Of late, my Talent Analytics sessions with the HR teams commence with the ‘Emperor’s Clothes Story’ that one of our clients had told his team at the kick-off meeting on Analytics. It was the story of two fraudulent artists imposing as ‘Royal Tailors’ who trick an Emperor into thinking that they own a superior breed of fabric, which is visible only to clever people. The ministers, common people and even the Emperor play along for the fear of being seen as ‘not clever’. The emperor realizes the truth only when a child innocently screams, “The Emperor is wearing nothing at all!” The moral of the story that the client wanted to put across was – do analytics because you want to decode specific people problems and address some business challenges, and not just because everyone is talking about Analytics! In my personal experience, successful implementation of Analytics projects have only taken place at those places where the Emperor’s Clothes syndrome doesn’t exist.

There are many myths that people have about talent analytics. I refer to four of the most talked about ones:

It is a new age thing: No! It’s not. There is a 90 per cent chance that you are doing or have done some analysis on your people in the past. Just the complexity has increased in the form of the number of people and the parameters that you can track. The computing power has also increased allowing you to crunch data in an effective fashion.

It is just about Technology: It is actually less about technology and more about logic. It starts with your business objectives, knowing what data you need, listing the parameters you can track, aligning all your primary and secondary data in one place and figuring out which parameters correlate more to the business outcomes. Technology is a means to make this logic faster, efficient and easy.

It is just a BI problem: Talent Analytics is not just about taking the existing data and trying to infer trends about your people. Human beings are complex. Data from HRMS or Resumes is not enough to derive meaningful insights. You need to use tools to capture the emotive side of your people to draw meaningful insights.

It is fancy: It is much more than the fancy-looking charts or graphs that you can present to your boss. It is a lot about cleaning, sorting and aligning the data to the business outcomes.

So what does one need to do to make Talent Analytics successful? Based on our client case studies on what goes into the successful implementation of Talent Analytics projects, here are three conclusions that I have drawn:

Data Visualization vs Analytics

This was what the GM-HR at a manufacturing company initially thought about Analytics– we would take the existing data that the company has and display that as charts and graphs. Once he realized that data visualization was just one small part of the entire Data Analytics project, his perspective towards Analytics changed. Analytics is inferring patterns from data and figuring out which parameters correlate more to business outcomes.

Emotive plus Resume/HRMS Data

In a large BPO set-up, we were trying to find the ‘attrition profile’ or causes of 90-day attrition. We tried plotting all possible parameters of the employees - from the number of dependents, to college and education background, to distance from office, to leave patterns but we were unable to see any meaningful correlations. We then measured the scores of the candidates on a personality assessment, a moodometer and a passion survey. One of the insights derived after the analysis was people who stayed within 10 kms from the office, who had two social media profiles, who scored low on stress tolerance and monotony tolerance, and had frequent swings on the moodometer tend to quit in the first 90 days.

Workforce Planning vs. Predictive 
Analytics

In successful projects, the HR teams don’t confuse between manpower planning and predictive analytics. Manpower planning is about historic trends and future planning for salary, headcount, diversity, skillsets etc. Predictive Analytics is about identifying the causes of good performance, failures, attrition among others.

Hope these findings help your Analytics project. Happy Analyzing!