2010
The power of “naturally-occurring experiments”
Creating actionable business intelligence on the people side of the business requires being able to isolate the relationship between various measures of human capital management (HCM, the management and development of people) and an organization’s key performance indicators (KPIs).
Only rarely is it possible to do so through a true experiment, including control and treatment groups. But if you are clever, it is almost always possible to use “naturally occurring experiments” as a way to learn about the linkage between HCM and KPIs.
For example, even though your organization may be governed by a single set of HR policies and procedures, there is usually remarkable variation in how well and effectively these policies and procedures are implemented across different office locations or departments.
It is this variation that generates a naturally occurring experiment (e.g., think of an organization with 30 different sales offices, all selling the same products, but having very different rates of success). The variation in success rates represents a potential gold mine for learning about the human drivers (or impediments) of organizational results. It can provide an answer to executives who view the results and want to know why some sales teams are more effective than others.
In other contexts, the questions might be “Why do we have much higher rates of employee turnover among some call center teams than we do among others?” or “Why are some managers able to attract and retain a highly diverse workforce, while others are not?” There is no end to these “why” questions (or the stakeholders who want answers to them). They exist for every aspect of the business that is touched by human behavior – that is, every aspect of the business.
By collecting the right data and analyzing it in the right way, it is entirely possible to capitalize on natural experiments and answer these questions. (Multiple examples of such applications are described in our new HR Analytics Handbook). The insights resulting from the analysis can then be used to foster fact-based decision-making in service of improved organizational performance. That is the essence of HR analytics.
Alternatively, some people find it helpful to think in “Six Sigma” terms. Six Sigma began as a set of analytic practices targeted at improving processes and eliminating defects (undesirable outcomes) in manufacturing, but its principles can be applied in a variety of other business contexts. The insight here is similar to that which comes from viewing an organization as a naturally-occurring experiment: there are negative (and positive) sources of variance responsible for defective (good) results, and it is possible to deploy analytic techniques to identify these sources of variance – in this case, the people-related sources. Knowing the sources of variation in results is the key to laying the foundation for an evidence-based approach to decision-making.
Tagged natural experiments