3. Value Creation and Advanced Analytics

The reason human capital is so difficult to commoditize is because it possess something other forms of capital do not: asymmetric or private information. Unlike financial or technical capital, if it so chooses human capital can keep what it knows to itself, withhold effort, or, frankly, just leave. To maximize the contribution of human capital, employers develop incentive contracts—policies and practices developed to enable and motivate the workforce. When these practices and policies that maximize human capital’s contribution are combined with other complementary forms of organizational capabilities (for example, work processes, technologies), it becomes nearly impossible for the competition to replicate, consequently providing a powerful and sustainable competitive advantage. The trick is to determine exactly which policies and practices for your specific human capital and organizational capabilities optimally promote your organizational objectives.

Getting human capital management (HCM) policies and practices right provides two main benefits. One, you will be much more likely to retain the human capital that is associated with your success, and two, you will have greater output, more innovative products and services, much more satisfied customers, and greater product and service quality. Consequently, this will result in cost savings (employee turnover is very expensive) and greater growth.

The use of advanced analytics can assist in making these determinations much more accurately. You can do this in a number of specific ways, including the following:

• Predict much more accurate outcomes.

• Use better models of organizational value creation.

• Model how humans actually decide.

• Utilize agent-based modeling.

• Recommend optimal practice and policy choice.

• Determine optimal policy and practice.

• Deep Q&A expert systems.

• Signal more accurately ability and potential.

• Determine optimal selection and promotion.

• Use bibliographical data.

• Map individual and team performance to organizational outcomes.

• Performance management and incentives.

• Map contribution to organizational objectives.

• Share knowledge and know-how.

• Evaluate the impact of planned and potential policy and practice changes.

• Optimize employment levels, hours worked, and benefits.

• Diagnose problems and recommend solutions.

The existence of asymmetric information is the reason why the effective management of human capital is so critical for organizational success. This is also one of the primary reasons why it is critically important to retain key human capital. One of the primary reasons employee turnover is so costly is because employees have a bad habit of taking what they know with them when they leave. Human capital has the unique ability to decide what information it cares to share and what information to keep to itself. However, there is another source of information that also has tremendous value to organizations, and that information is found in data, the data organizations keep on their human capital.

Most organizations have everything they need to constantly be evaluating and, more important, experimenting with new ideas and ways to engage, enable, and excite the human capital. This does not mean that you have to roll out a tremendously expensive new program, but you can experiment at one work site or with a work team and see how the new initiative is received. We should always have the answer to this question: Is the program worth it? Or, did the program work? For instance, one type of program that has been around for a long time is the organizational wellness program. Do these programs save the organization money or not?²

Take, for example, Yahoo!’s new policy that requires everyone to come into the office. There is an identifiable date, so assuming that they have the necessary data, they have everything they need to determine whether (after the introduction of that new policy) employee turnover decreased, increased, or stayed the same. They can look at absenteeism, employee morale, whether the company developed more innovative products, the number of customer complaints, and so on, ultimately determining if there is a relationship between the introduction of this new policy and outcomes of interest.

Panel data analysis provides us with the means to determine the holy grail of econometric analysis. This type of analysis allows us to evaluate (and hopefully establish) a cause and effect relationship, enabling us ultimately to say declaratively that a particular policy and practice was closely associated with a specific outcome. Time series or panel analysis consists of evaluating the impact the introduction a policy and practice has on variables of interest. For example, you just put in place a child-care facility in your office in New York, but not in California. Once some time passes, say three months, you have pretty much everything you need to conduct impactful econometric analysis, establishing (or not) a cause and effect relationship between the use of a specific practice and the impact on outcomes of interests. Those outcomes can include everything from employee turnover, absenteeism, sales, and customer satisfaction, to name a few.

Organizations are constantly conducting cost benefit analysis associated with market plans, new products, and so forth. However, the same sort of analysis associated with costing HCM issues is done less often. Organizations can experiment in different ways to evaluate and refine policies and practices that are in place and thus tailor their activities to obtain the maximum long-term positive impact.

One frustration I had when working for organizations was that we never really had time to evaluate the impact of the practices we were putting in place. We were so tied up with the immediate day-to-day activities that evaluating the effectiveness of what we were doing was not on the radar. At the time, I knew that was suboptimal, because conducting this sort of analysis could provide extremely valuable information on the impact of the policies and practices in place, and also inform future decisions about what to do. We (I was really just involved in the project at the very beginning) did end up getting a number of organizations to participate, and those that did participate got some really great analysis, and scholars at the CEP were able to publish some great papers using the data.

When people think about valuable data, they often think about data on consumer spending habits or people’s search results, which is information that is ultimately very valuable to advertising organizations. Less often considered is the treasure trove of data organizations have on their own primary driver of organizational success: human capital.

Organizations usually have a large amount of data on employee salary histories and demographic data. There is also data on training and performance appraisals and possibly employee satisfaction surveys. In addition, and very importantly, there is data on the date a new incentive plan was put into place (for example, when the child-care facility started or when a wellness program was established). Couple this with organizational performance data and you are able conduct a panel data analysis that provides a clear cause and effect relationship. Often, unfortunately, at best data is used to show correlations between two variables. Yes, information from correlations may point us in the right direct (or the wrong one), but it says nothing about cause and effect. Here again is an opportunity provided to us by big HCM data and sophisticated analytics. (I have not talked about how critical data integrity is, but it is a fundamental starting point.) We can use this information to both evaluate what we have done and to make better decisions about what to do in the future.

• Measurement error: Are you accurately measuring the concept you are attempting to measure (for example, employee morale and satisfaction)?

• Omitted variable bias: Is the observed result influenced by factors not included in your model (for example, variation in the quality of managers)?

• Reverse causality: Do higher profits make happier employees or do happier employees make higher profits?

Increasingly, the question is being asked: What drives intangible capital formation? While a substantial amount of the discussion has revolved around the issue of the measurement of intangibles,⁴ the work of those like Robert Kaplan and David Norton has gone a long way toward identifying how it is formed and developed. One of the key drivers of intangible assets is inputs from human capital.

One definition of organizational capital is:⁵

The knowledge used to combine human skills and physical capital into systems for producing and delivering want-satisfying products. It relates but is not limited to the following: (a) operating capabilities; (b) investment capabilities; and (c) innovation capabilities.⁶

Others view organizational capital as primarily residing within human capital,⁷ and still others view it as embodied in the organization itself.⁸ Organizational capital recognizes that inputs from human capital combined with other organizational capabilities are the primary value creation mechanisms within organizations.

Yu Peng Lin and I were interested in just what role organizational capital played as a mechanism for why we were seeing better performance in companies that broadly distributed stock options. As mentioned, organizational capital is meant to identify organizational capabilities, human and technological. By evaluating a measure of organizational capital, we were able to determine whether the higher degree of output was due to this combination of human and technical capital. We found strong evidence that stock options were associated with greater organizational capital and greater output.⁹

This research showed that much of the value created from stock options was due to an increase in organizational capital, and we concluded that this was also associated with much lower employee turnover. This finding, along with substantial other research, points us toward making certain that employee turnover is kept as low as possible. This, in turn, means that the value-creating mechanism within an organization is largely a function of inputs from human capital. This is not to suggest that the other inputs are not necessary, but maximizing the contribution of human capital is the way to create significant value in the organization.

It is difficult to adequately emphasize just how critical it is, for the well-being of the organization, to keep key employee turnover as low as possible. When an organization finds someone who is an especially good fit, the organization should ensure that they are motivated, satisfied, and challenged. Due to its vesting requirement, broad-based stock options are especially good at keeping employees tied to the organization. This “retention effect” ultimately results in greater output and financial returns.

Again, I want to emphasize that this does not mean that every organization everywhere would be better off if they offered stock options to everyone. The key is to have an engaged workforce, and there are many ways in which to accomplish this, depending on the nature of the organization and the workforce.

• Who does what?

• How do we get them to do it?

You are also losing all their accumulated organizational-specific human capital. We will see that keeping employee turnover as low as possible makes especially good sense if you have employees who have access to very valuable information in the form of information on customers, new products or services, product quality, and their own human capital; and this means just about everyone.

Keeping employee turnover as low as possible justifies investing in programs and policies that are focused specifically on that end. For instance, child-care services, or a health club, or deferred compensation, whatever policies fit your specific organization. As mentioned, the company SAS has onsite health care, child care, and social workers available to assist with elder care and other family issues. These programs all allow employees to focus on their jobs, and they also serve to keep employee turnover low. The related costs of these programs are more than paid for by the 4% or 5% annual employee turnover that this company sees versus the 20% or higher rate in the rest of their industry.

Oddly, many organizations and some of the research has long looked for best practices. These are practices and policies that work every time for everyone; essentially, one size fits all. Does following the best practice philosophy actually optimize the contribution from human capital? The answer is no. What motivates a cowboy will not motivate a derivative trader, and, for that matter, what motivates one cowboy will differ significantly from what motivates another.

The stock option example I mention in the first chapter is another example. Yes, broadly dispersed stock options are associated with greater levels of productivity, but the important qualifier is that this in no way means that everyone should give options to everyone or to anyone for that matter. Nor does it even mean that every technology firm should use stock options. The company I mentioned earlier, SAS, a tremendously successful privately owned company (giving phantom options in private firms is done often) does not give anyone stock options. They have one of the lowest turnover rates in the industry. The practices and policies you put in place depend entirely on what you are trying to accomplish and with whom.

Effective human capital management has been defined as follows:

A pattern of planned human resource deployments and activities intended to enable an organization to achieve its goals.¹⁰

An issue that needs immediate attention in a world of practically an unlimited number of potential HCM deployments and activities is this: Which specific ones are the most effective in any particular situation? Recently, there has been some controversy surrounding the optimal choice of policy and practices.

There are two aspects to strategic HR issues: first, the determination of the optimal way in which the HR policies and practices can support the objectives of the firm; and second, the execution of these policies and practices. The use of advanced analytics and emerging technologies can assist considerably with the determination of optimal HR policy and practice choice, and the emerging technologies can go a long way toward the execution of the various policies and practices.

This is the domain of strategic HR management, and ideally, to make optimal choices, you want to consider a combination of micro and macro factors. What practices make the ideal choice under specific situations, and how can advanced analytics assist with that?

An equation for this model is as follows:

Organizational Strategy + HR Policies and Practices = Organizational Success

• Universalistic: The universalistic approach holds that there are certain best practices that should be adopted by everyone everywhere because they will universally promote superior performance. The specific practice choice in the universalistic approach is not well defined, and this approach consists largely of general concepts (for example, extensive training, decentralized decision making, extensive information sharing).

• Contingency: The contingency approach holds that the choice of practice is contingent on the specific situation. This view adopts a best fit approach contingent on factors such as firm size, skill level and tasks of the workforce, labor market conditions, and so on.

• Configurational: The configurational approach consists of a systems approach in which the various HR functions (for example, recruitment, selection, training, compensation) complement one another.

Assumed within these approaches is that the potential performance effect is multiplicative; that is, the more of them you use, the greater the impact on performance. There seems to be general agreement that the “one size fits all” approach is faulty and so, instead, an integrated approach is the most sensible approach.

Consequently, how do we model ideal HR practice choice? According to Kaufman and Miller, management scholars generally consider economic models to be too simplistic,¹³ and economists view management as light on substance and heavy on description and prescription.¹⁴ However, the authors draw from both the management and economic traditions to determine the factors to consider when deciding what practices to put in place that are most closely associated with successful outcomes. More recently, there has been a focus on what has been referred to a “high performance work practices” (HPWP) or practices that engage and motive the workforce. However, it is also recognized that these practices vary by situation.

This model is well suited for explaining (or in our case determining) practice choice across industry, organizational life cycle, firm size, country, and so on and predicts the use of the various practices within a particular setting. The Xi variable is the one of interest to us. This variable is actually a vector (or list) of practices associated with optimal HCM practice choice.

• Firm size: The demand should increase with the size of the organization.

• Wages: If you are paying above market rate, expending effort on maximizing the contribution of human capital is very important.

• Technology: Team production.

In essence, the model uses the following equation when attempting to determine optimal choice:

HRMi=f(Qi,Wi,Xi)

The next question is this: What practice do we use when? The answer to this question will vary considerably depending on the situation you are in. In many cases, the default has been to put in place the same HR policies and practices that a competitor (or an organization that is operating in the same labor market, which may or may not be in the same industry) uses. Because there are few true apples-to-apples situations, it is most efficient to take the time to carefully evaluate your individual situation. This is not to say that knowing a competitor’s policies and practices is not valuable, but the adoption of exactly what they are doing is rarely advisable.

Suppose, for instance, that you are about to open a new plant in a different part of the country, or maybe in another country. You have some data on local pay rates and the kind of practices that the competitors have in place. However, it is a green field site in a business your organization is new to. How do you decide what policies and practices to put into place? You need to abide by existing laws and the other conventions of the region or country, but you still have plenty of latitude to choose policies and practices that you believe will maximize organizational efficiency.

The information necessary to make policy and practice determination is easily handled by tools associated with advanced analytics, but not so easily handled by us. The choice of policy and practices and how they align with the other HR policies and practices can quickly become complicated. This is due to the large number of different HR policies and practices that exist. You have to consider which ones maximize the potential for meeting organizational objectives as well as how they interact with all the other organizational and functional policies and practices in place.

The use of analytics has a long history within HR as well, with many practitioners and academics alike making formidable contributions to the discipline. Much of this work has concentrated on the use of analytics to establish a connection between HR activities and the performance of the firm.

Increasingly, strategy maps are being used for more robust analysis. This is important because to date many of the analytics available only allowed for simple descriptive analytics. More recently developed analytical tools enable you to be much more declarative about whether two variables have a causal relationship. For example, these systems should allow an answer to questions such as this: Has the introduction of a new on-site child-care facility resulted in an increase in employee morale and a decrease in employee turnover? Then, ideally, it would be of further benefit to determine the exact dollar impact the adoption of a child care facility. Are the costs associated with setting up and running an on-site child-care facility more or less than the cost savings associated with reducing employee turnover and absenteeism? Also, are there added benefits associated with the establishment of the on-site child-care facility? For instance, is there a benefit associated with the attraction of potential new employees?

The challenge with decision science is the emphasis has long been on how people should make decisions rather than on how people actually do make decisions. Fast forward to today and sophisticated BI analytics are mostly found within finance and general strategy products. They include such products as IBM’s CFO Performance Dashboard version 3 Advanced Edition and SAS’ Strategy Management.¹⁷ One big advantage of these systems is that they allow for rigorous evaluation of the relationships between variables. IBM’s CFO Performance Dashboard,¹⁸ for instance, provides financial key performance indicators (KPIs), and it integrates IBM’s Cognos BI software and IBM’s SPSS statistical capabilities. This provides a financial intelligence and allows for “what if” analysis and also includes predictive analysis using causal modeling that provides potential outcomes associated with specific business decisions and scenarios.

Although considerable variation exists across industries, on average 70% of the cost of doing business is due to human capital costs. Consequently, the more this resource is optimized, the better; and advanced analytics provides a number of tools that can assist with this. Advanced analytics offers a number of potential ways in which to make better decisions about HR policy and practice choice, including the following:

• Dashboards, scorecards, and strategy maps can be used to better understand relationships between variables and assist in establishing line-of-sight causal relationships between performance outcomes.

• Advanced analytics with predictive capabilities can establish connections between programs and policies and organizational performance outcomes (for example, productivity, profitability, employee turnover, employee morale).

• Q&A expert systems can assist with the determination of optimal HR policy and practice choice.

• Applications can help diagnose problematic outcomes such as employee turnover.

Recent work has established a connection between effective HR management and firm performance.¹⁹

The alignment of HCM practices with business objectives has evolved into how these policies and practices map to organizational objectives. Many of the corporate performance management strategy products display the relationships between various metrics and how they co-vary. This largely takes place though the use of dashboards, scorecards, enterprise metrics frameworks, and the analysis of structured and unstructured data.²⁰ According to Forrester Research, a number of BI vendors offer “consolidated HR analytics solutions.”²¹ The benefit of these sorts of products is that they allow integration across a variety of different enterprise resource planning (ERP) vendors.

• Machine learning

Identify why turnover is taking place.

“Learn” what characteristics are associated with superior performance.

• Predictive modeling

Use machine learning to better predict whom you will need to hire in the future.

• Deep Q&A expert systems

Get advice based on rigorous research rather than one person’s opinion.

• Prescriptive recommendations

What can deep Q&A systems assist us with relative to determination of policy practice and choice?

Again, the policies and practices in place are often a function of tradition or benchmarking. What may have made sense 10 years ago or relative to the plant next to you in a totally different industry and in a different point in the life cycle will almost certainly not be pertinent to your given situation.

Most people doing HCM are too busy with the transactional to have time to engage in analysis. The potential of these systems is substantial. For instance, it is possible to start with an inventory of employee skills to match those with the strategic objectives of the firm. This is what these systems now allow for. Much of the advanced analytics consist of the following three capabilities:

• Forecasting capabilities are used to accurately assess future needed skills. It is also possible to evaluate the needed skills mix using a variety of different scenarios. An example is the potential entrance into new markets requiring a new set of capabilities.

• Predictive modeling allows for an analysis of past events to predict future outcomes and assess both areas of opportunity and risk. For instance, it is possible to identify employees who are at high risk of leaving the organization, allowing time to develop interventions to reduce undesirable turnover.

• Optimization provides a method to determine the ideal allocation of resources (for example, allocation of a bonus pool across employees while keeping an eye on internal and external equity).

A number of specific tools do an outstanding job of helping to make decisions that ultimately impact the success of the organization. One of the most integrated systems is SuccessFactor’sBizX, short for Business Execution Software.

The software offers a complete set of applications, including the following:

• Performance and Goals: Facilitates the communication of individual goals and enables managers and executives to monitor how individuals are progressing on goals and to issue rewards when objectives are met.

• Compensation: Ties performance appraisals and performance management to rewards.

• Recruitment: This application provides a means to track and manage perspective candidates and also provides access to social media and a means of collaborating within the organization to facilitate decision making.

• Learning: Mostly an e-learning or a learning as management solution (LMS).

• Collaboration: This refers to a mobile collaborative device providing a mechanism to assist with decision making and information sharing.

• Workforce Planning: Allows for forecasting the impact of a variety of strategies.

• Employee Central: A user-friendly HR self-service data center.

• Workforce Analytics and Reporting: Provides actionable intelligence to decision makers.

A number of significant factors are currently impacting ERP software systems.²² One factor is that more and more organizations are moving from an on-premise application of cloud-based software as service (SaaS) or platform as a service (PaaS) model. There is also a movement toward including a broader range of functionalities. Standalone applications such as applicant tracking systems are now integrated in with e-recruitment software, which are further integrated with workplace planning software, compensation, corporate leaning, collaboration systems, and so on. These combinations are called talent management suites and offer the advantage of integration of information. Finally, there will be further development of mobile apps and interaction with external data (for example, with social media).

That may well be the mission of Amazon when their data analytics recommend a book that we would never have chosen on our own but end up loving. This sort of computation logic allows for not only utilization of historical time series analysis but is now able to further interact with real-time data as well.²³

• Talent Acquisition: Their system provides an analysis of costs and the time it will take to acquire talent. It also analyzes the current pool of talent and how accurate its source is (for example, executive search firm).

• Succession Planning: Tracks current employees to identify and fill vacant positions.

• Talent Retention: Tracks the retention of employees.

• Talent Development: Measures costs and effectiveness of training programs on the skills and development of employees and how well they meet organizational goals.

SAS HCM software includes a number of useful tools that enable forecasting, predictive modeling, and optimization. One feature allows for the identification of top-performing employees who are at risk of leaving. In addition, SAS offer both time series and structural equation modeling. This provides significant help with establishing cause and effect between variables. Other advantages include the following:

• The system provides prepackaged metrics that provide a view of metrics such as revenue per employee and how close it is to established goals.

• Allows for “what if” analysis, providing a means of better anticipating a variety of workforce planning scenarios.

• Integrated in with the other SAS solutions and so can be used to see where the organization is relative to goals.

• Establishment of a link between strategy and execution. It does this by tracking KPIs.

• Establishment of cause and effect relationships (one of its biggest benefits). This is possible through evaluating variation between KPIs and goals.

• Alignment. One key benefit of these systems is that you can customize the metrics to your situation and thus allow for alignment of practices with strategy.

• Determination of potential challenges and opportunities through the use of alerts.

According to Kaplan and Norton:

Executives in all sectors and in all parts of the world were facing the dual challenges of how to mobilize their human capital and information resources.²⁶

According to Gartner, a large portion of these systems were used for reporting. More recently, as I have already discussed, scorecards and dashboards have been added. However, these too can often be used primarily in a descriptive manner, rather than for predictive purposes.

Integrated systems such as these offer a number of advantages. There are advantages associated with having all the information in one place, and, of course, there are cost efficiencies associated with such integrated systems.

Increasingly, HCM software has consolidated. So, what were once independent functions are now all included in one suite.²⁸ These suites include some or all of the following functions:

• Workforce planning

• Talent acquisition

• Compensation

• Performance management

• Career development

• Succession planning

• Corporate learning

In addition, the inclusion of the following functions would prove valuable:

• Integration with broader ERP (for example, finance and operations)

• Social networking

• Collaborative decision-making software