Untitled, by Kevin Hale.
Is there a simple and straightforward approach to applying human resource analytics to emerging issues? Yes, there is. Two authors have identified a few basic tips that make or break the creation of a meaningful analytics team. First, you must get analysts whose subject matter is human beings. And second, you must use a consulting approach when deciding how to meet business needs.
I would like to endorse the opinions of Alec Levenson and Alexis Fink in their article Winning the HR Analytics Arms Race from April 2017. Their portrayal of what is happening behinds the scenes in human resource analytics very closely matches my own experience of what is really going on.
The article flags that human resource executives are putting top priority on leveraging their data, but they put a low priority on predictive analytics. Just about every other priority area in the top-six are things that could be enhanced by people analytics, like succession planning, workforce planning, and diversity. But notably, executives have determined they are absolutely not ready to enter the ethereal world of predictions. The field is not fully developed and the data set is not yet mature.
The article describes the large number of people who can do clever analytics in different fields, such as engineers, accountants, and rocket scientists. But there’s a problem. Some of these people have never studied employees and their motivations. The authors favor industrial psychologists. My personal experience is that a blend of multiple social sciences is good, as long as everyone has an analytic bent. However, if someone spent the last decade crunching numbers in a field which does not consider the human soul, they’re factually a novice.
The authors also pan the tendency for analysts to try to make something with the data they have available. I ran through this loop in my first four months on the job. The readily-available data is stuff that the payroll and statutory compliance teams needed to get their own work done. But sometimes the most interesting information is stuff that is hard to get. Sometimes you need to create new information from scratch. And often you need to choose a higher standard of data quality.
Levenson and Fink favor a consulting approach, although they don’t call it that. The analyst must meet with the client and figure out what problems keep them up at night. The client won’t name data, they will name a nuanced person-puzzle for which analytics might be finessed into a useful tool. But the solution is several steps in. You have to start with the consulting question. And indeed, this is what business partners do. It is what business analysts do. Face it, if you cared about people, how could you not ask what’s important in the mind of your client?