business leaders are induced or held accountable for HR measures. HR organizations often point proudly to the fact that bonuses for top leaders depend in part on the results of an HR “scorecard.” For example, incentive systems might make bonuses for business-unit managers contingent on reducing turnover, raising average engagement scores, or placing their employees into the required distribution of 70 percent in the middle, 10 percent at the bottom, and 20 percent in the top.

Reprinted by permission of Harvard Business School Press, from Beyond HR: The New Science of Human Capital by John Boudreau and Peter M. Ramstad. Boston, MA, 2007, pp. 189.

Figure 1-2. Hitting the “Wall” in HR Measurement.

Yet, having business-leader incentives based on HR measures is not the same as creating organization change. To have impact, HR measures must create a true strategic difference in the organization. Many organizations are frustrated because they seem to be doing all the measurement things “right,” but there is a large gap between the expectations for the measurement systems and their true effects. HR measurement systems have much to learn from measurement systems in more mature professions such as finance and marketing. In these professions, measures are only one part of the system for creating organizational change through better decisions.

Many HR measures originate from a desire to “justify” the investments in HR processes or programs. Typically, HR seeks to develop measures to increase the respect for (and potentially the investment in) the HR function and its services and activities. Contrast this with financial measurement. Although it is certainly important to measure how the accounting or finance department operates, the majority of financial measures are not concerned with how finance and accounting services are delivered. Financial measures typically focus on the outcomes—the quality of decisions that affect financial resources. In contrast, most HR measures today focus on how the HR function is using and deploying its resources, and whether those resources are used efficiently. To the extent that the HR organization is ultimately accountable for improving talent decisions throughout the organization, HR professionals require a more holistic perspective on how measurements can drive strategic change.

Correcting these limitations requires keeping in mind the basic principle expressed at the beginning of this chapter: Human capital metrics are valuable to the extent that they improve decisions about talent and how it is organized. That means that we must embed HR measures within a complete framework for creating organizational change through enhanced decisions. We describe that framework next.

Without a proper logic, it is impossible to know where to look for insights. The logic element of any measurement system provides the “story” behind the connections between the numbers and the effects and outcomes. The chapters in this book provide logical models that help to organize the measurements, and show how they can articulate useful decision frameworks. Examples include the elements of turnover costs, the conditions that determine the value of enhanced selection, and the connections that link employee health and vital organizational outcomes. Missing or faulty logic is often the reason why well-meaning HR professionals generate measurement systems that are technically sound, but make little sense to those who must use them. With well-grounded logic, it is much easier to help leaders outside the HR profession understand and use the measurement systems to enhance their decisions.

For example, recall Figure 1-1, which shows how finance organizes its measures of return on equity to reflect the logic that equity is used to purchase assets, which are used to generate sales, which, in turn, produce profits. The logically derived measures include leverage (assets divided by equity), asset productivity (sales divided by assets), and margin (profits divided by sales). You can directly calculate return on equity simply by dividing profits by equity, but that would obscure the logical connection points that are vital to make decisions about equity, assets, and sales effectively. The power of the framework is to embed the measures within a logic that enhances decisions.

In the field of human resources, there are many logical frameworks, including salary structures, workforce-planning models, and even labor contracts. All are useful, but they are not sufficient to connect decisions about investments in HR programs to strategic outcomes. In contrast, some authors have proposed a “service-value-profit” framework for the customer-facing process. This framework calls attention to the connections between HR and management practices, which, in turn, affect employee attitudes, engagement, and turnover; which, in turn, affect the experiences of customers. This, in turn, affects customer-buying behavior, which, in turn, affects sales, which, in turn, affects profits. Perhaps the most well-known application of this framework was at Sears, which showed quantitative relationships among these factors and used them to change the behavior of store managers.^[5]

The measures part of the LAMP model has received the greatest attention in HR. As discussed in subsequent chapters, virtually every area of HR has many different measures. Much time and attention is paid to enhancing the quality of HR measures, based on criteria such as timeliness, completeness, reliability, and consistency. These are certainly important standards, but lacking a context, they can be pursued well beyond their optimum levels or they can be applied to areas where they have little consequence.

Consider the measurement of employee turnover. There is much debate about the appropriate formulas to use in estimating turnover and its costs, or the precision and frequency with which employee turnover should be calculated. Today’s turnover-reporting systems can calculate turnover rates for virtually any employee group and business unit. Armed with such systems, managers “slice and dice” the data in a wide variety of ways (ethnicity, skills, performance, and so on), each manager pursuing his or her own pet theory about turnover and why it matters. Are those theories any good? If not, better measures won’t help. That’s why the logic element of the LAMP model must support good measurement.

Precision alone is not a panacea. There are many ways to make HR measures more reliable and precise. An exclusive focus on measurement quality can produce a brighter light shining where the keys are not! Measures require investment, which should be directed where it has the greatest return, not just where improvement is most feasible. Organizations routinely pay greater attention to some elements of their materials inventory more than others. Indeed, a well-known principle is the “80-20 rule” that suggests that 80 percent of the important variation in inventory costs or quality is often driven by 20 percent of the inventory items. Thus, although organizations indeed track 100 percent of their inventory items, they measure the vital 20 percent with greater precision, more frequently, and with greater accountability for key decision makers.

Why not approach HR measurement in the same way? Employee turnover is not equally important everywhere. Where turnover costs are very high, or where turnover represents a significant risk to the revenues or critical resources of the organization (such as when departing employees take clients with them or when they possess unique knowledge that cannot be re-created easily), it makes sense to track turnover very closely and with greater precision. However, this does not mean simply reporting turnover rates more frequently. It means that the turnover measurements in these situations should focus precisely on what matters. If turnover is a risk due to the loss of key capabilities, turnover rates should be stratified to distinguish those with such skills from others. If turnover is a risk due to losses of clients with departing employees, turnover rates should not focus on skill differences, but instead should be stratified according to the risks of client loss.

Lacking a common logic about how turnover affects business or strategic success, well-meaning managers draw conclusions that might be misguided or dangerous. This is why every chapter of this book describes measures, as well as the logic that helps explain how the measures work together. For example, Chapter 4, “The High Cost of Employee Separations,” deals with turnover.

Even a very rigorous logic with good measures can flounder if the analysis is incorrect. For example, some theories suggest that employees with positive attitudes convey those attitudes to customers who, in turn, have more positive experiences and purchase more. Suppose an organization has data showing that customer attitudes and purchases are higher in locations with better employee attitudes? Does that mean that improving employee attitudes will improve customer attitudes? Many organizations have invested significant resources in programs to improve frontline-employee attitudes based precisely on this sort of evidence of association (correlation).

The problem is that this conclusion may be wrong, and such investments misguided. A correlation between employee and customer attitudes does not prove that one causes the other, nor does it prove that improving one will improve the other. Such a correlation also happens when customer attitudes actually cause employee attitudes. This can happen because stores with more loyal and committed customers are more pleasant places to work. The correlation can also result from a third, unmeasured factor. Perhaps stores in certain locations attract customers who buy more merchandise or services and are more enthusiastic. Employees in those locations like working with such customers, and are more satisfied. Store location turns out to cause both store performance and employee satisfaction. The point is that a high correlation between employee attitudes and customer purchases could be due to any or all of these effects. Sound analytics can reveal which way the causal arrow actually is pointing.

Analytics is about drawing the right conclusions from data. It includes statistics and research design, and then goes beyond them to include skill in identifying and articulating key issues, gathering and using appropriate data within and outside the HR function, setting the appropriate balance between statistical rigor and practical relevance, and building analytical competencies throughout the organization. Analytics transforms HR logic and measures into rigorous, relevant insights.

Analytics often connect the logical framework to the “science” related to talent and organization, which is an important element of a mature decision science. Frequently, the most appropriate and advanced analytics are found in scientific studies that are published in professional journals. In this book, we draw upon that scientific knowledge to build the analytical frameworks in each chapter.

Analytical principles span virtually every area of HR measurement. In Chapter 2, we describe general analytical principles that form the foundation of good measurement. We also provide a set of economic concepts that form the analytical basis for asking the right questions to connect organizational phenomena such as employee turnover and employee quality to business outcomes. In addition to these general frameworks, each chapter contains analytics relevant specifically to the topic of that chapter.

Advanced analytics are often the domain of specialists in statistics, psychology, economics, and other disciplines. In fact, HR organizations often draw upon experts in these fields, and upon internal analytical groups in areas such as marketing and consumer research, to help augment their own analytical capability. Although this can be very useful, it is our strong belief that familiarity with analytical principles is increasingly essential for all HR professionals and for those who aspire to use HR data well.

The final element of the LAMP framework is process. Measurement affects decisions and behaviors, and those occur within a complex web of social structures, knowledge frameworks, and organizational cultural norms. Therefore, effective measurement systems must fit within a change-management process that reflects principles of learning and knowledge transfer. HR measures and the logic that supports them are part of an influence process.

The initial step in effective measurement is to get managers to accept that HR analysis is possible and informative. The way to make that happen is not necessarily to present the most sophisticated analysis. The best approach may be to present relatively simple measures and analyses that match the mental models that managers already use. Calculating turnover costs can reveal millions of dollars that can be saved with turnover reductions, as discussed in Chapter 4. Several leaders outside of HR have told us that a turnover-cost analysis was their first realization that talent and organization decisions had tangible effects on the economic and accounting processes they were familiar with.

Of course, measuring only the cost of turnover is insufficient for good decision making. For example, overzealous attempts to cut turnover costs can compromise candidate quality in ways that far outweigh the cost savings. Managers can reduce the number of candidates who must be interviewed by lowering their selection standards. The lower the standards, the more candidates will “pass” the interview, and the fewer interviews that must be conducted to fill a certain number of vacancies. Of course, lowering standards can create problems that far outweigh the cost savings from doing fewer interviews! Still, the process element of the LAMP framework reminds us that often best way to start a change process may be first to assess turnover costs, to create initial awareness that the same analytical logic used for financial, technological, and marketing investments can apply to human resources. Then the door is open to more sophisticated analyses beyond the costs.

Education is also a core element of any change process. The return-on-investment (ROI) formula from finance is actually a potent tool for educating leaders in the key components of financial decisions. In the same way, we believe that HR measurements increasingly will be used to educate constituents and will become embedded within the organization’s learning and knowledge frameworks.

In the chapters that follow, we suggest where the HR measures we describe can be connected to existing organizational frameworks and systems that offer the greatest opportunity for using measures to get attention and enhance decisions. For example, the accounting and finance systems in organizations currently pay a great deal of attention to escalating health-care costs. The cost measures discussed in Chapter 5, “Employee Health, Wellness, and Welfare,” can offer additional insights and more precision to such discussions. Moreover, starting by embedding these basic ideas and measures into the existing health-care-cost discussion, HR leaders can gain credibility to be able to extend the discussion to include additional logical connections between employee health and other organizational outcomes, such as learning, performance, and profits. What began as a budget exercise becomes a more nuanced discussion about the optimal investments in employee health, and how those investments pay off.

You will see the LAMP framework emerge in many of the chapters in this book, to help you organize not only the measures, but also your approach to making those measures matter. Our next section illustrates how some alternative measurement frameworks can help us understand the benefits and limitations of several of today’s most popular approaches to HR measurement.

The first row of Table 1-1 describes measures focused on “efficiency” (see also Figure 1-1). These measures are usually expressed in terms of “input-output” ratios, such as the time to fill vacancies, turnover rates, turnover costs, and compensation budgets compared to total expenses.^[7] These approaches are compelling because they connect HR processes to accounting outcomes (dollars), and because they can show that HR operations achieve visible cost reductions, particularly when compared to other organizations. They are frequently a significant motivator for HR outsourcing. Many applications of Six Sigma to HR tend to focus on such measures to detect opportunities to improve costs or speed. One of the major limitations of these types of measures, however, is that they are not really HR measures at all—instead, they are efficiency ratios that can be used to monitor overhead costs in nearly any staff function. As a result, efficiency-focused systems can omit the value of talent. Fixating on cost reduction alone can lead to the rejection of more expensive decision options that are the better value. Efficiency-based measures alone, no matter how “financially” compelling, cannot reflect the value of talent. Finally, they focus almost exclusively on the HR function, and not on the decisions made elsewhere within the organization.

The next row of Table 1-1, “HR activity and ‘best-practice’ indexes,” directly measures the association between the reported existence of HR activities, such as merit pay, teams, valid selection, training, and so on, and changes in financial outcomes, such as profits and shareholder-value creation.^[8] Some results show strikingly strong associations between certain HR activities and financial outcomes, which has been used to justify investments in those activities. However, most existing research cannot prove that investing in HR activities causes superior financial outcomes.^[9] Another limitation of such measures is that they use one description of HR practices to represent an entire organization, when in reality HR practices vary significantly across divisions, geographic locations, and so forth. This may partly explain why managers in the same organization might inconsistently report the frequency of use of human resource management (HRM) activities.^[10] Also, such systems typically only measure the existence of HRM activities or practices, but not their effects. Even when an actual relationship exists, simply duplicating others’ best practices may fail to differentiate the organization’s competitive position. The best the organization can hope to achieve is to become a perfect copy of someone else.

These limitations can be seen by an analogy to advertising. It is quite likely that studies would show an association between financial performance and the presence of television-advertising activity, perhaps even that advertising activity rises before financial outcomes rise. This would suggest that among organizations that compete where advertising matters, advertising decisions relate to financial outcomes. Would it also mean that every organization should advertise on television? Obviously not.

Thus, these approaches shed some valuable light on the important question of whether HR activities relate to financial outcomes, and they have made important contributions to HRM research. However, even their strongest advocates agree that they do not measure the connections that explain why HRM practices might associate with financial outcomes, and they do not reflect other key elements of strategic success. They leave unanswered whether and how groups of employees significantly affect key processes and outcomes.

The third row of Table 1-1 describes HR “scorecards” or “dashboards,” inspired by Kaplan and Norton,^[11] who proposed adding measures of “customer” (such as customer satisfaction, market share, and so on), “internal processes” (such as cycle time, quality, and cost), and “learning and growth” (systems, organization procedures, and people that contribute to competitive advantage) to traditional financial measures. HR scorecards include measures aligned and arranged into each of the four perspectives.^[12] Such approaches tie HR measures to a compelling business concept and, in principle, can articulate links between HR measures and strategic or financial outcomes.

Today’s scorecards or “dashboards,” built on data warehouses, allow users to “drill down” using a potentially huge array of variables customized to unique personal preferences. For example, HR training costs conceivably can be broken down by location, course, and diversity category, and then linked to attitudes, performance, and turnover. Although impressive, in the hands of the unsophisticated, such approaches risk creating information overload, or even worse, a false certainty about the connection between talent and strategic success. As Walker and MacDonald observed in describing the GTE/Verizon scorecard, “The measures taken in isolation can be misleading.” They describe one GTE/Verizon call center where, “when HR reviewed the call center results from the HR Scorecard...the HR metrics showed a very low cost per hire, a very quick cycle time to fill jobs, and an average employee separation rate ... the staffing metrics showed a high efficiency and cost control.” However, the call center accomplished this by “changing talent pools and reducing the investments in selection methods [that] kept costs low while bringing in applicants who were ready to start quickly but were harder to train and keep ...a bad tradeoff.” GTE/Verizon was fortunate to have HR analysts who discovered this flaw in logic, but the example shows that even the best scorecards and drill-down technology alone do not necessarily provide the logical framework users need to make the best talent decisions.

HR scorecards are also often limited by relegating HR to measuring only the “learning and growth” category, or by applying the four categories only to the HR function, calculating HR-function “financials” (for example, HR program budgets), “customers” (for example, HR client-satisfaction surveys), “operational efficiency” (for example, the yield rates of recruitment sources), and “learning and growth” (for example, the qualifications of HR professionals). Both lead to measurement systems with weak (if any) links to organizational outcomes.

When we work with scorecard designers, they note that the majority of scorecards measure only HR operations and activities, the elements of efficiency and effectiveness in Figure 1-1. Scorecards admirably draw attention to impact, but the actual link between logic and measurement is often superficial, such as linking the organizational goal of “speed to customers” with the HR scorecard measure “faster time to fill,” or linking the strategic goal of “global integration” with the HR scorecard measure of “number of cross-region assignments completed.” Still, the scorecard-design principle of connectedness has promise, as we shall see in Chapters 3 through 11.

The bottom row of Table 1-1 describes causal-chain analysis, which focuses on measuring the specific links between HRM programs or individual characteristics and business processes or outcomes. Recall our earlier example, where Sears, a large U.S. retailer, used data to connect the attitudes of store associates, their on-the-job behaviors, the responses of store customers, and the revenue performance of the stores. This measurement approach offers tangible data and frameworks that actually measure the intervening links between human capacity (in this case, store-associate attitudes reflecting their commitment or motivation) and business outcomes (such as store revenues). In terms of Table 1-1, causal-chain analysis comes closest to mapping all the linking elements.

The drawback is that all causal chains simplify reality. At the same time, they are so compelling that they might motivate oversimplification. Finding that employee attitudes predict customer responses, organizations may invest heavily to maximize employee attitudes. At some point, other factors (such as employee knowledge of products) become more important. Continuing to raise attitudes can actually be suboptimal, even if it produces small additional changes in business outcomes. It’s important to have a logical framework that can reveal the new paths as they emerge.