Since there is a large disparity in where organizations are regarding the sophistication of their analytics, it is helpful to discuss how they collect and report data in terms of a human capital analytics continuum. This new view of human capital analytics starts with simple, commonly used techniques and compares analytic maturity to mountain climbing. The methods become more difficult as companies climb the mountain, but they are more rewarding as they reach higher ground.
The ascent begins with anecdotes or storytelling. Robert Brinkerhoff has done some of the best work in this area with his success case method, which uses surveys and interviewing techniques to derive stories of real results people have experienced after applying learning in the workplace.
Scorecards and dashboards are other important techniques. Scorecards are strategic performance management tools that can leverage automated surveys to track how an organization executes strategy and the consequences arising from business processes, most commonly referred to as activity metrics. Scorecards characteristically have a mixture of financial and non-financial measures, each compared to their targets, all within a single concise report. They serve an important role on the continuum, because this step is where you must lay out the basic assumptions of what your strategies are and the various ways in which you will measure those strategies.
Dashboards share those characteristics. A dashboard is a distillation of the most important key performance indicators of a company that an executive can view at a glance. They might be an ad hoc effort put together on spreadsheets or they may involve specialized programming.
Benchmarks are another step. Benchmarking has long been used as a standard tool; the idea is that studying the best-run companies in a specific area can be very beneficial in terms of setting things such as salary, training levels and desired turnover rates.
Scorecards, dashboards and benchmarks typically lead toward correlations, which are the next stage. Correlations describe the statistics where two or more variables are moving together, but you don’t know specifically why. They are rich data mines for understanding business resources and human capital. But correlation does not imply causation. Strictly using correlations, we might infer that umbrellas need to be banned because it ends up raining on days when everyone carries them. So be wary of making business decisions solely based upon correlational information.
Causation is the next level beyond correlation. Causation shows the proof of why a metric or key performance indicator has changed. In their book, Investing in People: Financial Impact of Human Resource Initiatives, Wayne Cascio and John Boudreau identify three criteria of causation:
1. Two events must show a clear and statistically significant connection.
2. One event must precede the other.
3. All other plausible causes must be ruled out.
The final stages of the continuum are predictive analysis and optimization, the holy grail of human capital measurement. Predictive analysis is having the intelligence to understand where the investment is working and where it is not by segmentation — job title, tenure, department, business unit and region. This is intimately wrapped up with causation. Without understanding all of the factors that determine impact, it is impossible to correctly assess it.
Understanding where the investment is working and not working allows you to make changes to improve future investments. For example, if you know how much tenure influences performance, and which tenure levels benefit the most from a training investment, you can now specify where training should be focused and what new programs need to be created for those tenure levels not realizing the same benefits.
All the analytic levels on the continuum are useful. The rigorous analysis that provides true impact and helps you optimize your investments is the most important goal in human capital analysis.