top of page

Workforce Transformation: Moving beyond a walled garden view of HR Analytics

Big Data unchains us from reliance on one data set and gives us access to many sources of data. It moves us from a walled garden view of the world where data is limited and scarce to one of openness and abundance. And this abundant and robust world of information and data will allow us to gain insights that in the past we only dreamed about. It will allow us to make connections about people, and between people, and between people and results that will allow us to increase innovation, sales and revenue.

However, I continue to read articles and papers purporting to show the limitations of big data, or for the purpose of this article, “HR Analytics”. Most people, in my humble opinion, make two key mistakes when talking about the potential and the limitations of “HR Analytics”.

The first mistake that many make is that they discount the potential power of big data (HR Analytics) because the underlying “data” isn’t perfect. But data has always been, and will probably always be flawed. It is the unavoidable nature of data. Data is noisy and messy. If your company is determined to wait for perfect data before pursuing a big data people analytics program then you might as well give up now. The solution to the problem of imperfect data is to have access to as many data sets as possible to create a more holistic and robust view of the subject. This is what Big Data offers.

This type of thinking can be seen in a recent article in the Harvard Business Review titled Most HR Data is Bad Data (link here). Citing a growing body of research the author makes the strong point that the typical employee review process that most companies use is fatally flawed for a number of reasons, not the least of which is because of a handful of studies that clearly indicate that reviews tend to reflect the biases of the reviewers more than the strengths or weaknesses of the reviewed. This is fairly intuitive and I think that most of us who have gone through, and/or ran, annual reviews realize the limitations. The problem is that like many in the space the author then goes on and makes the claim that therefore “HR data is bad” and needs to be “fixed” before the promises of big data can be realized by HR.

This leads us to the second mistake, which is that when many talk about “big data” they are actually talking about “small data” and confusing and conflating the two. For instance, the author in the article makes the critical, yet common, mistake in assuming that those flawed reviews, as well as other bits of information generated by management, collected by HR and stored in the employee records are what comprises the totality of "HR Data". It is what I call a walled garden view of data. In other words, it thin slices Joe and Mary Employee down to a very limited view of who they are, what value they provide, etc. It is a profile based on reviews, raises, titles, sick days, vacation days, etc. And because the data is limited and fairly one dimensional it restricts and limits the types of queries I can make and the type of insights I can gain.

This is why “Big Data” is fundamentally different. By definition it means data gathered from many sources. So in the era of big data when we want to find “leaders” in our company we won’t depend solely (or maybe even at all) on reading and analyzing the employee’s reviews but will instead seek out employee’s with large social networks, who are thought leaders as defined by the number of followers their blogs and twitter feeds have, whose internal social networks are large, who hold leadership positions in their spare time in non-profit ventures, who coach little league teams, who set and achieve significant goals in their private lives (such as running marathons or climbing Mt. Everest), and the list goes on and on. And this is all made possible because the amount of data available about everyone online is increasing exponentially. It is robust.

So the answer isn’t to fix, necessarily, the data that the HBR author defined as “bad”, although it would be nice if that happened. The answer is to recognize that data in general is messy and noisy, move beyond a walled garden view of data, re-define “HR Data” and start about the task of figuring out how to use big data to gain insights that just a few years ago would have seemed impossible.

Featured Review
Tag Cloud
No tags yet.
bottom of page