The HR Analytics Conundrum
The clamour for HR to embrace analytics has been going on for quite a while. However, many HR departments are still struggling to make a breakthrough in this area. I personally feel that implementing HR analytics need not be a complicated issue. Here are some of my thoughts on why HR may have found the going tough and how HR can overcome them.
Poor understanding of subject matter
Many HR departments today report headcount and attrition but fail to realise that such reports are a form of analytics . Strip away the terminology and analytics may not be as daunting a subject as many think it would be. This lack of understanding also blindside HR management and prevent them from allocating proper resource to deliver analytics.
Blind Adoption of Best Practice
Google’s success with people analytics has created a buzz and many HR departments are now attempting to recreate a similar setup. What many fail to realise is the infrastructure and human capital necessary to deliver analytics on a similar scale. This may be due in part to the lack of understanding of the subject.
Issue of Data Quality
Several HR departments tend to point towards poor data quality as a reason for not embarking on analytics. However, the reality is that most HR data are often patchy, especially since most ERP software was never designed with analytics in mind. While the ERP software helps to ensure some uniformity in the data, the quality is only as good as the person who is keying in the data and as long as there is human activity, the data will never be perfectly clean.
How should HR then approach this issue?
Start by going back to the basics
Understand the purpose of HR analytics and what is involved in analytics. I define it as the identification of trends and patterns and/or the discovery of correlating factors that may or may not be causal.
Define the business issue
What are you attempting to study? Only with an understanding of the issue can any form of analysis begin.
Tell a story
What does the pattern or correlation suggest? For instance, if you are analysing attrition, a pattern or trend may suggest a profile that is at-risk. Or there may be correlating factors that help you to predict attrition (e.g. spike in medical leaves correlating with high attrition).
What do you intend to do?
Ultimately, HR departments must take proactive measures and act on what the analysis may be suggesting. Back to the example of the identification of an at-risk profile, a possible action may be to engage the at-risk group and understand what might cause attrition. Only with such an understanding can solutions be designed.
I believe in keeping things simple and the approach described above does not require heavy investment. However, I do not deny certain infrastructure (e.g. data warehousing) or tools (e.g. visualisation software) may help to make the process more effective, but these should not be reasons for not starting out on HR analytics.