Opportunity is missed by most people because it is dressed in overalls and looks like work.

—Thomas Edison

People Analytics is a growing area of data science that holds great promise for the future. It’s likely to become a commonplace practice in the coming years as the people, processes, and systems mature and make People Analytics something that any organization can do. The field of People Analytics as a discipline at the intersection of business analytics and HR is still in relative infancy, and technology is finally beginning to catch up to fulfill the promise for what organizational leaders have wanted for many years. For example, some companies such as Starbucks, Limited Brands, and Best Buy can precisely identify the value of a 0.1 percent increase in employee engagement at a particular store. At Best Buy, for example, that value is more than $100,000 in the store’s annual operating income.¹

It’s true that advanced analytic techniques of employee data that companies have at their disposal can help answer some of the critical questions surrounding the value of human capital, such as: How do investments in employee programs actually impact workforce performance? How can you motivate employees to succeed? and Who are your top performers? However, most companies are not currently able to answer these questions; they struggle with answering even basic workforce analytics questions, much less questions having to do with more sophisticated techniques. Futhermore, we believe the use of advanced People Analytics is intensifying and will soon become the new normal for businesses.

One day soon, people will stop discussing the merits of the emergence of People Analytics, which companies are using advanced People Analytics to optimize their human capital spend, or whether People Analytics is critical to their business. It will just be a given that every company must cope with the stream of Big Data from workers, and must have a People Analytics strategy, as well as use various data assets and tools to augment the data they collect internally. In other words, the formal use of People Analytics within HR and across business units will become as ubiquitous as data itself has become. In the same way that most companies have strategies for learning and development, onboarding, training, and resource planning, they will also have a formal strategy for People Analytics.

A wide range of trends are only just beginning to pave the way for advanced People Analytics, including improved technology, machine learning techniques, and data visualization. These trends are making applications of computational People Analytics possible that were not possible even five years ago. In this chapter we will discuss some of those, as well as other trends that will influence the future state of People Analytics.

We believe the future of advanced People Analytics is strong and that the future holds several key trends related to People Analytics, each of which we will discuss in the chapter. Specifically, we think that in a People Analytics future:

• There is a rise of employee behavioral data.
• People Analytics moves beyond the averages.
• Predictive analytics becomes the new standard.
• Big Data analytics becomes automated.
• Big Data empowers employee development.
• Models become the gold of People Analytics.
• People Analytics becomes more accessible to the nonanalyst.
• People Analytics becomes a specialized department.
• Employee data privacy becomes top of mind.
• There is quantification of HR.

RISE OF EMPLOYEE BEHAVIORAL DATA

Another future trend related to People Analytics will be the proliferation of various forms of employee data that can be tracked, analyzed, and modeled. The intersection of computer science, information technology, and psychology of how computers can effectively interact with humans is driving part of this new data on employee behavior. We are regularly gaining new knowledge on how humans can effectively interact with technology, as well as the tools to create that knowledge, which is being applied to the workplace more regularly. One of the oldest examples of this is GPS-based technology. GPS allows for the tracking of location and movement, and saw widespread usage initially in the shipping and logistics industry, where companies wanted to make sure drivers and packages arrived on time. And its use is on the rise. A 2014 study by the research firm Aberdeen found that 54 percent of companies that send employees out on service calls use some sort of location-based tracking system—and this is up from 37 percent found in 2012.²

What about employee behavioral data of the future? What types of technology are on the horizon to enable employees to be more successful and employers to use People Analytics to model successful outcomes? One example is eye-tracking technology. It’s one form of human-centered computing whereby a camera can track the eye movement of someone viewing a computer screen, recording data about where and how often the person viewed different areas of the screen. It has been applied in areas as diverse as website usability testing, sports medicine, automobile testing, geriatric research, training simulators, and infant research. Some employers are already using it now to monitor whether employees are paying attention during training sessions.

As we look into the future, we expect more devices will interact directly with the human body and, as a result, generate data from those interactions in the workplace that needs to be analyzed. In terms of the future applications of human-centered computing, we think the following applications will be relevant for the field of People Analytics:

• Wearable computing/smart fabrics
• Consumer health informatics
• Brainwave measurement
• Facial recognition
• Emotional recognition
• Exercise informatics
• Body scan technologies
• Gesture-based interfaces
• Motion-detection devices
• Molecular computing

Furthermore, many of the hundreds of millions of mobile apps include some type of location-based tracking. Also, stores are experimenting with location-aware services that enable retailers to track and serve offers to users in a specific location. This same technology could be used to interact with employees based on their locations.

Certain companies are already creating multidimensional employee behavior tracking tools that are being used by some organizations. For example, imagine a tiny microphone embedded in the ID badge hanging from the lanyard around your neck. The microphone is gauging the tone of your voice and how frequently you are contributing in meetings. Hidden accelerometers measure your body language and track how often you push away from your desk. At the end of each day, the badge will have collected several gigabytes’ worth of data about your office behavior. Sound like science fiction? It’s not, and is being touted as the next frontier in office innovation by Boston-based creators Humanyze, which developed a badge that tracks employees’ daily movements, how much they speak up during meetings, and whether they need a break or are going strong before storing nearly four gigabytes in data for employers to analyze at the end of the day. The company believes information like this will help with productivity in the office.³

As organizations begin to leverage employee behavioral data of different types, People Analytics will need to cope with policies about how to store, sort, analyze, and use this data for the benefit of organizations while respecting the privacy rights of employees.

PEOPLE ANALYTICS MOVES BEYOND THE AVERAGES

Analytics of any kind relies on describing patterns in the data. One such example of a metric that reflects the pattern in data is the “average,” which is something most people learn during their early math education. To calculate an average, take all the data you have about something and divide the sum total into equal parts. Average is used frequently in People Analytics. For example, average cost per hire, average speed to answer, average time to productivity, average customer resolution rate, average talk time of a sales rep, and average hours worked are common metrics used to evaluate worker performance in organizations.

However, often the average doesn’t do a great job at helping us understand the individual variations and nuances in your workforce. For example, if you have a customer service team of 100 and half of them have a very high talk time when helping customers and half have a very low talk time, calculating an average won’t be very insightful. You’ll end up with a metric that doesn’t describe anyone in either group and, if used, would give a false sense of failure or success as you moved forward with initiatives to drive this metric.

A future trend in People Analytics will be to move beyond the averages and create analytical models that describe the complexity of the workforce. We think People Analytics will eventually move toward analytical models that help optimize each worker’s performance such that analytical insights will help each person perform as best he or she can in a chosen career. Rather than being compared against a large population, the practical use of People Analytics will be to personalize our understanding of the unique qualities of each person, helping provide insights that benefit both the employer and the worker.

For example, imagine that we create an employee retention model that scores employees in terms of how at risk they are to leave the organization in the next 90 days. We might use metrics in the model such as absenteeism rate, hours worked, start time, projects delivered, and calls made, among others, to attempt to predict whether the worker is at risk of leaving the organization. If we used a simple average comparison of a specific person across an entire workforce, we might get a lot of false positives and think someone will leave when they won’t, as well as fail to detect all those who are truly at risk of leaving. However, if we go beyond the averages and use, say, a ratio of the change in performance over time for each employee, we are much more likely to account for individual differences in how employees do their jobs. As a result, the key model inputs will be changes in how each employee works as a predictor of likelihood to quit, rather than evaluating how that employee compares to an average.

In the future, People Analytics will go beyond the averages and our models will be able to predict worker dynamics at an increasingly more granular level—even down to the level of a specific worker.

PREDICTIVE BECOMES THE NEW STANDARD

We provided overviews of predictive analytics and its importance throughout this book. We believe that as we progress toward the future, predictive analytics will become more widespread and evolve into the norm for all analytics. People-related analytical techniques will need to have a predictive component in order to be considered business-relevant or effective, not metrics to describe merely “what has happened,” but to help describe “what will happen.” This will require more sophisticated statistical techniques, data integrations, and more computational power, all of which are becoming possible. It will also require the expertise of analysts and HR that can develop people-related predictive models effectively and understand how to learn, test, and optimize using predictive analytical techniques. There are certainly many examples of predictive analytics applied to human capital, many of which you have seen illustrated in this book, and we expect to see it become the standard across all industries in the future.

AUTOMATED BIG DATA ANALYTICS

Machine learning is a branch of artificial intelligence concerned with the design and development of algorithms that allow computers to learn from processing real data and to become more proficient over time. In the future, artificial intelligence will start showing up in more and more unexpected places, including the software used by most employers. A major focus of machine learning research is to automatically learn to recognize complex patterns in Big Data and make intelligent decisions. Some examples of the current applications of machine learning include:

• Search engines
• Medical diagnosis
• Bioinformatics
• Cheminformatics
• Detecting credit card fraud
• Stock market analysis
• Classifying DNA sequences
• Speech and handwriting recognition
• Robot locomotion
• Aircraft autopilot
• Computational finance
• Sentiment analysis
• Recommender systems

Examples of automated Big Data analytics applied to employee data already exist. For example, Kanjoya, Inc. has developed sentiment analytics software that automatically sifts through thousands of open-ended employee comments to understand whether employees are satisfied or frustrated. For companies with thousands of employees in multiple locations, automated analytics like this can really improve productivity, letting leaders focus on changes to enhance employee engagement, rather than sift through data to understand it. Kanjoya’s automated analytics, developed in collaboration with scientists at Stanford University, can also uncover subtle differences in opinions, attitudes, and sentiment in written conversations and can detect the earliest signs of workplace bullying, harassment, and discrimination.⁴

Machine learning is a rapidly evolving field that has the potential to have a great impact on People Analytics. However, what is not clear is how the future advances in machine learning will impact the need for trained analysts and other human resources. As machine learning models and techniques improve over the long term, it is possible there will be a reduction in demand for humans with that specialized skill set. Additionally, as sophisticated applications are developed, it will make it easier to run larger, more complex organizations with fewer people, possibly leading to corporate consolidation and the ability to do more with a smaller workforce.

BIG DATA EMPOWERS EMPLOYEE DEVELOPMENT

The days of the stressful annual performance review involving a big buildup of activity to get a long document written and delivered are starting to change. Going away are the days when feedback is given in large mega-doses that don’t always lead to meaningful development. The availability of Big Data analytics is a key driver of this shift.

The broad trend is that companies will forgo the annual review process and traditional engagement surveys and replace them with a more flexible and ongoing feedback system enabled by Big Data management by apps. These systems will allow employees to give and receive feedback anytime throughout the year on a variety of issues. The new surveys would be shorter and more pointed than the long annual engagement survey. They’ll also be designed for smartphones so they can be taken anywhere at any time. Companies like Adobe, Accenture, Starbucks, and General Electric have moved away from the old type of annual performance review to more open and ongoing feedback. Some of these companies are running their employee performance management by an app, from which employees can get feedback from their managers and peers and also communicate with them.

According to the head of human resources (HR) at Accenture, “you know the world is no longer working in yearlong cycles, so to set objectives at the beginning of the year and revisit them at the end of the year simply to see how people are doing is not really relevant any longer. Doing away with the annual review cycle really makes sense for us when you consider that two-thirds of our employees are millennials, who are used to giving and receiving constant feedback in their daily lives, whether it be through Instagram likes or writing and reading reviews on Amazon.”⁵

Through leveraging Big Data analytics, companies will be able to analyze and identify, in real time, early signs of employee retention or morale issues, providing a closed-loop response to employees that empowers them and makes them heard rather than waiting around for the next annual performance review or employee survey.

MODELS BECOME THE NEW GOLD OF PEOPLE ANALYTICS

This may seem counterintuitive, but we believe data will become less valuable in the future. Even now, data are everywhere and people and organizations are overwhelmed with data. In the future, having a treasure trove of data about your employees will not, by itself, hold much value. However, we do expect analytical models to become more valuable at the same time the data by itself becomes less valuable. In other words, the companies that have the ability to create actionable knowledge-based tools from employee data, either using their own data or using someone else’s data, will see the business benefits. This will take many forms: everything from applications that sift through employee data for nuggets of productivity insight to personalized algorithms that allow an individual employee to analyze his or her own behavioral performance data to help understand something about themselves and take action.

PEOPLE ANALYTICS BECOMES MORE ACCESSIBLE

Another trend that will shape the future of People Analytics is that analytical techniques will become more accessible to the general business user, enabling nonanalysts and people outside of HR to take an analytical approach to employee and human capital challenges. As general knowledge of analytics spreads and software providers automate the use of techniques (e.g., hide the actual steps of data analytics from the user, such as data modeling, text analytics, web analytics, and segmentation through automation), people with little or no analytical background will be able to run models and take business action from the results. We can already see instances of this occurring in certain analytical disciplines. For example, the rise of automated online survey tools over the past 10 years has led to people in all departments of organizations creating and analyzing their own customer surveys. This sometimes causes frustration for the marketing research experts, as survey questions are sometimes poorly worded or statistically significant differences or margins of error are not considered. However, we do believe this trend will be net positive for the influence of analytics, but we’re sure there will be pain along the way. For example, untrained analysts are likely to apply analytical techniques improperly or in an inappropriate manner with incorrectly prepared data. As a result, there will be instances of confusion and frustration as those who know People Analytics help those who do not. However, we believe the social pressure that will result will eventually lead to People Analytics being used more effectively, not only by HR, but across the organization.

PEOPLE ANALYTICS BECOMES A SPECIALIZED DEPARTMENT

Although we expect that People Analytics techniques will become more available to nonanalysts across the enterprise, we also see the future of People Analytics being one where specialized departments are created to address the needs of People Analytics across the organization. In the current state, analytical professionals are often fragmented across the enterprise, frequently in different departments with labels such as business intelligence, statistician, survey researcher, web analyst, data scientist, and HR analyst. Furthermore, most companies have very few, if any, professionals who specialize in people-related analytics. We expect that most companies will move to centralize People Analytics expertise in a specialized department either in HR or in finance. The model for some companies may be to have People Analytics as a formal, centralized, shared service, and for others, as a center of excellence where People Analytics professionals are kept close to the business units they serve, yet have accountability to and participate in an analytics center of excellence. Regardless, the most successful companies of the future will recognize and prioritize the importance of People Analytics and related analytics professionals, making them into a formal business function in the same way that it is commonplace in many companies today to have such departments as engineering, marketing, customer service, technology, and finance.

EMPLOYEE DATA PRIVACY BACKLASH

Employee data privacy will be a hot topic in the future of People Analytics as employees continue to grapple with the notion that many of their activities are being tracked in great detail by their employer, both while they are at work and while they are at home if they’re using work-provided equipment. Also, as analytics become more sophisticated and human-like, employees may get an uncomfortable feeling when they are given workplace insights and recommendations they didn’t even know were possible. Recall the case we reviewed in Chapter 6 of Xerox’s insight that customer service rep success is due, in part, to personality traits.

Already employee data privacy is a hot topic with many companies grappling with how best to use employee data. One position is that analytics software should be applied to data only where an employee has no expectation of privacy. For example, Intel is very thoughtful about only monitoring employees’ sentiment through their communications in the workplace where privacy would never be assumed. “We’re only going to do it where it’s very clearly been an employee statement in a known public forum where they know their stuff is being looked at,” says Intel’s Richard Taylor. On internal Intel blogs, for example, employees must attach their real names to comments. In Intel’s eyes, such posts would be acceptable for analysis. However, employee e-mails are seen as private and Intel won’t analyze them. “We would lose the trust of our employees if we did that,” he says. “That would be the worst thing” with this issue.⁶

The global economy will make the employee data privacy backlash even more complicated for multinational firms as different countries develop different standards regarding what is acceptable use and acceptable privacy rights of employee data. For example, an October 2015 ruling by the European Court of Justice makes it difficult to transfer employee data about a German worker outside of Germany without express permission. If the ruling stands, it may force all organizations that currently rely on the ability to access the data of their European Union (EU) partners and subsidiaries to seek alternate modes of data transfer or risk legal liability from being in noncompliance with EU data protection requirements.⁷

We expect that, as the future progresses and as data and People Analytics become more important in our workplace lives, there will be a data privacy backlash where employees will demand more awareness about what is being tracked and even more governmental involvement in employee data privacy standards and protection in the same way that Equal Employment Opportunity Commission guidelines have become an important role of the government. We do not know what form this will take, whether it will be a credit agency model whereby all information is centralized in a few organizations or it will be distributed across each employer whereby employees have the ability to decide which information about them is shared, hidden, or permanently deleted. Either way, the field of People Analytics must take note and engage in the conversation, as employer–employee dynamics will be directly impacted by any changes in data privacy policies and standards.

QUANTIFICATION OF HR

Another trend that will affect the future of People Analytics is what we call the quantification of HR. Those of you in HR leadership know you’re under pressure: under pressure to quantify such things as people-related programs, demonstrate the ROI of training and development initiatives, and show efficient use workforce talent. This all takes hard data—data that you may not have, data that may be difficult to understand, and data that may be confusing to analyze. However, the CEO, CFO, and shareholders will continue to put pressure on the return from HR and how your organization knows it’s getting the most from its people assets. Therefore, the future of HR looks a lot more like a quantitative discipline than just a people-friendly discipline.

So far, many organizations are already behind. The Sierra-Cedar HR Systems Survey, now in its seventeenth year, gathers information from more than 1,000 organizations across the globe that is validated against publicly available financial and market data. The longevity of the HR Systems Survey affords a historical perspective that allows a look back year over year at factors that have an impact on business outcomes. As part of that survey, Sierra tracks the adoption and deployment of HR analytics solutions, gathering data on process maturity as well as the type and amount of data HR organizations are capturing.

The results are surprising. Although many companies used HR dashboards and reports of some kind, very few used true analytics, with workforce analytics used by only 16 percent, predictive analytics used by only 9 percent, strategic workforce planning analytics used by only 11 percent, and Big Data analytics used by only 9 percent.⁸

However, because Sierra is tracking business outcomes with its study, the analysts were able to compare the business outcomes of HR organizations that were data driven versus those that were not. They wanted to see whether organizations that gather more data, share that data openly, and leverage it in processes and decision making see improved outcomes from their efforts.

First, they had to define the quantified HR organization. Starting with key findings from the 2013–2014 survey results about business intelligence (BI) adoption, they determined that a quantified HR organization would be one that invests in HR measurement technologies, processes, and practices that enable it to improve workforce operations and achieve organizational goals in an environment of data-driven decision making. Specifically, they identified several selection criteria that were based on an organization’s leadership in four areas:

1. Business intelligence process maturity.
2. Managers’ direct access to HR analytics and BI that supports workforce decision making.
3. More data sources regularly juxtaposed with workforce data.
4. More overall categories of HR metrics.

Sierra found that company size didn’t matter, with successful companies in this area ranging in size from small (with workforces of just over 100) to very large (with workforces of more than 400,000). Many quantified HR organizations were also global, operating in an average of 29 different countries.⁹

The findings were clear. Quantified HR organizations outperformed even the top-performing organizations of the most recent Sierra study, achieving higher levels of financial performance, as well as positive HR and talent outcomes. However, the most dramatic difference in outcome analysis was in the return on equity.

There is more work to understand exactly how the quantified HR organization achieves results, as surely many factors have an impact on an organization’s overall success. However, there is enough data from multiple sources to conclude that the data-driven HR organization is real, and that it isn’t tied to a single type of technology, organizational makeup, or industry. Becoming an analytically driven HR function is achievable for any organization willing to take an honest look at its data and analytics and use that information to make workforce decisions.