HR Analytics for Competitive Advantage, On and Off the Field
By Keri Pearlson, Jun 03, 2016
Managing talent analytics is not a new concept. Tom Davenport, Jeanne Harris and Jeremy Shapiro published an article in Harvard Business Review in October 2010 on this very topic. Their article, Competing on Talent Analytics, suggested six types of talent analytics: Human-capital facts, analytics HR, human-capital investment analysis, workforce forecasts, talent value modeling and talent supply chain.
But more recently, talent analytics is moving front and center in many organizations. Spurred on by the famous movie Moneyball, many sports teams now have found ways to use talent analytics to optimize their roster of players and their playbooks. One team to do so is the Texas Rangers baseball team, in Arlington, Texas.
Best Practices Institute recently held a talent analytics workshop at Globe Life Park in Arlington where the Texas Rangers play, and the agenda included both Rangers business and analytics leaders as well as professionals outside the MLB world. The event, Talentball, was a 2-day workshop designed to bring the power of sabermetrics (baseball analytics) used in professional baseball teams to HR organizations of the participants. I was a facilitator of some of the small group discussions. Everyone who participated left with actionable insights on improving their talent analytics.
To set the stage, let me share a short overview of the program. There were three main components of the workshop: A series of presentations by Haig Nalbantian, Senior Partner and Cofounder of Mercer’s Workforce Sciences Institute, a series of discussions from key leaders in the Texas Ranger organization, one of the most successful organizations to utilize sabermetrics, and a series of small group discussions focused on applying talent analytics. The journey baseball has taken to engineer this transition to more analytics based decision making in such a short amount of time has produced lessons that may help speed the development of advanced workforce analytics in other industries.
Here are some of the insights I brought back:
Building a robust model that adds value to HR and talent management is more than just identifying the business questions to answer. Tools and techniques from predictive and prescriptive analytics are useful. But a key insight is that it’s necessary to find ways to isolate both the level of analysis and key demographics that influence the business question. To do that requires knowledge of the employee population and the surrounding community.
Sabermetrics has grown from a fringe practice of a few visionary teams to a mainstream management practice. Every team now has an in-house analytics function. Analytics inform most significant baseball decisions such as player acquisition and development, contract negotiations, and in-game strategies. This has allowed small-market teams to successfully compete with larger market teams with much bigger budgets.
Baseball clubs view players as investments, given the high salaries they are forced to pay to get players. This development has transformed the way baseball teams look at player salaries and increased the consideration of a likely ROI of taking on the player. Player contribution and performance metrics use newer, more refined measures like Wins Above Replacement (WAR), Value Over Replacement Player (VORP) and Total Pitcher Index (TPI). These types of measures provide a way to evaluate players across teams and from very different team situations.
Three key challenges that workforce analytics can address are gauging the true contribution of employees individually and in groups (departments or teams), determining the right level of analysis to see true contribution, and optimizing in-game strategies (employee selection versus employee behaviors).
Because of the role of situational factors, the true contribution of individuals or groups may not be accurately captured by absolute measures. For example, by ‘old standards,’ Felix Hernandez, from Seattle, was among the worst starting pitchers contending for the 2010 Cy Young award. But considering the hand he was dealt, such as the lack of run support from his team, the Baseball Writer’s Association of America voted him the best pitcher in the league by a wide margin. His WAR score was 7.1 and the next closest competitor, Clay Bucholtz from Boston, had a WAR of a 5.6.
Individual performance assessment can be contaminated by environmental factors. In one case study, centralized performance management practices evaluated using traditional measures showed one set of ‘leading managers’, but when environmental conditions and market dynamics were taken into account, a different set of managers rose to the top performers’s list. Because of the susceptibility to random environment factors, absolute or “raw” performance measures often camouflage the actual performance of individuals and groups in organizations. Modeling techniques enable comparisons of ‘expected’ and ‘actual’ performance, which is a much better way to understand contribution. And failure to differentiate between performance and risk leads to high compensation costs, lower engagement, and higher turnover.
Modern organizations are exploiting ‘big data’ to develop two types of facts critical to human capital decisions: Dashboards are enabling counts, rates, and tabulations common from business intelligence software that are useful in reporting, tracking and responding to queries about the workforce. Deep-dive analytics use statistical modeling joined with expertise to prove inferences about cause and effect relationships useful for strategy, forecasting and problem solving. It’s no longer about anecdotes and more about predictive modeling in the modern workforce planning practices.
In one study done by Mercer, comparing what “employees said” to “what they do” revealed very different conclusions about workforce turnover behavior. Focus groups, HR interviews, employee surveys, and comparative patters in databases suggested a number of ways to reduce turnover. These methods suggested that pay and workload were most critical to employee commitment and retention. But analysis of actual behavior, using advanced analytics, showed that career development and management stability were the most impactful.
Every employer has unique internal labor market dynamics based on three highly inter-related labor flows: career levels, promotions and lateral moves, and exits. Analytics allows evaluation of the key factors that influence these labor flows, and allows isolation of the key variables that directly impact the outcome. For example, in one study, the value of tenure for front-line workers was found to drop below employee costs after about 10 years of service, but this was not apparent in the raw productivity numbers since these workers got better and more inherently productive routes, masking their real productivity contribution.
Human capital factors affect performance through selection and behavior of team members. Business impact modeling at Mercer revealed significant value in firm-specific experience and reward practices. For example, in a healthcare company, managers used part-time workers to fill in as a way to manage workloads and budgets. But analytics revealed that this strategy was actually causing productivity losses of about 3% of the annual revenue of the organization because of the extra work that these part-time workers had to do to compare with the full-time workers. This was a cautionary tale for companies thinking of using this workforce strategy as a way to manage the requirements of the Affordable Care Act.
The Texas Rangers use analytics in many of their business processes outside of HR, such as in ticket sales, where they monitor, analyze and predict ticket sales for each game, and make key marketing and sales decisions based on what their models suggest. Tools like Tableau provide visualizations of model results which helps make the insights easier to understand.
HR leaders can learn a lot from the experiences of MLB teams and sabermetrics. Teams who successfully use these tools have been able to field better teams, recognize true contributors, identify top performing teams, and understand the really contributing factors to better performance. Those are the same goals our organizations have. Ultimately, it’s about bringing the right people to the right roles and the right time—for baseball teams and corporate teams.
About the author
Dr. Keri E. Pearlson is an expert in the area of managing and using information. She has worked with CIOs and executives from some of the largest corporations in the world. She has expertise in helping executives create strategies to become Web 2.0-enabled enterprises, designing and delivering executive leadership programs, and managing multi-client programs on issues of interest to senior executives of information systems. Keri specializes in helping IT executives prepare to participate in the strategy formulation processes with their executive peers. Current issues include web2.0/3.0 strategy, ecoinformation systems, finding additional value from current investments, project vulnerability analysis, and succession preparation. She’s a skilled relationship manager and an accomplished meeting facilitator. She’s the Founding Partner and President of KP Partners, a CIO advisory services firm.
Keri has held various positions in academia and industry. As Vice President-Leadership Development for nGenera (formerly the Concours Group), she designed and delivered executive-level workshops for CIOs and their direct reports, and she led research programs on issues of importance to CIOs. She was a research and program director at the Research Board, a small, private think tank for CIOs, from 2001- 2003. From 1992-2000, she was a member of the information systems faculty at the Graduate School of Business at the University of Texas at Austin where she taught management information systems courses to MBAs and executives. Keri was also a research affiliate with CSC-Research Services where she conducted a study of the design and execution of mobile organizations. From 1986 to 1992, she did research for faculty at the Harvard Business School and for CSC-Index’s Prism Group. Prior work was at AT&T and Hughes Aircraft Company.
Keri holds a Doctorate in Business Administration (DBA) in Management Information Systems from the Harvard Business School and both a Masters Degree in Industrial Engineering Management and a Bachelors Degree in Applied Mathematics from Stanford University.