9

_______________

Talent Measurement and Analytics

Beyond Measures to the LAMP

How many times have you heard that the reason HR doesn’t get the respect it deserves is because it’s soft and doesn’t have the measures that accounting, marketing, and other areas have? It is very common for experts to assert that decisions about talent and organization would be dramatically improved if only the HR profession would develop more or better “numbers,” usually designed to provide objective evidence that HR investments pay off. This chapter applies the principles of talentship to measurement and illustrates that many popular notions about HR measurement are simply wrong, beginning with the idea that more measures equal better decisions.

A great deal of the earliest work attempting to connect HR activities to organizational outcomes was motivated by the goal of developing measures to calculate the costs and benefits of specific HR programs, such as staffing, training, and pay.¹ In the 1980s and 1990s our own experience was similar, in that we approached the task of enhancing talent and organization decisions by developing and understanding HR measurement.

We were frequently asked to help HR leaders develop more and better HR measures, because HR was seen as soft and the key to getting support for HR investments was more measures that would be accepted by business leaders using accounting and finance models. The paradox was that in organization after organization, the problem wasn’t a lack of measures at all. In fact, these HR organizations had amassed hundreds of measures of their activities, some of which even quantified effectiveness outcomes such as learning, attitudes, and turnover. What we found was that HR measurement is useful, but a fixation on measurement can prove to be very dysfunctional.

Developing more measures wasn’t going to solve the fundamental issue. The real need was for decisions about talent and organization investments to become more systematic, consistent, and shared between HR and non-HR professionals. It was understandable that HR leaders would believe that the problem was that HR measures couldn’t compete with more well-developed measures from accounting and finance. Yet, as we have seen, the fundamental power of those decision sciences emanates not just from their measures but, more important, from the logic on which they are developed and presented.

As we discussed earlier, one hallmark of a well-developed decision science is when the measurement and data systems work synergistically with the decision framework and the alignment of processes and competencies inside and outside the particular function. The framework provides the logic that identifies what measures are needed and where measurement can most affect decisions that have a large impact on sustainable strategic success. The measures populate the logical framework, that framework evolves based on the findings from the measures, and so on. At the same time, the measurement framework provides the data for tracking processes and teaching competencies so they are aligned with the decision framework. Later in this chapter we’ll present a detailed example of precisely this kind of evolution applied to talent and organization measurement.

The HC BRidge model provides a useful logical framework for designing and refining measurement systems so that measurement development is focused where it can have the most impact. In this chapter we will show how you can use the HC BRidge framework to diagnose and better understand how to improve HR measurement. More important, we will suggest a perspective that recognizes HR measurement as a catalyst for organizational effectiveness through better talent and organization decisions where they matter most. That requires looking well beyond the measures, to uncover the elements of a more complete system of decision support and organizational change.

Hitting the Wall in HR Measurement

Type “HR measurement” into a search engine, and you will get over nine hundred thousand results. Scorecards, summits, dashboards, data mines, data warehouses, and audits abound. HR organizations lament that their measurement efforts are stymied by limited budgets, but even among those with significant resources (in fact, especially in these cases), 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.² Despite ever more comprehensive databases, and ever more sophisticated data analysis and reporting, HR measures only rarely drive true strategic change.³

As figure 9-1 shows, over time the HR profession has become more and 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 held accountable for HR measures. HR organizations often point proudly to the fact that top leaders’ bonuses depend in part on the results of their scorecard measures, such as turnover, employee attitudes, bench strength, and performance distributions. For example, some incentive systems make bonuses for business-unit managers contingent on reducing turnover to a target level, raising average engagement scores, or placing their employees into the required distribution of 70 percent in the middle, 10 percent in the bottom, and 20 percent in the top.

Yet having business leaders manage to such numbers is not the same as creating organizational change. HR measures must create a truly strategic difference in the organization. As we have seen, turnover reductions, increased engagement, and performance differences are not equally pivotal everywhere. Understanding the nuances is often the key to avoiding costly mistakes based on well-meaning but misguided talent goals.

FIGURE 9-1

Hitting the wall in HR measurement

Many of the organizations we work with are frustrated because they seem to be doing all the measurement things right yet see the gap between the expectations for the measurement systems and those systems’ true effects.

Why do HR organizations hit the wall? As we have seen, the HR profession is on the cusp of extending its paradigm from focusing only on compliance and services to including a specific focus on talent and organization decisions. Recognizing the implications of this paradigm extension provides clues to how HR measurement systems can learn valuable lessons from measurement systems in more mature professions like finance and marketing. In these professions measures are only one part of the system for creating organizational change through better decisions.

Beyond Proving the Value of the HR Function

Here is a question we commonly ask HR audiences: “Would you like to have HR measures as powerful as accounting measures?” The answer is invariably an enthusiastic yes. Then we ask the follow-up question: “How many of the accounting measures tell you about how the accounting department is doing?”

The implication is clear. Many HR measures originate from a desire to justify the investments in HR processes or programs. Typically, HR seeks measurement not to improve decisions but to increase the respect for (and potentially the investment in) the HR function and its services and activities. Talent measurement is often a search for validation of the HR function more than a quest for better talent and organization decisions.

In financial measurement it is certainly important to measure how the accounting or finance department operates. Measures—such as transaction-processing time, benchmark staff levels, and so on—are important for internal functional control. The vast majority of measures used for financial decisions, however, are not concerned with how finance and accounting services are delivered. Financial measures typically focus on the outcomes—the quality of decisions that impact financial resources.

One of the implications of the traditional HR measurement approach is that it often puts HR professionals in a bind. If the measurements show that HR is doing well, no one cares. If the measurements show that there is a problem, even when the problem is not caused by the HR function, it is still assigned to HR to fix. A classic example is turnover. If it is low, HR gets no credit; and if it is higher than it should be, it becomes an HR issue rather than a business issue. This stands in significant contrast with financial measures. When a division is behind budget, the accounting department is rarely responsible for fixing the problem. Rather, accounting and finance are tasked with providing the insight, measures, and frameworks that highlight the issue and provide the mental models for the business leader to craft an appropriate response.

Most HR measures today focus on how the function is using and deploying its resources. Satisfaction with the department is also measured. Some HR organizations actually measure satisfaction with different programs as a first step in deciding what the HR department should and should not do. We have proposed that the paradigm shift toward the talentship decision science requires HR to be accountable for improving talent decisions throughout the organization. That requires a framework for connecting those investments to organizational effectiveness, but it also requires taking a more holistic perspective on how measurements can drive strategic change. We describe that framework next.

The LAMP Framework

We believe that a paradigm shift toward a talent decision science is key to getting to the other side of the wall. Incremental improvements in traditional approaches are not adequate. HR measurement can move beyond the wall using what we call the “LAMP model.”⁴ The letters in LAMP stand for four critical components of a measurement system that drives strategic change and organizational effectiveness. As shown in figure 9-2, the letters stand for logic, analytics, measures, and process. Measures represent only one component of this system. Though they are essential, without the other three components, measures are destined to remain isolated from the true purpose of HR measurement systems.

LAMP is more than an acronym; it’s also a metaphor for today’s HR measurement dilemma. Consider this illustration: one evening while strolling, a man encountered an inebriated person diligently searching the sidewalk below a street lamp.

“Did you lose something?” he asked.

“My car keys. I’ve been looking for them for an hour,” the person replied.

The man quickly scanned the area, spotting nothing. “Are you sure you lost them here?”

“No, I lost them in that dark alley over there.”

“If you lost your keys in the dark alley, why don’t you search over there?”

“Because this is where the light is.”

FIGURE 9-2

Lighting the LAMP

In many ways, talent and organization measurement systems are like the person looking for his keys where the light is, not where he is most likely to find them. This has been accelerated by advancements in information technology that often provide technical capabilities that far surpass the ability of the decision science and processes to properly use them. So it is not uncommon to find organizations that have invested significant resources constructing elegant search and interactive presentation technology around measures of efficiency or measures that largely emanate from the accounting system.

The paradox is that the real insights probably exist in areas where there are not standard accounting measures. Significant growth in HR outsourcing, where efficiency is often the primary value proposition and IT technology is the primary tool, has exacerbated these issues. Even imperfect measures aimed at the right areas may be more illuminating than very elegant measures aimed at the wrong place. As it’s been said, “Even a weak penlight in the alley where the keys were lost is better than a very bright streetlight directed somewhere else.”

As figure 9-2 shows, HR measurement will advance most quickly if it focuses on the ultimate objective of measurement in a decision science framework. Ultimately, measurement systems are only as valuable as the decisions they improve and the organizational effectiveness to which they contribute. Measures must enhance talent and organization decisions where they most affect strategic success and organizational effectiveness. Let’s examine how the four components of the LAMP framework define a more complete measurement system. In doing so, we will present the elements in the following order: logic, measures, analytics, and finally, process.

Logic: Implementing the HC BRidge Decision Framework

Throughout this book we have illustrated the significant power of a logical framework that connects talent and organization investments to strategic success. Such a framework provides a language for systematic and consistently in-depth conversations about how to improve the way organizations compete with and for talent and how organizations are designed. Once the logic is clear, measures emerge that were not obvious before. Recall two of our examples. Boeing measured the behavioral performance of its engineers differently when it realized that the pivotal aligned actions included facilitating global teams. Starbucks measured engagement of its baristas differently from fast-food outlets that rely on standardization because of the strategic importance of employee trust and discretion. The point is that measurement systems, like so much of HR and talent management, are most powerful when logic precedes measurement and when measures are closely tied to the logical pivot-points that make the biggest difference to strategic success and the organization’s unique position in its talent markets.

Measures: Counting What Counts

As noted earlier, the measures part of the LAMP model has received the greatest attention in HR. Lists of HR measures abound, often categorized into scorecards and dashboards. 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 optimal level or applied to areas where they have little consequence.

Measuring turnover offers a good example. HR organizations have spent countless hours debating the appropriate formula for turnover and the precision and frequency with which it can be calculated. One HR data warehouse team we worked with said, “We have built the most sophisticated turnover-tracking data and Web interface ever. Now we’ll put it out there and see what our managers do with it. They are strategic leaders, so they will help us understand how to analyze turnover data.”⁵ What happened is that managers began to slice and dice the data in a wide variety of ways, each pursuing his or her own pet theory about turnover and why it mattered. Some generated reports on turnover by ethnicity, others based on skill levels, others based on performance, and so on. Having no common logic about the role of employee turnover in affecting business or strategic success, well-meaning managers were drawing conclusions that might be misguided or dangerous.

As we’ve seen in earlier chapters, the implications of turnover rates or any other HR measure are very different depending on strategic and business context. Where talent is quality-pivotal, because applicants are well qualified and quickly master the job, turnover is important because of its costs and its effect on talent shortages. Turnover affects the organization mostly through the lack of a full complement of employees. Thus, filling vacancies more quickly addresses the business issue.

A completely different situation is where turnover creates a capability shortage in a position that is quality pivotal, such as where it takes time to learn the job and experienced individuals are being replaced by inexperienced ones. Reducing turnover or filling vacancies more quickly may not address the problem. Turnover can be reduced without increasing overall workforce experience if the number of departures among experienced employees rises and the number of departures of inexperienced employees falls by an equal or greater number. Here, the effects of turnover on workforce learning and quality is key, so reducing turnover levels and time to fill may be much less important than keeping experienced employees or speeding the learning among new employees.

Finally, as important as turnover is, we only rarely see organizations measure the destinations of the employees who leave. For many roles, this is actually the greatest economic impact of turnover. The most common distinctions are between voluntary and involuntary turnover and the reason someone leaves. However, knowing where people are going is often critical. For example, it is far different if someone voluntarily leaves to work for a competitor (perhaps taking valuable knowledge—or even key clients—with them) versus leaving the industry entirely.

Precision alone is not a panacea. There are many ways to make HR measures more reliable and precise, but an exclusive focus on measurement quality can produce little more than a brighter light shining in a place other than where the keys are! Measurement quality must be considered in the context of decision support. Improved measures require investment, which should be directed where it has the greatest return, not simply where improvement is most feasible.

Diagnosing Measurement Systems

The logical elements of the HC BRidge framework—efficiency, effectiveness, and impact—also provide a template for building measurement systems. Organizations can use the HC BRidge framework to determine whether their measures are properly representing the three anchor points. Conference Board research suggests that the vast majority of HR measures fall in the efficiency anchor point.⁶ The organization measures the resources spent on HR programs, the frequency or existence of HR programs, or in some cases the demographic characteristics of the workforce, as table 9-1 shows.⁷ Turnover and resignation rates are among the most common measures.

Recent research at the Center for Effective Organizations shows that having measures in all three areas—efficiency, effectiveness, and impact—is correlated with the degree to which HR leaders play a significant role in strategy formation. Table 9-2 shows that there is a significant correlation between the existence of HR measures in every category and the degree to which HR leaders perceive a stronger role in strategy.⁸

The data in table 9-2 shows measures ordered with impact at the top, effectiveness in the middle, and efficiency at the bottom. The averages show that organizations in the survey more often reported that efficiency measures existed now (numbers closer to four, on the four-point scale), while effectiveness and impact measures were more likely “being considered” (closer to one on the scale). However, when the existence of each category of measures was compared to the responses on a question of HR’s role in strategy, for almost every measure there was a relationship between its existence and HR’s stronger strategic role. Measures throughout the decision framework are needed; no one area is more related to strategic partnership than others. So it is important for HR organizations to consider carefully how well their measurement systems map the elements of the HC BRidge framework.

TABLE 9-1

Highest-frequency human capital measures

Turnover	96%
Voluntary resignation	84
Average compensation	82
Average workforce age	77
Diversity	76
Compensation/total cost	76
Average seniority	75
Work accident frequency	74
Percentage with variable compensation	71
Percentage with stock options	71
Source: Stephen Gates, Measuring More Than Efficiency, Research report r-1356-04-rr (New York: Conference Board, 2004).

TABLE 9-2

Measuring the anchor points and strategic partnership

Anchor point	Does your organization currently …	Avg.	Correlation with HR role in strategy
Impact	Collect metrics that measure the business impact of HR’s programs and processes?	2.7	.20*
Effectiveness	Use dashboards or scorecards to evaluate HR’s performance?	2.9	.31**
Effectiveness	Use measures and analytics to evaluate and track the performance of outsourced HR activities?	2.7	.30**
Effectiveness	Have metrics and analytics that reflect the effects of HR programs on the workforce (i.e., competence, motivation, attitudes, behaviors, etc.)?	2.7	.29**
Effectiveness	Have the capability to conduct cost-benefit analyses (also called “utility analyses”) of HR programs?	2.5	.19
Efficiency	Measure the financial efficiency of HR operations (e.g., cost per hire, time to fill, training costs)?	3.1	.29**
Efficiency	Collect metrics that measure the cost of providing HR services?	3.0	.24*
Efficiency	Benchmark analytics and measures against data from outside organizations (e.g., Saratoga, Mercer, Hewitt, etc.)?	3.0	.11
Response scale is 1 = “Not currently being considered” to 4 = “Yes, have now.” *p ≤.05 **p ≤.01
Source: Edward E. Lawler III, John W. Boudreau, and Susan Mohrman, Achieving Strategic Excellence (Palo Alto, CA: Stanford University Press, 2006).

In our work with organizations, we have found that using the anchor points as a diagnostic framework moves attention from simply listing measures or organizing them using standard scorecard categories to considering how each measurement element connects to others to tell the story about the logical connections embodied in the framework. We ask HR leaders where most of their measures lie. They usually conclude that the preponderance of their measures, and thus the focus of their key constituents, is in the efficiency part of the framework.

They also realize that there are many measures that already exist in other management systems that could be usefully incorporated into their talent and organization measurement approach, to reflect effectiveness and impact. For example, many measures exist for the vital processes and resources from the impact part of the HC BRidge framework, but they are the purview of other functions, such as the supply chain, information systems, manufacturing, and R&D.

When leaders inside and outside the HR function begin to have in-depth conversations about the constraints and vital pivot-points that we illustrated earlier, they usually discover that process owners can connect their key process measures with the more typical HR measures in the efficiency part of the HC BRidge framework. We’ll see the power of this connection in the example from Limited Brands later in the chapter.

Finding Talent and Organization Measures in Organizational Databases

The anchor points of the HC BRidge framework also provide a useful lens for understanding the connections between the structure of organizational databases and the talentship decision science. As we have seen, it is useful to consider how to populate the HC BRidge framework with measurements. It is also important to understand the structure of organizational databases and where those measures are likely to reside. Data warehouse technology can integrate data from different systems and processes. The HC BRidge framework can provide an organizing structure that applies such databases to talent and organization decisions.

The framework also provides a useful perspective on the typical maturity curve of measurement. HR and organizations must realize that connecting measures to a decision science takes time and that what’s feasible will depend on the existing measures and data. Few organizational measurement systems were designed to reflect a talent and organization decision science. They are much more likely to reflect financial, operations, marketing, or other perspectives. Thus, the measures available to support the HR decision science will vary with the system’s maturity. Figure 9-3 captures this relationship.

FIGURE 9-3

Metrics’ maturity over time

On the left side of figure 9-3 are the familiar anchor points of the HC BRidge framework. Here the framework is matched to the typical categories, or layers, of data as defined by most organizational data systems. The idea is to provide a map that organizational leaders can use to integrate the HC BRidge elements with the structure of the data they encounter in their existing information systems. Within each layer there is also a progression of data sophistication or maturity, which the diagram shows as moving from left to right within each layer.

Organizational Performance Layer

At the top is the organizational performance layer. Data in this layer reflects standard organizational performance measures, such as assets, cash, sales, and overall cost levels. Basic data systems typically contain information for financial reporting at the enterprise, business-unit, and functional levels. Such measures are important and useful, particularly for external financial reporting and for most management reporting. As we have noted, however, the data necessary to uncover the strategic pivotpoints is often deeper, embedded in measures of organization and business processes and resources, such as manufacturing, supply chain, R&D, sales, and customer relations. This sort of data often exists in the managerial accounting system that tracks these processes and resources, but it is not always readily available in the organizational data systems until they reach a later stage of maturity, as the diagram shows.

Human Capital Layer

The next row shows the human capital layer. The data in this layer is generally located in the HR information system. In the most basic systems, the data in this layer tends to reflect information necessary for government and financial reporting, largely focused on demographics and head-count in different jobs or organizational units. Closely related to data on demographics and headcount is data on employee movement into, out of, and between organizational positions.

As HR information systems mature, they typically incorporate the data that is collected through employee surveys, including attitudes. Such surveys eventually tap employee engagement and perceptions of the line of sight from their work to the larger organizational mission. Next, data systems will incorporate the data from the performance measurement system, including performance ratings and perhaps information on specific performance behaviors. Once it is possible to gather data from the performance management system, it is also possible to gather data on employee potential and readiness, particularly when those ratings are done through the same process as performance ratings (such as the common two-dimensional format with performance on one axis and potential on the other). Finally, sophisticated systems will track individual-level behaviors directly related to business outcomes. For example, some of the more sophisticated systems track specific sales and patents for particular individuals.

Connecting the organization performance layer and the human capital layer supports the impact element of HC BRidge. An example would be to correlate observed behaviors (such as performance scales based on behavioral anchors like “sharing information” or “completing customer relations paperwork on time”) with business outcomes (such as individual sales revenue or customer satisfaction scores). The relationship provides one way to measure the performance yield curves described earlier.

HR Systems Layer

The third data layer is the HR systems layer, which is also often found in the HR information system. This is focused on data about the performance and activity level of HR processes rather than measures connected to individuals. Basic data at this layer includes HR activities such as training, staffing, development, and rewards. It reflects such elements as the number of process transactions (requisitions filled, performance ratings completed, training days delivered, etc.). As data systems mature, we encourage collecting data on the necessary conditions for system success that we described in earlier chapters.

Refocusing measurement on these conditions has a powerful effect on the decision processes and accountability relationship between HR and its clients. For example, early learning management systems might measure the number and types of training provided or attended. A more advanced system might develop measures of the conditions required for learning to occur and be used. These conditions might include readiness for the learning experience (such as being motivated and understanding how the learning will relate to one’s work); learning (the actual level of knowledge or skill attained); and transfer (the degree to which opportunities are provided to apply the learning after the learning experience). As we noted earlier, the readiness and transfer conditions are largely affected by the immediate supervisor of the learner and less by the learning experience itself.

To see how the measurement system supports elements of talentship, recall our earlier example in which employee performance is low due to a lack of readiness or transfer. The typical measurement system reflects only the training provided. So, when the employees’ manager complains that the training isn’t effective because the trained employees are not performing better, the HR measurement system shows that those employees were appropriately trained.

Now consider what happens if the system contains data on the relative levels of readiness, learning, and transfer for different managers. Such data might reveal that this manager’s employees are learning as much as those of other managers but that they are significantly lower on readiness and/or transfer. While the HR function has a great deal of influence on learning, the manager has the greatest influence on readiness and transfer. We believe that more advanced HR systems measures will increasingly distinguish between conditions that are primarily under the control of leaders outside of HR (such as preparing participants to learn and providing opportunities to apply the learning) versus those that are under the control of the HR function (such as the quality of a training program).

Many of the conditions that most significantly affect the success of HR systems are more controlled by leaders outside rather than inside the HR function. Just as accounting shows which units are performing above or below budget, more advanced measures at the HR systems layer will show the relative performance of units on important conditions defined by a decision-based approach to those systems.

Connecting the HR systems layer to the human capital layer often provides data to inform the effectiveness anchor point of the HC BRidge framework.

HR Investments Layer

The final data layer is the HR investments layer. Such systems virtually always begin with a focus on the resources used in the HR function, including the accounting budget and HR headcount. As the systems mature, they expand to include the basic HR deliverables, such as the number of programs, the number of employees using them, their frequency, and the time and money expended on specific HR initiatives.

Connecting the data in the HR investments layer with the data in the HR systems layer supports the efficiency anchor point of the HC BRidge framework.

A full treatment of the measurement implications of the HC BRidge framework is beyond this book’s scope. Our purpose here is to illustrate how the HC BRidge framework provides an alternative to scorecards or other systems that focus on only one part of the logic connecting talent to strategic success. Using the four layers, organizations can begin discussions about how they would actually measure the HC BRidge linking elements in a way that uniquely captures the vital connections between talent, organization, and strategic success.

Analytics: Finding Answers in the Data

Even a very rigorous logic with good measures can flounder if the analysis is done incorrectly. For example, it is logical to suggest that when employee attitudes are positive, employees convey those attitudes to customers, who in turn have more positive experiences and purchase more. Many organizations test that logical premise by correlating employee attitudes with customer attitudes across different retail locations. If customer attitudes and purchases are higher in locations with higher employee attitudes, that is interpreted to mean that improving employee attitudes will improve customer attitudes. Many organizations have invested significant resources in programs to improve frontline employees’ attitudes based precisely on this sort of correlation evidence.

Of course, this conclusion may be wrong, and such investments may be misguided. A simple correlation between employee and customer attitudes does not prove that one causes the other, nor that improving one will increase the other. For example, a high correlation between employee and customer attitudes can occur because stores that are in locations with more loyal and committed customers are a more pleasant place to work. Customer attitudes can actually cause employee attitudes. Or the relationship could be due to a third factor: location. Perhaps stores in better locations attract customers who buy more and who are more enthusiastic about new offerings. Employees in those locations may like working with such customers better and be more satisfied. Store location turns out to cause both store performance and employee satisfaction.

How Analytics Supports Better Decisions

Analytics builds on the science of determining the right conclusions from data. It draws on statistics and research design and then goes on to include identifying and articulating key issues, gathering and using appropriate data inside 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 data and measures into rigorous and relevant insights.

The more abundant data becomes, the more essential is analytical capability. Without sufficient analytical capability, HR and business leaders can fall victim to improper conclusions or be misled by superficial patterns and make poor human capital decisions. Analytics ensures that insights from HR data provide legitimate and reliable foundations for human capital decisions. Thus, analytics is an essential addition to rigorous logic. Analytics often provides a prominent way to connect the decision framework to the scientific findings related to talent and organization resources and decisions, which we noted earlier as an important element of a mature decision science. Frequently, the most appropriate and advanced analytics are found in scientific studies.

Finding the Talent Analytics in Organizations

Increasingly, organizations are devoting specific resources to improving analytics applied to talent and organization decisions. Analytical methods have long been a standard part of training social scientists in areas such as psychology, sociology, and economics. Many HR organizations already employ research teams. Such teams often comprise social scientists with PhD-level training in designing and carrying out research.

Other organizations rely on analytical capability outside the HR function. For example, organizations with very strong capabilities in customer and market analysis often engage their analysts on HR issues. It is not unusual to find market researchers called in to look for patterns of employee attitudes and to identify employee types, just as they might identify customer segments. Engineers may be adept at data mining and identifying patterns in things as varied as oil deposits, customer demographics, and flows through the supply chain; they are sometimes asked to find useful patterns in data on employee demographics, movement patterns between jobs, turnover, or attitudes. And some HR organizations call on the analytical capabilities of a wide variety of commercial vendors or universities.

HR analytics teams are also often called on as subject matter experts to support other HR professionals and are asked to educate their peers to help raise the level of analytical awareness in the HR function. For example, Sun Microsystems created an R&D laboratory for HR, and over time this laboratory evolved from a source of very specific research on the effects of HR programs, to a source of analytical expertise for others in HR, and finally to a source of forward-looking research on issues deemed to be critical to the strategic future of the organization, such as virtual work.⁹

Whether the analytical skills reside within the HR function, in other parts of the company, or with an outside organization, HR analytical teams today are typically focused on fairly narrow HR domains. It is not unusual for internal HR research groups to attend exclusively to attitude surveys, to compensation market data, or to mapping flows of employees through different roles and positions. These skills are increasingly valuable outside these rather specialized areas. Analytical skills are even appearing in competency models.¹⁰ The challenge is to create an HR measurement system and organizational structure that successfully engages these skills where they can have the greatest effect. As we have seen, the most interesting and important decisions in HR span the functional specialties of HR and often require understanding relationships between talent and organization elements such as resources, processes, and differentiators. Thus, future talent and organization analysts will increasingly integrate and build cross-organizational databases and design research that incorporates business, economic, and strategic contexts.

The talentship perspective allows us to envision a future in which talent and organization analytics are much more closely tied to mainstream analytics in areas such as marketing, finance, operations, and information systems. Today’s talent analytics are often separate from the more mature functional analyses and are often completed only after other analyses are finished. As we have seen throughout this book, there is no need for such a separation. Rather, traditional strategy and business analysis will be enhanced by incorporating insights from talent and organization and will be aimed at improving talent and organization decisions.

Process: Making the Insights Motivating and Actionable

The final element of the LAMP framework is process. In talentship the ultimate criterion for HR measurement is how it affects organizational effectiveness and sustainable strategic success. Measurement affects these outcomes through its impact on decisions and behaviors, and those decisions and behaviors occur within a complex web of social structures, knowledge frameworks, and organizational cultural norms. Thus, a key component of effective measurement systems is that they 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.

For example, research shows that if managers don’t perceive HR issues as strategic and analytical in the first place, they may simply ignore numerical and analytical information about HR.¹¹ They seem to place HR into a soft category of phenomena that are beyond analysis and therefore are only really addressable through opinions, politics, or other less analytical approaches.

So an initial step in effective measurement is to get managers to accept that HR analysis is possible and informative. The way to do that is often not to present the most sophisticated analysis right away. The best approach may be to present relatively simple measures that clearly connect to the mental frameworks that managers already use. In some organizations calculating the costs of turnover reveals that millions of dollars might be saved with turnover reductions. Organizational leaders 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. As we noted earlier, overzealous attempts to cut turnover costs can compromise candidate quality in ways that are far more significant to cost savings. Yet the best way to start a change process may be to first present turnover costs to create needed awareness that the same analytical logic used for financial, technological, and marketing investments can apply to HR.

We noted earlier that a significant element of the evolution of a talent and organization decision science will be a shift from creating influence by responding to client requests or telling constituents what is required, to enhancing HR’s influence through educating constituents about the principles and logic they can use to make better decisions.¹² Education is also a core element of change processes. The ROI formula from finance is actually a potent tool for educating leaders in the key components of financial decisions. In the same way, as the talentship decision science takes hold, HR measurements will educate constituents and become embedded within the organization’s learning and knowledge frameworks.

Let’s turn to a comprehensive example of the principles we’ve discussed, at a global retail organization: Limited Brands.

Limited Brands’ Store-Level Measurement Evolution

Limited Brands is a globally known retailer that operates a balanced portfolio of retail brands, including intimate apparel (Victoria’s Secret), general apparel (The Limited), and personal care (Bath & Body Works).¹³ A core process for Limited Brands is store operations, and the company uses a decision science approach to allocating scarce resources, such as real estate, technology, money, and talent.

Typical of many retailers, Limited Brands had sophisticated measurement systems and decision frameworks for many of its key resources in stores, but the models and measures used for talent and organization were rudimentary. In 2004 the organization set out to change this. Its experience vividly illustrates the power of adopting a decision science approach to talent and organization measurement, connecting measures with a shared and logical decision model and developing measures that deeply reflect the core processes and resources, not just the top-level outcomes.

From “Accounting for Payroll” to “Deploying Talent Resources”

Toyin Ogun, a VP of HR at Limited Brands, observes that the motivation for the talent and organization measurement makeover was the realization that decisions about talent and organization were founded largely on an accounting-based system of payroll tracking and allocation. Figure 9-4 shows the contrast between the existing system on the left and the aspiration on the right. Notice that both systems can be described as based on measurement, facts, and evidence. Indeed, in many organizations a scorecard that could track the allocation of payroll to stores would be regarded as a business-relevant measurement of human capital. Comparing the left to the right sides of figure 9-4 shows that although different measures would populate the new system on the right, the essence of the change was to literally reverse the measurement logic from top-down to bottom-up. Instead of starting with payroll and allocating it across stores based on activity, the objective was to begin with an in-depth decision model of optimal store activity and use that to make decisions about talent and organization, which would then determine the necessary resources, such as payroll, and their return.

FIGURE 9-4

Reversing the measurement logic at Limited Brands

Source: Toyin Ogun, “Limited Brands Talent Measurement” (Metrics and Analytics Executive Program, Center for Effective Organizations, University of Southern California, 2005). Reprinted with permission.

Making the Subjective More Objective

As Ogun observes in figure 9-5, much of the decision making about talent in the existing system was subjective and not based on data or a shared logic. Store and business leaders were working hard to make good decisions, but a lack of measures, analytics, logic, and a repeatable process all contributed to potentially less-than-optimal decisions. As with so many talent decisions, well-meaning business leaders were using their own logical algorithms, such as catering to stores that asked first, sending talent to stores with the fastest-growing sales, or providing more labor hours to stores that were open the most hours. Each has some logical justification, but none was based on a deep and logical connection between talent and business success.

Seeing Store Operations Through the Lens of Talent Decisions

Limited Brands set out to develop a decision science about the right roles, aligned actions, and capacities that were connected to vital store processes and resources. Figure 9-6 shows the logic the company developed. Notice how this is not an HR diagram at all. Rather, it is an actual overhead perspective of the floor of a retail space, used to depict the vital processes and potential talent contributions. For example, salespeople at the front of the store are scheduled according to each store’s unique traffic forecast (Point 1 in the diagram), while selling experts are deployed at the high-fashion locations during peak sales hours (Point 3). The pivotal contribution of store associates is different at the front of the store than at the expert-driven locations that involve high fashion. Freight processing activities are driven by a set of more specific labor standards based on the time it takes to unload trucks (Point 7).

FIGURE 9-5

Limited Brands’ traditional allocation of payroll within stores

Source: Toyin Ogun, “Limited Brands Talent Measurement” (Metrics and Analytics Executive Program, Center for Effective Organizations, University of Southern California, 2005). Reprinted with permission.

FIGURE 9-6

Limited Brands’ science of right people/place/time/role/skills

Source: Toyin Ogun, “Limited Brands Talent Measurement” (Metrics and Analytics Executive Program, Center for Effective Organizations, University of Southern California, 2005). Reprinted with permission.

The fact that this diagram looks like a store is important in subtle ways. The process element of the LAMP model emphasizes that the best talent and organization measurement systems will not only get the story correct and measure it well; they will embed it within the organization’s mental models so that it is an effective decision support system both inside and outside the HR function. We will return to this idea in chapter 10.

Notice how differently the measurement question is framed when it is depicted this way. It is no longer simply an issue of allocating payroll or even of implementing HR programs. The connection between decisions about store performance and talent deployment is now seamlessly obvious to leaders inside and outside the HR function. The decision framework and the measurement model now fit with the processes that the organization naturally uses in its planning, budgeting, and strategy analysis.

In our work with organizations, we have found this to be a significant opportunity. We routinely coach HR and organizational leaders to use talentship and the HC BRidge framework to logically and systematically identify, measure, and analyze where talent decisions can make the biggest difference—and then to consider the organization’s mental models and to present the findings within those frameworks. The framework in figure 9-6 was a perfect metaphor for a retail store, but we have seen frameworks based on brand management, materials flow diagrams, and other evocative metaphors, in different situations.

Allowing the Logical Model to Drive Measures

At this point Limited Brands has used logic and process to frame the talent measurement and decision question in a way that makes the former payroll-up system clearly obsolete. Now the task is to develop or find measures to populate the model. Figure 9-7 shows how Toyin Ogun describes the company’s measurement approach.

Limited Brands realized that the measures necessary to populate its new decision framework lay both inside and outside the usual HR layers of its data systems. In fact, the measurement system would require constructing measures that could track actual in-store movements of customers and employees.

FIGURE 9-7

Limited Brands’ science of right skills

Source: Toyin Ogun, “Limited Brands Talent Measurement” (Metrics and Analytics Executive Program, Center for Effective Organizations, University of Southern California, 2005). Reprinted with permission.

Notice the “In-store observations” notation under “Study customer behavior.” The next time you enter a Limited Brands store, look up at the ceiling, and you’ll notice small, round units with cameras inside. These are used to record real-time data on the volume, distribution, and movement of customers and employees. Limited Brands combined this data with information from its standard sales tracking system (customer relationship management, point of sale, customer interviews, and customer demographics), as shown under the “Analyze customer data” bullet in figure 9-7. In terms of HC BRidge, the result was the ability to statistically connect very in-depth information about the actions and interactions of store talent and the process outcomes that mattered most. This is a far cry from simply measuring the HR practices used in a store and relating them to store sales, and it is much more detailed than relating the general HR practices of Limited Brands with the company’s overall financial performance. The depth of the data here allows much deeper and competitively unique insights.

Making Analytics Interactive

Finally, Limited Brands developed a new approach to analytics by constructing a mathematical simulation tool. The tool integrates and combines data about customers, talent, employee behaviors, and store performance, allowing planners to run what-if scenarios that seamlessly integrate talent and organization decisions with decisions about other vital store resources and that examine what elements and combinations really move the needle in store performance. Figure 9-8 shows the architecture of that simulation. Notice the focus on optimization, not just maximization. Notice that the model emphasizes mitigating risk through modeling different scenarios. This sounds like operations engineering, and that’s no coincidence. Limited Brands drew on the expertise of its operations engineers to apply analytical logic to talent. This integration of mental models and logical frameworks across disciplines is a hallmark of a mature decision science.

Competing Better

Limited Brands combined logic, analytics, and measures. What did the company learn? One lesson is that once customers decide to enter the fitting room, they are much more likely to purchase something. This led to investments in talent systems that inform store associates and managers about how often they succeed in getting customers into fitting rooms. By directing this information to store associates and managers, Limited Brands increases the opportunity for those employees to make decisions about their own talent, actions, and interactions that can enhance the store’s success. With logic supported by measures, this clear line of sight from actions to store performance becomes a reliable part of store operations, not just a matter of luck.

FIGURE 9-8

Limited Brands’ labor scheduling architecture

Source: Toyin Ogun, “Limited Brands Talent Measurement” (Metrics and Analytics Executive Program, Center for Effective Organizations, University of Southern California, 2005). Reprinted with permission.

Limited Brands also learned that a significant amount of its sales were on weekends. Before this analysis some stores had offered their best employees time off on weekends as a reward, perhaps hoping to reduce turnover by providing the best associates with something they wanted. These are worthy goals, but a deeper decision-based logic reveals the fallacy in what might have appeared to be logical reasoning.

In fact, data showed that this was exactly the opposite of what was optimal for the store and exactly opposite of what should be done to maximize commissions for the best associates. It turned out that many of the best associates were quite happy to work weekends once the relationship between weekend work and higher commissions through bigger sales was made clear. Moreover, having the best store associates working during the highest traffic times for shopping malls enhanced store performance.

This is also a very good example of the synergy between demand and supply that we discussed in earlier chapters. If competitors continue to operate with less rigorous logic, it creates a competitive opportunity for Limited Brands to attract, retain, and deploy the best store associate candidates.

Conclusion

The Limited Brands example illustrates how a holistic approach to talent and organization measurement must encompass all elements of the LAMP framework. It also shows how systematic development of a logic that connects talent and organization investments to vital organization processes is the key to making sense of the thousands of available measures that could connect talent and organization decisions to strategic success. Limited Brands’ logic is consistent with the HC BRidge framework, while the application of that logic reveals relationships that are unique to Limited Brands’ own strategies, competitive intent, processes, and resources.

This chapter showed how the talentship decision science provides an approach to talent and organization measurement that emphasizes decisions, organizational effectiveness, and using measures to articulate and teach a systematic logical connection between talent and organization decisions and the vital elements of strategic success. Measurement has been termed the “Achilles’ heel” of HR management, but as we have seen, the real solution goes well beyond merely improving measures. In chapter 10 we will extend this theme to show how leading organizations are making talentship real by integrating it deeply into their strategy, performance, planning, and budgeting systems.