Moving From Data to Insight

According to a 2013 Executive Board survey of more than 1,500 senior business leaders, only 15 percent agreed that HR analytics has led them to change a business decision in the last year.

There are a handful of organizations that are using data-generated insight to make good HR decisions. Google is one. To find out what manager competencies were most important to produce optimal outcomes from employees in their environment, the company used HR analytics to find the answer.

In an effort internally referred to as “Project Oxygen,” Google used statistical analysis to identify and prioritize a list of eight characteristics that made a boss more effective. Technical expertise ranked last. Taking an interest in employees’ careers and lives was most important. The results from Project Oxygen had a material impact on the way Google hires, manages, develops and retains employees.

Building solid HR analytics capability is not for the faint of heart. The following four-step approach may help:

Step one: Ad hoc metrics and reporting: Most organizations start here. This is a necessary first step focused on creating and communicating retrospective information such as turnover, headcount and headcount to budget. Ensure the data is accurate. Bad data or fuzzy math are not an option. Partner with finance to ensure it is on board.

Step two: Descriptive benchmarking and dashboards: Adding benchmarking provides more depth to reporting by allowing the organization to see how internal trends compare to other organizations in the company’s geography or industry. Adding effective dashboards requires that the organization determine what metrics matter most for its specific objectives. Answer the question: “What problem are we trying to solve?” The answer should be for the organization, not just HR.

Step three: Identify and implement an HR business intelligence tool: Tying data from tools such as new hire and engagement surveys to existing metrics such as top talent productivity can have value. The output from this type of tool only has value if it helps the organization make better business and talent decisions. Therefore, find the right tool. Initially it needs to pull together data from multiple HR systems. Having the ability to include, and eventually correlate, data from non-HR systems will be necessary for organizations that want to get to step four.

Step four: Predictive analytics: This is the holy grail of HR analytics. Predictive analytics uses a variety of techniques from statistics, modeling and data mining to analyze current and historical data to make accurate predictions about future events. Find and secure a resource capable of doing the required statistical analysis. Many claim to know how to correctly do this type of work, but few truly do.

Instincts, hunches and feelings should not be the basis for important decisions. But without implementing this four-step process, or something like it, many leaders will fall back on their hunches or instincts to make talent decisions. Organizations don’t generally need new data. Instead, they need a strategy that enables them to make better use of data they already have.

David Almeda is chief people officer at Kronos Inc., a global workforce management company. He can be reached at