Navigation SearchNavigation ContactNavigation Products
Leader Pulse
Leadership ideas, trends, and smarts

For Downstream HR Analytics Impact, Take an Upstream View

By Rich Wellins, Ph.D. and Evan Sinar, Ph.D.

Richard S. Wellins, Ph.D.

Business analytics (BA) is a set of skills, technologies and processes used by organizations to gain deeper and more actionable insights into their business. Though the visibility of BA has spiked in recent years, it has recently surged in visibility. These practices date back to 1865, when Richard Millar Devens coined the term “business intelligence” to describe how a banker, Sir Henry Furnese, increased profit by analyzing and acting on business about his environment and as a result, gained a competitive advantage. Moving the clock forward, modern business analytics popularity can be attributed to thought leaders such as Howard Dresner, a past Gartner Research Analyst, and Tom Davenport, Babson College, who continues to write prolifically on the subject.

Examples of BA prevail in everyday company decisions, from the price hotels set for rooms, which change constantly, to driving the books Amazon recommends to you, based not on a hunch but rather, on analytics predicting the likelihood you will “add to cart”—and through evidence-based healthcare, big data predicting clinical outcomes based on choice of treatment, hospital, and physicians. And, an entirely new analysis application has emerged around DNA. TIME’s article, How Blood Tests Are Changing Medicine, July 2015, told us what a single drop of blood could tell us about a person.

What are leadership analytics?

Business analytics has recently (and finally) come of age in the HR arena. When it comes to leadership, our main focus of this piece, DNA research offers a great analogy. What can the “DNA” of current or potential leaders tell us? Will they be great? Mediocre? Or fail miserably? In which situations will they succeed or struggle? And, how will their behaviors shape their companies’ performance?

Just imagine if you could:

  • Determine which early life experiences predict later leadership excellence (e.g., international experience? Non-traditional degree paths? Membership in school/college clubs?)
  • Hone in on the leadership behaviors driving employee engagement?
  • Confirm you have talent in the right roles and locations to meet three-year growth targets and if not, what skills are needed to close the gap?
  • Anticipate and proactively support leaders at risk of turnover?
  • Increase the “hit rate” for high potential leader identification?
  • Optimize the learning environment for faster skill acquisition and growth through manager reinforcement, job/business relevance, and application opportunities?

The reality is all these questions and many more are now answerable. A growing number of companies are busy building teams of financial analysts, psychologists, and economists to collect and mine their data—all aimed at making better leadership talent decisions. These progressive companies are still far rarer than their counterparts; however, for the majority of companies, the lack of progress is appalling. Data from the multinational version of DDI and The Conference Board's Global Leadership Forecast 2014|2015, shows that 41 percent of surveyed companies aren’t currently doing any form of leader-focused analytics well. And, only six percent are effective across a key set of analytic metrics.

Given the potential of deep analytics to drive both the effectiveness and perception of the HR function, rapid movement is critical. The Harvard Business Review cover story, “It’s Time to Blow Up HR” (July/August, 2015), not a gentle title, spotlights the importance of HR analytics and how, as a profession, we need to step it up.

That’s the bad news. But, there is hope on the horizon. Industry analyst Josh Bersin notes that progress toward the planning, correlation and prediction facets of people analytics has accelerated dramatically between 2015 and 2016, with the proportion of companies excelling in relating people data to business performance and using people data to predict performance more than doubling in just one year.

Making the most of leadership analytics

The Leadership Analytics process is about exploring relationships among various pieces of data aimed at answering your burning questions, similar to the examples we gave earlier. Based on these questions, you can planfully mine your database, supplementing with new or third-party data as needed, to find the answers. Of course, your success will depend on having an accurate and robust database in the first place. In many cases, you may need to revert back to data collection to tell a complete and coherent story.

Certain categories of leadership data, when used together, follow a logical path connecting a leader’s background, with experiences, and subsequently to outcomes. The first category contains “upstream” data about a leader’s background, work experiences, development opportunities, and so on. This category of data—often undervalued in its role in enabling analytical results to be translated into responsive action—can be extremely important because of its link to downstream placement and promotion decisions, allowing you to spot and guide existing talent early enough to matter, and to target your search for new target matching the most predictive profiles. The second category of “midstream” data is around the personal performance of the leader. The final category of data is around “downstream” company or business unit outcomes. The table below shows examples of data in each category:

Data Category Examples
Upstream:
Personal Background/Experience
Formal education
Company training
Work experiences and assignments
Scores for pre-hire assessment tools (e.g., tests, assessment centers)
Roles and/or job functions held
Midstream:
Individual Leadership Performance
Personal engagement
Demonstrated skill acquisition
Performance evaluations
Number of promotions
Entry and success in a high potential program
Scores from post-hire or developmental assessments (e.g., 360 feedback tools)
Downstream:
Business Outcomes
Team engagement
Innovation throughput
Employee retention
Revenue growth
Profitability
Productivity

Once this logical data structure and sequence are in place, you can begin exploring relationships. Let’s consider an example: Assume you want to explore the relationship of leadership quality and business unit growth. You might start by exploring the relationship between certain combinations of training and job experiences that lead to improvement across a set of five to six key leadership behaviors. In turn, you could find that those leaders with stronger behaviors positively influence team-level engagement. Finally, you show that higher levels of engagement drive higher financial performance to provide the business-scale payoff of your leadership development initiatives.

In a second example, you might want to predict the likelihood of predicting the success of three new retail stores you are about to open. Upstream and midstream analyses paint a clear picture of the types of store managers who have been successful in your other nationwide stores. Using well-matched assessment tools and recruiting strategies, you can now promote or select those store managers who are more likely to hit their (and your) numbers.

Not all forms of analytics are created equal

In a previous DDI article (GPS for Talent Management Analytics), we outlined a framework to understand the growing field of talent management analytics. There is often confusion (and occasional snobbery) when it comes to defining what really constitutes “analytics.” We take the view that almost any form of data can fall under the analytic banner—IF it contributes to the end-to-end prediction model described above. The value is derived from how you use the data. Higher value is derived from the earlier examples: Predicting which leaders, in which situations, stay longer, engage better, and hit their performance targets harder. This prediction path starts with early background and experience information—without which companies will be limited in not just understanding the linkages, but influencing and improving them.

Determining your analytic questions depends largely on your strategies and appetite for data-driven decision making. However, we do have two pieces of advice. First, view your analytics quest as exploratory rather than fixed. A fixed approach might start by exploring the relationship between leaders who are adept at influencing and business unit revenue growth. You may or may not find a relationship. An exploratory approach might attempt to look at the relationship between a range of leadership skills with an array of outcome measures. You might end up with surprises, but nonetheless, useful insights. One of the authors of this article was talking with the head of a major law firm association exploring the relationships between a number of variables and promotion to partnerships. They expected to find things like prestige of school or serving on the colleague’s law review publications. Neither mattered much. What did drive promotion was whether or not the student had a job during college and law school.

Finally, while we hesitate to discount any form of analytics, our own recent research (Global Leadership Forecast 2014|2015) showed that the analytics used most by multinational organizations were least related to financial success. That is, the lesser used (and harder to capture) analytics had the most impact (see the graphic below). Overall, financially outperforming companies were six to nine times more likely than their lower performing counterparts to deploy advanced leadership analytics.

The Misguided Flow of Leadership Analytics

Clearly, existing analytical models that are akin to writing a novel by beginning with the last chapter aren’t enough. These approaches start the story at the final link between a leader’s performance and a business outcome, yet expect the readers—business decision-makers in this case—to still follow along with, care about, and act on what ultimately happens. These models produce temporary satisfaction but do little to instill or incite sustained interest and action. Starting the analytics sequence much earlier, far upstream of the end state, is vital for understanding not only why the story ended the way it did but also how to instead rewrite it toward a happier ending.

In this article, we’ve covered the What and Why of Leadership Analytics—but what about the How? In Part II, we explore the readiness of HR professionals to conquer the world of analytics.

Rich Wellins, Ph.D., is senior vice president at DDI and coauthor of Your First Leadership Job: How Catalyst Leaders Bring Out the Best in Others. He is passionate about helping organizations employ alignment and analytics to realize the potential of their leadership capability.

Evan Sinar, Ph.D. is DDI’s Chief Scientist and director of DDI’s Center for Analytics and Behavioral Research (CABER). Evan is a thought leader on talent management analytics and data visualization.

Posted: 22 Jul, 2016,
Talk to an Expert: For Downstream HR Analytics Impact, Take an Upstream View
* Denotes required field
 *
 *
 *
 *
 *
 *
 *
 *
Consent to DDI Marketing *

I consent to DDI collecting and processing my personal data in the provision of services to me and for the purposes of marketing and research. I am aware of my rights and the ways in which my data will be used as referenced in DDI’s Data Privacy Policy. I am aware I have the right to revoke this consent at any time.

Please enter the number from this image
 Security code