2

_______________

A Decision Science Applied to Talent

Understanding the Necessary Components

Chapter 1 defined the goal of the talentship decision science: “to increase the success of the organization by improving decisions that depend on or impact talent resources.” In this chapter we define the concept of a decision science and its necessary components. We concentrate here on the elements that characterize all successful decision sciences and how they apply to talent. Chapter 3 will more fully describe the specific decision framework that we use as a pillar of the talentship decision science—the framework that is the basis for much of the rest of this book.

Why a “Decision Science”?

In about 1999 we began using the term decision science to capture the nature of the essential evolution for HR. Since then, its use has become increasingly common among HR executives, thought leaders, and academics. The 2005 book of essays by thought leaders on the future of HR that was copublished by the Society for Human Resource Management contains an entire section entitled “See HR as a Decision Science and Bring Discipline to It.”¹ This section includes a chapter from us that applies talentship concepts to the sustainable enterprise.² It also includes other chapters, such as “Science Explodes Human Capital Mythology”; “Human Resource Accounting, Human Capital Management, and the Bottom Line”; “Improving Human Resources’ Analytical Literacy”; and “The Dual Theory of Human Resource Management and Business Performance.”

Yet there is no widely accepted definition of a talent decision science. For decades there has been a general science of decisions and decision making, producing insights about how decision makers behave and the factors that enhance and reduce their rationality and accuracy. Our concept of a decision science for talent draws on this research. As we shall see, the components of a decision science help define the necessary elements for improving talent decisions and the relevance of the HR profession. First, let’s consider the power of combining decisions and science.

Decisions

Why do we focus on decisions? As we saw in chapter 1, marketing and finance evolved to the strategically influential functions they represent today in large part by extending their paradigm from compliance to services to decisions. Functional service excellence alone cannot achieve strategic success through these resources because they are integral to the ongoing success of the organization, not isolated within a single function. The majority of decisions that depend on or impact financial capital or customers are made by those in general leadership roles outside the finance and marketing functions. This is true for talent decisions as well.

When we ask line or HR leaders to think of a decision that depended on or affected talent resources but in retrospect was not made well, even companies with best-in-class HR functions can describe numerous examples. Many of the examples are remarkably consistent. The talent decision mistakes are not typically made by HR professionals. Poor talent decisions seldom have poor HR programs as the root cause; instead, they’re made by well-intentioned leaders with unintended talent implications.

For example, one highly specialized high-tech firm made a decision to relocate to be closer to its key customer, one that accounted for well over 50 percent of revenues. The decision logic was that the organization could more efficiently and quickly serve this large customer by locating its operations closer. What was overlooked was that the key services required several sophisticated and highly specialized experts. When the move was announced, more experts than anticipated left the organization, creating a disruption that was far more damaging to client relationships and the company’s reputation than the benefits of proximity. The decision required integrating three perspectives: financial, marketing, and talent. The financial and marketing elements were logically considered, but the failure to accurately consider the talent implications undermined the intended benefits.

Several senior executives we have worked with have noted that HR strategies often reflect traditionally critical industry needs, such as “avoid employee strikes” in companies where heavy manufacturing is vital and where leaders often worked in labor relations before advancing to top HR roles. In many sales organizations the rallying cry is “reduce turnover” because turnover costs are so apparent. HR leaders point out that without a logical decision framework, such goals can become so prominent that they mask other significant organizational needs.

The greatest opportunity to improve talent and organization decisions is by improving those decisions that are made outside the HR function. Just like with decisions about financial and customer resources, talent decisions reside with executives, managers, supervisors, and employees who make decisions that impact talent, including their own as well as those they are responsible for or interact with. Even in core HR processes—such as succession planning, performance management, selection, and leadership development—potential improvements in effectiveness rely far more heavily on improving the competency and engagement of non-HR leaders than on anything that HR typically controls directly.

The relocation example is typical in that most significant business decisions impact multiple resources, so the objective should be to equip leaders with more comprehensive decision frameworks. It’s not a matter of choosing between people versus profits, with the organization’s financial controller arguing for profits and the HR leader acting as the employee advocate. Instead, the goal is a decision science that enables leaders to integrate talent resources with other vital resources. To be sure, this improves talent and organization decisions, but its ultimate goal is to improve strategic decisions more broadly.

Science

Why use the term science? Because the most successful professions have decision systems that follow scientific principles and that have a strong capacity to quickly incorporate new scientific knowledge into practical applications. Disciplines such as finance, marketing, and operations not only provide leaders with frameworks and concepts that describe how those resources affect strategic success; they also reflect the findings from universities, research centers, and scholarly journals. Their decision models are compatible with the language and structure of the scholarly science that supports them.

For example, in operations research there is often a very close connection between the technical tools used in industry and the scholarly research that informs them. In the arena of total quality management (TQM), the decision frameworks used by managers reflect fundamental logical elements—such as plan, do, check, and act—that translate into logical connections with such processes as inspection, maintenance, adjustment, and equipment replacement. This logic allows managers’ decision models to be quickly informed by research on topics like statistical process control, control charts, and time-series statistical analysis. It also provides a context for researchers, who frame their research questions consistently with the logic and practical issues facing leaders who apply TQM.³

With talent and organization, the logical frameworks used by leaders often bear distressingly little similarity to the scholarly research in HR and human behavior at work.⁴ As we will see throughout this book, this is regrettable because there is much that leaders can learn from scholarly findings and much that scholars can learn by better incorporating business leaders’ insights into their research.⁵ Compare the approach to bond ratings in finance and employee assessment practices in HR. Both strive to provide a valid and reliable measure of the future performance of an asset with some risk. A treasury department is expected to purchase information on scientifically rigorous bond ratings in its investment decisions. HR functions often lack support for scientific employee assessment investments (valid tests, interviews, assessment centers, surveys, etc.) because they don’t see the value. In fact, frameworks for comparing the costs of employment testing to their benefits have existed since the 1940s, but they are not widely used, in part because organizations’ decision frameworks have few connections to the logical principles of these models.⁶

A decision science also approaches decisions through a scientific method, which means that questions are framed so that they are testable and falsifiable with data-based results. It means that the logic supporting the decision science is modified when new findings make old ideas obsolete. It means that the decision framework clearly translates new scientific findings into practical implications. This scientific method includes, but goes well beyond, a fact-based approach to HR. Many articles that carry the label of “decision science” are about improved analytics, measurement, or scorecards. Data is certainly an element of a decision science, as we will describe later, but much of the data being used by HR lacks the logical framework and the analysis required to use it to advance either decisions or science.

A true decision science does more than just incorporate facts and measures. A decision science draws on and informs scientific study related to the resource. There is a vast array of research about human behavior at work, labor markets, and how organizations can better compete with and for talent and organization resources. Such disciplines as psychology, economics, sociology, organization theory, game theory, and even operations management and human physiology all contain potent research frameworks and findings. Unfortunately, the transfer of such findings into actual decisions is often woefully slow or nonexistent. A decision science connects research to the practical dilemmas facing decision makers in organizations. It also provides a means to apply research on talent and organization to other fields and to bring insights from other scientific fields (such as operations, strategy, marketing, etc.) to bear on talent decisions within organizations.

Components of a Decision Science for Talent

To implement a decision science for talent, it is important to consider what is required. Based on what we have learned from fields outside HR and our experience in implementing a decision science for HR within organizations, we find that there are five important elements in a mature decision science:

• Decision framework

• Management systems integration

• Shared mental model

• Data, measurement, and analysis

• Focus on optimization

Decision Framework

The decision framework defines the logical connection between the decisions about a resource and the organization’s ultimate goals. It defines how the organization should think about the talent and organizational implications of business decisions in a common and consistent way. It provides the basis for evaluating and improving the decisions that involve the resource.

An effective decision framework provides a consistent logical model of the chain of causal connections divided into independent elements. The DuPont return on equity (ROE) model is a good illustration.⁷ There is a goal ROE segmented into three elements (margin, asset productivity, and leverage), as shown in figure 2-1.

ROE is a consistent and logical model that can be expressed as an algebraic formula. It also depicts the causal chain of capital:

Starting at the denominator in the right corner of the model, the causal chain can be described as follows:

FIGURE 2-1

DuPont return on equity model

Reprinted by permission of the Business History Review 49, No. 2. Exhibit from “Management Accounting in an Early Integrated Industrial: E. I. DuPont de Nemours Powder Company, 1903–1912,” by H. Thomas Johnson, 1975. Copyright © 1975 by the President and Fellows of Harvard College; all rights reserved.

• Equity (investment) is used to acquire assets (the ratio of assets to equity is leverage).

• Assets are used to generate sales (the ratio of sales to assets is the asset productivity).

• Sales generate profits (the ratio of profits to sales is the margin).

There is a significant amount of independence between margin, asset productivity, and leverage. While there is almost never complete independence between the elements, a good decision framework will achieve as much as possible. Within the DuPont model, lowering the cost of goods sold could increase margins without affecting asset productivity or leverage. Increasing the accounts receivable could improve asset productivity without affecting margin or leverage. Finally, reducing equity by increasing debt could increase leverage without affecting asset productivity or operating margin.

The same standard chain of causal connections can also be seen in the decision frameworks from marketing. While there can be many variations, a typical decision framework for marketing is:

HC BRidge: The decision framework for talentship. The causal chain for talent and organization decisions can be described as:

Figure 2-2 illustrates the causal logic of finance, marketing, and talentship that we have described. While the analogy is not perfectly precise, you can see that the underlying logic is similar. The point is not that the talentship logic maps perfectly against marketing and finance, but rather that it is logically consistent with them. Resources are expended on activities or assets; those activities or assets produce changes in targets, such as sales, customers, and talent pools; those targets produce changes in financial outcomes or other sustainable strategic success factors.

The segments are also independent. You can spend the same amount to produce training activities (efficiency) and get far different results from the training programs and practices (effectiveness). Likewise, you can enhance skills in different talent pools (effectiveness), but the outcomes achieved through the new skills can vary significantly. The same level of efficiency can produce different levels of effectiveness. The same level of effectiveness can produce significantly different impact. As we shall see, when organizations lack such a framework, they mistakenly consider only one part of the logical chain (such as squeezing HR budgets to produce more efficiency without considering effectiveness or impact), or they mistakenly assume that improving one element improves the others (such as assuming that if employees have more training, the organization will compete better on its unique knowledge).

We call this decision framework for organization and talent the “HC BRidge framework” and refer to impact, effectiveness, and efficiency as “anchor points.” (The “HC” stands for “Human Capital” and the “BRidge” reflects the metaphor of three anchor points supporting a set of linking elements that collectively span the logical connections between investments in talent and organization programs and practices and the ultimate goal of sustainable strategic success. The capitalized letters B and R in HC BRidge symbolize Boudreau and Ramstad.) This framework will be described in more detail in chapter 3, and each anchor point and linking element will be used throughout the book to describe how the decision framework reveals new insights about competing for and with talent.

FIGURE 2-2

Finance, marketing, and talentship progressions

Traditional HR Decision Frameworks. HR leaders’ common first reaction to talentship and the need for a decision framework for HR is “We already have many systems designed to help leaders outside HR make decisions, like salary structures and competency systems.” Consider salary structures. Virtually everyone knows their salary grade, and employees and managers routinely use salary structures in their decisions about budgeting, headcount planning, merit pay, and other rewards. Because salary grades are often the only available framework for mapping the organization’s talent resources, they can become the default framework for things such as signature authority, participation in leadership programs, parking space allocation, and many other decisions unrelated to the original purpose. The salary grade system certainly affects decisions, but it is not a decision framework in the way that we have defined it. It is an organizing framework for the delivery of an HR system, not a decision framework for the resource.

Competency systems that span an enterprise serve a similar purpose. When done well, they provide a common architecture for defining, measuring, and developing capabilities within an organization, including not only the requirements for a job but also the logical progression and important transition points between jobs. Such systems can be used to align key talent management systems, and individual measurement to help develop both individuals and the talent pool overall. Like salary grade structures, they are very important and useful, supporting both decisions and data analysis. Even well-developed versions of competency systems, however, do not provide necessary insight into important business questions such as the talent implications of alternative business models, organizational structures and design, competing talent market value propositions, and the value of HR investments.

Salary structures and competency systems are very similar to the organizing frameworks such as the “chart of accounts” in accounting that existed before the decision science of finance emerged.⁸ The chart of accounts provides a classification system to organize large amounts of data, and it certainly improves the consistency of decisions. Like the chart of accounts, the salary grade system provides its primary value through systematic management control, and it emerged during the control era of personnel that we described earlier. It often creates more value in restraining excessive investment than in identifying areas where increased compensation would optimize organizational value. Competency models emerged more recently, during the services stage of the HR profession, to integrate and align HR services. When compensation and competency systems are integrated within a decision science framework, they become even more powerful and offer potentially greater impact on the organization, because they can be deployed in a more strategically relevant and integrated perspective.

Management Systems Integration

The next component of a decision science is the integration of its decision framework and key principles into the general management systems. The decision framework must be integrated so that decisions about different resources seamlessly support the overall organization and business processes. When this is done well, it induces decisions that consider all the key resources, including talent, rather than focusing on resources in isolation. This integration also requires more alignment between the general business planning processes and functional planning processes. For example, planning for the finance and marketing functions is closely tied to planning for the overall organization.

Financial frameworks are well integrated into general management systems in most organizations. In fact, one challenge to improving talent and organization decisions is that companies only have financial management systems and lack well-developed decision systems for other important resources. For example, many organizations have strategic-planning processes that seem to be far more focused on preparing the long-range financial plan than on optimizing the strategic position. It is not so much that the finance function intentionally dominates but rather that its processes and frameworks are much more mature and integrated. The same integrated attention is rare in strategic planning for the leadership capabilities that support the financial plan.

Two types of management systems must be integrated with a mature decision science for organization and talent:

• Management systems outside the HR function, where talent issues should be considered and addressed, which include strategic planning, product line management, corporate development (e.g., mergers and acquisitions), operational budgeting, and capital budgeting

• Talent management systems (often within the HR function) that must consider the strategic context of the organization, including workforce planning, staffing, development, performance management, compensation, and succession planning

Perhaps the most difficult way to change organizations is to impose new decision frameworks that compete with useful management processes already in place. Too often, HR organizational effectiveness leaders roll out their planning and decision systems as distinct additions to existing processes. In our experience we’ve encountered workforce planning systems that have little commonality and integration with the long-range planning processes. HR leaders would often better serve the organization by improving the business planning process in partnership with other leaders rather than deploying new HR processes that are potentially poor substitutes. We find that the more HR and talent planning processes are separate and distinct from the core management processes, the less they are strategically effective.

Let us illustrate how talent decision frameworks can integrate with management systems using the capital budgeting system, which is well refined in most organizations. One of the basic financial assumptions is to expect a higher return on investments with higher risk. Many of the potential risk factors in capital investments are linked to talent resources. One way that capital budgeting could be more integrated with talent resource decisions is to specifically identify the talent and organization risk factors associated with capital investments (availability and quality of leadership, degree of organizational change required, experience in the organization with the technology, etc.) and then set a higher return hurdle rate for investments that are higher risk due to organizational factors.

Integration of talent and organization systems with management systems must also consider the sequence of planning processes. It is not unusual to encounter organizations where the talent planning process occurs long after the organizational planning process has concluded, and where the HR budget basically distributes the allotted budget for head-count or training. In some organizations HR planning couldn’t possibly affect organizational strategic decisions because the HR process occurs after all the key strategic decisions are made. Even something as simple as timing the HR planning process so its results are available to the broader budgeting and planning process can significantly enhance integration with the management processes and the quality of decisions that result.

Leaders at Pepsi in the 1980s referred to decisions about hiring and training levels as the “human capex,” meaning the human capital expenditure plan. They believed that organizations should treat the human capex process as equally important as the financial capex. As John Bronson, former executive vice president of HR at Pepsicola Worldwide, recalls:

One of the legacies of Andrall [Andy] Pearson at PepsiCo was the MRPA, the management resources planning audit. It was his audit of the talent of the organization. He was unabashed that it was his process, not HR’s process, and not even the division presidents’ process. He expected leaders to treat their human capex with the same importance as the more traditional financial capex. Andy believed that blue-chip companies required blue-chip players. He was relentless in driving the process through the PepsiCo organization. He was like a merchant banker reviewing a financing plan. If the business-unit CEO couldn’t explain how the talent plan supported the growth, budget, and financial capex, not only would his plan be in jeopardy; he might lose his job. To Andy, if you didn’t have a solid plan for how talent supported both superior business performance and growth, you weren’t a serious contender for larger jobs.⁹

A fundamental requirement for this kind of business leader accountability is a shared decision science that aligns functional and general management systems. HR is a less mature profession, so there is often less consistency in the decision models used within the profession when compared to more mature fields like finance. Here is one of our favorite discussion questions for business and HR leaders: suppose you asked ten controllers to address a specific financial challenge, and you asked ten HR professionals to address a specific organizational or talent challenge. Where would the responses be more consistent and aligned, among the controllers or the HR professionals?

The answer is nearly always that there will be less consistency and alignment on the people issues than the financial ones. In addition, HR leaders’ different backgrounds, experiences, and perspectives will drive variation among their responses. This can be seen when there is significant misalignment between HR professionals at headquarters and the business-unit HR professionals. Examples include different approaches to goal setting in performance management and compensation, and different assumptions and models for individual development from the professionals who drive the design of succession-planning and development systems.

This is a symptom of HR functions operating without a common point of view. Without a consistent decision framework connecting the various elements of the talent management systems and decisions, they lack a consistent message to integrate their core systems, much less affect the broader management systems outside the HR function.

Thus, aligning HR systems, such as talent planning and HR functional planning, with the decision framework is a vital requirement of a mature and effective decision science and a platform for integrating talent decisions within the broader management systems.

Shared Mental Model

A successful decision science is used by organizational leaders as a natural part of their work. Its logic elements are a part of the mental models and mind-set of key decision makers inside and outside the profession. All organizational leaders are expected to be conversationally competent in the basic principles of the decision framework and are required to have the skills and professional support to use the systems that require their direct involvement. Every business leader must be conversant with principles from the finance decision science, such as net present value and assets and liabilities. Likewise, they must be familiar with marketing decision science concepts, such as customer segments and product life cycles. The marketing and finance functions provide support and deeper professional capabilities, but all managers know that they cannot abdicate the basic knowledge of finance and marketing principles to others. These same principles operate in management systems like budgeting, so leaders not only understand why they are important, but they routinely use the fundamental decision science concepts. So there is less resistance and more opportunity to improve decision making with those systems. It doesn’t seem strange if marketing or finance suggests enhancements to the decision systems for general managers because those managers are already accustomed to working with marketing and finance principles.

Today, individual leaders too often approach talent and organization decisions with vastly different mental models, divergent logical principles, and a focus on very different factors. The sources of principles vary—from motivational speakers and high-profile executives to successful athletes and an occasional college professor. Without a common understanding of the key principles, talent and organization management systems lack context and are seen as administrative or bureaucratic. As the talent and organization decision science matures, its principles will become more consistent and a more natural part of the mind-set of both line leaders and HR leaders, with the appropriate level of sophistication.

To achieve this goal, HR leaders will need to focus more on teaching than telling, a significant change. Finance and marketing are effective in part because their principles have been taught to business leaders in business school, followed up with executive development, and reinforced with real-world practice and career experiences in which leaders are usually coached by functional specialists. Even with very mature and strong staff functions within finance and marketing, the need for senior leaders with well-developed competencies in finance and marketing is seen as important. Talentship will produce decision frameworks that will be consistently taught to organization leaders, becoming a natural part of their work and decisions.

The need for a talentship mind-set is also vital for HR functional leaders. In our experience there are almost always some HR professionals in any organization who effectively understand, teach, and enhance decisions based on how talent connects to strategic success. These HR professionals typically admit that they learned to provide this kind of support in their own way, with little systematic instruction or development. One HR professional put it well: “This capability is critical to our future, but it doesn’t scale because everyone does it and learns it differently.”¹⁰

A decision framework contributes to scale by developing, using, and teaching a consistent, logical point of view about how to connect talent resources to strategic success. A logical point of view provides a consistent script for an ongoing dialogue about talent and strategy, allowing more reliable and consistent diagnosis, analysis, and action on talent issues throughout the organization.

Data, Measurement, and Analysis

A mature decision science has data, measurement, and analysis aligned with its decision framework principles. These are refined and deployed through management systems, used by leaders who understand the principles, and supported by professionals who add insight and expertise. Today finance reflects this level of maturity almost everywhere, and marketing approaches this level of maturity, particularly in industries where competitive dynamics hinge on marketing sophistication, such as consumer products and multilocation retail.

These systems have evolved over decades, and today we hardly notice how well integrated financial measurement and analysis processes are with financial decision models. It seems to have always been that way. For example, as described in table 2-1, today the ratios commonly measured in financial decisions and the structure of accounting statements link directly to the DuPont decision framework. Similarly, marketing decision frameworks provide the logical structure for customer relationship management and customer analysis systems, which use vast amounts of data mining and advanced analytics to produce competitive insights.¹¹

In stark contrast, HR data, information, and measurement face a paradox today. Although there is increasing sophistication in technology, data availability, and the capacity to report and disseminate HR information, frustration increases when investments in HR data systems, scorecards, and integrated enterprise resource systems fail to create the strategic insights needed to drive organizational effectiveness. One reason for this paradox is that the technological advances have outpaced the fundamental logic connecting talent and organization decisions to strategic success. Major elements of marketing and finance are well over fifty years old, so those decision sciences were far more mature by the time technology advanced. The computer-enhanced systems could build on well-developed decision frameworks, integrated management systems, and shared mental models, making information technology much more valuable.

TABLE 2-1

Ratios measured in financial decisions

	Common supporting analysis ratios		Source of data for the numerator and denominator
Margin	Gross margin		Both are from the operating statement (profit and losses)
	Cost of goods sold
	Sales, general, and administrative expense ratio
	HR as a percentage of revenue
	Operating margin
Asset Productivity	Accounts receivable (in days)		One is from the operating statement, and the other is from the balance sheet
	Accounts payable (in days)
	Inventory turns
	Cash operating cycle
Leverage	Debt-to-equity ratio		Both are from the balance sheet

HR has no such decision science and no decision framework to organize information technology. Thus, technology has found its greatest value in automating areas with more established organizing frameworks, such as payroll, but has not reached the level of impact in supporting more strategic decisions. For example, HR functions often brainstorm their own unique employee turnover classifications when installing a new software system. It’s not surprising that even after years of using such systems, there remains too little insight into the factors that affect employee turnover, how it affects the organization, and what to do about it.

As we will discuss later, HR measures exist mostly in areas where the accounting systems require information to control labor costs or monitor functional activity. Efficiency gets a lot of attention, but effectiveness and impact are often unmeasured. While there have been significant advances in applying analytics to the field of HR management—including high-level data analysis approaches like social network analysis and multivariate regression—such methods often suffer from the lack of a more comprehensive decision framework. For example, a statistical method from marketing, called “conjoint analysis,” has been applied to employee survey data to see which work elements most significantly associate with employee engagement or turnover.¹² This provides insights about how HR programs might enhance those work elements, but it often fails to identify where engagement and retention matter most and why. Advanced analytics hold great promise for enhancing talent decisions, but as we shall see later, it is often the logic, not the analytics, that creates the big breakthroughs.

A decision framework provides the logical structure to organizational data, measures, and analytics, and identifies gaps in existing measurement systems. Armed with such a decision science and framework, organizations can avoid investing in sophisticated data and analysis that fails to achieve its potential because the tools don’t address the important questions.

Focus on Optimization

The final pillar of a mature decision science is that its logic reveals how decisions can optimize the returns from a resource, rather than simply describing them or only partially maximizing them. A mature decision science reveals how to optimize results by balancing trade-offs instead of assuming that more is better.

Finance provides a good illustration. Before the DuPont model (which marks the beginning of the decision science of finance), the goal was simply to maximize profits. Disproportionate amounts of capital were directed to businesses with large profits, often resulting in high margins but low return on capital. Instead of maximizing profits in isolation, the DuPont model strove to optimize profits by recognizing the constraints on financial capital resources.¹³ As the finance decision science matured, other factors were integrated. Financial decision models now not only maximize returns but use decision frameworks, such as portfolio theory, to optimize return in the context of risk and liquidity. Further refinements revealed how to balance liquidity in the broader strategy, by investing in ways that consider the range of future strategic options that investment might enable.

By contrast, many HR decisions often try to increase learning, engagement, or retention without limit or context. This is very different from optimizing a portfolio of HR practices against the organization’s unique resource opportunity costs and constraints.¹⁴ For example, if more sales training increases product knowledge, which increases selling success, a less mature decision framework might apply training more broadly. Having proved the value of training by linking it to increased sales, the right decision seems to be to acquire more training. Several executives we have worked with have termed this the “peanut-butter” approach, because it spreads something equally across the entire organization.

In fact, considering the necessary investments (time, money, etc.) to achieve increased selling success through training, enhancing product knowledge from an already high level may be very expensive. The optimal solution might involve less product knowledge and more motivation, and thus less training and more incentives. The key point is that a mature decision science frames the question in terms of optimal solutions rather than just describing relationships or increasing one desired outcome out of context. Even when optimal decisions can’t be precisely defined, the logic that a focus on optimization provides will often lead to insights that are missed by a less comprehensive approach.

Distinguish Average from Marginal Value. A core principle in optimization is the difference between average and marginal (or incremental) impact. Although something can be highly valuable, increasing or decreasing the amount of it may not have a big effect. For example, suppose an organization has a hundred sales representatives and total revenue of $50 million, making the average sales per sales representative $500,000. What is the optimal number of sales representatives? You can’t tell from the average. Optimizing requires that you know the potential effects of increasing or decreasing the sales force. If the same $50 million in revenue could be generated by forty representatives, then the marginal value of the last ten representatives would actually be zero. On the other hand, if these sales representatives were working to their capacity and there were available sales territories without adequate sales coverage, additional reps would create significant incremental sales.

A mature decision science clearly articulates the difference between resources or activities that provide high average value and those that provide high marginal value. We use the word pivotal to describe the marginal effect of resources, activities, and decisions. Pivotal captures the idea of a lever, where a small change at the fulcrum causes very large changes on the other end. Highly pivotal areas are those where a small change makes a big difference to strategy and value. A resource, decision, or activity can be highly valuable and important, even if it is not pivotal. Some resources, decisions, or activities are both important (highly valuable on average) and pivotal (small changes make a big difference).

A good example is product design. Consider how two components of a car relate to a consumer’s purchase decision: tires and interior design. Which adds more value on average? The tires. They are essential to the car’s ability to move, and they impact both safety and performance. Yet tires generally do not influence purchase decisions, because safety standards guarantee that all tires will be very safe and reliable. Differences in interior features—optimal sound system, elegant upholstery, portable technology docks, number and location of cup holders—likely have far more effect on the consumer’s buying decision. In terms of the overall value of an automobile, you can’t drive without tires, but you can drive without cup holders and an iPod dock. Interior features, however, clearly have a greater impact on the purchase decision. In our language, the tires are important, but the interior design is pivotal.

Figure 2-3 shows this example in the form of what we call a “performance yield curve,” a fundamental idea underlying talentship that is used throughout the book. The performance yield curve for tires is much higher than for interior features, which reflects tires’ importance. The yield curve for tires is relatively flat across a large range of performance levels, and it drops quickly on the left if tire performance falls below a certain level. Tires create tremendous value and are very important, but once they reach a certain level, increasing their performance does not add value to the consumer’s purchasing decision. Yet, if they fall below the minimum standard (as happened with the Firestone tires on Ford SUVs, resulting in a massive recall in 2000), the result is very bad indeed. The key to optimizing tires against their effect on the initial purchase decision is to get them to standard, not significantly higher. Beyond this point the incremental cost of increasing tire performance exceeds the incremental value in customer purchases.

FIGURE 2-3

Yield curves for automobile components: Tires vs. interior design

The marketing and finance decision sciences have sophisticated systems to exploit the distinction between marginal and average value. As the marketing decision science evolved, the concept of segmentation was applied at multiple levels, including markets, customers within markets, products, and as we just discussed, product features within products. Conjoint analysis and other statistical tools use data from extensive consumer research to produce deep insights about the incremental value of features, which informs decisions about product designs. They carefully isolate those attributes (such as safe tires) that are core or expected (sometimes referred to as “table stakes”) from those where differences drive perceptions of value (such as the interior design of a new car). Optimization requires investing based on the incremental contribution, not the average contribution. To do this, features must be segmented on their marginal value (pivotalness), not their average value (or importance). Failing to segment based on pivotalness often results in equal investments, even when the potential marginal return is significantly different, as shown in figure 2-4.

FIGURE 2-4

Segmentation and the dangers of equal investments

This seems fundamental, but it is frequently misapplied in organization and talent decisions. What we have called “talent segmentation” is still very rudimentary.¹⁵ We see this in the frequent tendency to do the same thing across a wide range of jobs or talent pools. Examples include: “If stock options are good for executives, then they should be expanded to all employees” and “If it’s important to increase attention given to our customers, then everyone should have thirty hours of training in customer awareness” and “If weeding out the bottom 10 percent of performers makes sense in our sales force, let’s weed out the bottom 10 percent in every job.”

Effective segmentation based on marginal value in talent and organization decisions helps answer questions such as “Where does my strategy require increasing the performance of our talent, and how it is organized?” The answer cannot be “everywhere,” because that is cost prohibitive. The answer also cannot be “nowhere,” because competitive advantage must have some source—one or more multi-incumbent roles, or talent pools, where superior talent quality makes a significant strategic difference. Lacking a decision science that guides this kind of talent segmentation, organizations typically invest too little in talent pools that are most pivotal and too much in talent pools that are important but far less pivotal. The idea that talent and organization decisions are vital to competitive advantage is a virtual truism today, as we noted earlier, yet the very essence of competitive advantage is finding unique and different ways to advance a particular value proposition, seize specific market opportunities, or leverage distinctive strategic resources. Still, today’s decisions about organization and talent are often made with an eye toward duplicating the practices of other successful companies, rather than in-depth internal analysis to find the appropriate investments for specific contexts. The absence of a decision science that distinguishes marginal from average value is a significant cause.

Segmentation in Auto Insurance. Segmentation based on marginal impact produced a competitive advantage for Allstate Insurance Company. Allstate (then a division of Sears) was one of the first companies to adjust rates based on age, auto usage, and claims history—a revolution at the time.¹⁶ This easily described idea had massive implications for virtually all aspects of the auto insurance business. Allstate was able to extract much more value from the insurance market and provide much greater value to its customers by adjusting what it charged according to customer characteristics related to the probability of accidents and other factors. Allstate has continued its tradition of innovative differentiation in its pricing models by bringing in a variety of new factors that it now markets under the brand “Your Choice Auto Insurance,” which is based on sophisticated pricing models. This changed the game from a product definition perspective by providing customers a much wider variety of choices. Before, customers could choose their level of coverage and size of deductibles. Now they can customize policies on features such as accident forgiveness and what type of rewards they would like associated with good driving records.

Allstate’s research revealed interesting and surprising patterns in auto safety among different consumer groups. A BusinessWeek item noted:

For decades, Allstate had lumped customers into three main pricing categories, based on basic details such as a customer’s age and place of residence. It now has more than 1,500 price levels. Agents used to simply refer to a manual to give customers a price; now they log on to a computer that uses complex algorithms to analyze 16 credit report variables, such as late payments and card balances, as well as data such as claims history for specific car models. Thus, [drivers who are safe bets] are rewarded, saving up to 20% over the old system, and high-risk drivers are penalized, paying up to 20% more. It has worked well enough that All-state now applies it to other lines, such as homeowners’ insurance.¹⁷

Thinking Differently Using Decision Science Principles

A decision is an invitation to think differently. Historically, as decision sciences become embedded within organizations, natural synergies emerge across the five decision science elements. This creates tangible but very organic changes in the way business and HR leaders, employees, investors, and potential employees converse about a strategic resource. Consider the power of just three changes in the way your organization approaches its talent and organization decisions.

First, clearly distinguishing between pivotalness and importance motivates a focus on the marginal value of talent decisions. This is as important for talent as the distinction between the marginal value of advertising and the overall importance of advertising, for example. It helps decision makers avoid getting lost in a sea of important initiatives and set priorities correctly.

Second, consistently use performance yield curves to identify the nature of pivot-point slopes and shapes. Not only does this kind of discipline help identify where decisions should focus on achieving a standard versus improving performance, it also provides a way to think about the risks and returns to performance at different levels. It helps people avoid making decisions based on well-meaning but rudimentary rules such as “get the best person in every job.”

Third, focus on optimization, not just maximization. This creates an environment in which trade-offs can be discussed with less of the emotion that usually prevents good decisions and often leads to decisions like “Let’s just do the same thing for everyone to be fair.” Optimization presumes that talent investments will be unequal but also creates a high standard for analyzing and communicating good reasons for such unequal investments.

Thus, even in the early stages of implementing a decision science, tangible changes occur in how talent and organizations are understood and made. History shows that it is from these small tangible steps that significant untapped strategic success flows.

Conclusion

When new decision sciences emerge, they typically present difficult changes in social, organizational, and personal traditions. Before the new logic is used by competitors, a failure to make decisions more optimally doesn’t create any relative disadvantage, so less sophisticated decision systems still allow organizations to stay competitive. Before Allstate applied the decision science principle of customer segmentation and optimization, no one did any worse by following the old model. Yet first movers who apply a new decision science often create formidable competitive advantages. Once Allstate generated value by adopting a more sophisticated decision science, its competitors were at a disadvantage. Soon everyone began to realize the power of the new decision framework and tried to catch up.

We believe that HR, and the larger domain of organization and talent decisions, is at precisely this historical point. A few organizations are beginning to develop some elements of a more mature decision science. For example, at Corning, HR leaders who support key divisions use talent-focused strategy analysis during their annual strategy sessions.¹⁸ At The Hartford Financial Services Group, Inc., investments in HR programs are allocated in part based on where they will have the most pivotal effect.¹⁹

Still, because the effects are so isolated, there is not yet an urgent need to evolve. The vast majority of organizations can compete effectively, even while making more traditional talent and organization decisions. A new decision science, however, will emerge for talent, just as surely as it emerged for other resources. The first organizations to apply the new decision framework will achieve significant first-mover advantages, forcing others to react. In time we envision the talent decision science becoming as natural a part of management thinking as finance and marketing are today, but before that happens there are opportunities for game-changing strategic decisions by organizations that apply it first.

The remainder of this book describes this emerging decision science: talentship. It rests on the pillars of management systems integration, shared mental models, and aligned data, analytics, and measures. The decision framework based on impact, effectiveness, and efficiency is the core pillar around which all these decision science components revolve. So we have organized the book around the decision framework and its vital core elements. Next, we describe those core elements.