HR analytics

By // 1 comment:

I have been in HR function for the last 5 years, and I have never published any post related to HR ever in this blog. Even this was an observation from one of my friend. With Linkedin’s new Pulse platform for article writing, I posted an article which I have been thinking about for a while now. Here is the article from Linkedin pulse page.

Here is the article from Linkedin I published,

There was a famous analogy on Big data by Dan Ariely that went on rage - Big data is like Teenage sex everyone talks about it, nobody really knows how to do it, everyone thinks everyone else is doing it, so everyone claims they are doing it. I think the analogy can be extrapolated to HR Analytics. With the growth of digitized data there is an explosion of Talent data and every body is talking about HR Analytics / HR Big data in most places I have seen. But there are only in very few places I have seen they really convert those analytics work into meaningful insights that had some real impact on the business.

I look at this analytics stack at three layers

  1. Data Aggregation and quality management - This is hygiene factor that ensures we are able to get all the various HR data points and we pull the right quality data for our work
  2. Data analysis and visualization - This step is churning the data on various statistical (or even simple) analysis and visualize them in graphs, tables, dashboards etc.
  3. Drawing Insights - This is the step we look at the data and pull various insights by adding other contextual and anecdotal information we have about the business. These insights convert to actions for our business.

In what I am observing most of the places we are doing a great job in the top 2 layers - Aggregating quality data and do a deeper analysis also. We even visualize with fancy dashboards these days with Power BI and other cool data visualization tools available now. But when it comes to Insights , I see most of the report outs are trends which can be any way seen by visualizing. We really don't pull out insights which are actionable for business. Because we really don't add context to the data and present as insights most of the time.

In the places I have observed, we are missing some of the basics done in research methodology. In research methodology the researcher will do some exploratory research to come up with a Hypothesis. Then he does all the survey etc to do data collection then does the analysis to validate the hypothesis. But in our situation we really don't start with some hypothesis to do analysis. Rather we do regular churn of data on try to get some insights. In any company there is no shortage of metrics and dashboards anyway.

So we should start focusing with Step 3 - on identifying the real business problems we are trying to solve and formulate few hypothesis based on the anecdotal data we have. Here the contextual data is not just HR data - it is about the business related information we have as well. Once you have the hypothesis validate them through various analytics work. Like any research problems if we form say 10-20 hypothesis we may be able to validate say 1 or 2 hypothesis that is actionable intelligence for business.

Instead of starting with HR Analytics work as a project - we should start calling as data driven HR where we validate various intuitions and heuristics we have already by using the talent and business data analysis. Then business would really see value when we give actionable insights and also an ROI on our investments in data analytics.

(Image Credit :