Talent Analytics: Predicting HR’s Way Out Of The Fog

Here’s the average amount of time recruiters spend looking at a resume: 6.25 seconds. That’s how long it takes to evaluate, by brain, whether or not a candidate is the right fit for a job. And here’s another stat: this is the thirteenth month in a row when 200,000 plus jobs have been created.

So, theoretically, to fill 200,000 jobs would require 347 hours of brain time. That doesn’t count all the other candidates who didn’t make the cut. Or what happens after the hire.

The variations on success or failure in HR are always endless: we’re human, not robot, for one thing. But given our profoundly transformed world of work, the variables are now also epic. From multiple generations to global organizations to the enormous impact of Big Data, there’s no turning back. Our era has been called the Talent Age, the Social Age, the Mobile Age. What it’s not: the Pile Of Resumes Age.

Hence HR’s present headache. There’s a lot of talk about how we need to change the culture: become more people-centric, understand what the millennials and innovative talent wants (start by accepting that they are The Future of Work), figure out how to foster engagement and express recognition and make sure no one leaves. And then there’s all this data. Big Data can seem part ether, part mega-entity. As someone told me, it’s like a fog machine was left on and filled the conference room as we all sat there, stunned. And the word unstructured can strike fear in the hearts of even the most seasoned talent managers.

But it’s not a fog of data, it’s our own fog. We need to approach epic change in an epic way and be very clear about it. To really leverage human capital now, we need to turn to the data that is constantly forming, streaming, reforming. Passive and active candidates, onboarding, training, engagement, retention, attrition, performance, recognition: it can all be predicted with Big Data.

The key is that we are not just gazing into a crystal ball, we’re looking with clarity, knowing that the more information, the more time, the more data points, the more accuracy. But this is about modeling, and about forecasting:

  1. Turnover.Predicting the risk for the most turnover — in which functions, which units, which locations, and what positions, and modeling the scenarios in advance
  2. Churn / retention.Identifying where the highest risk of churn is going to be, and who is at risk for it. Determining what resources should be turned to them in terms of retention activities and / or training.
  1. Risk.Building realistic profiles of which candidates are risk for leaving prematurely, and when. Creating models of which candidates are likely to experience drop in their performance.
  1. Talent.Forecasting who, among new hires, are going to be the high achievers and high performers, and decide should they be shifted into fast track programs. 
  1. Futurecasting. Modeling the various changes that an organization may experience, from global to political, and what the impact of talent hiring, retention and engagement could be.

We need technology that can be used through mobile devices, is interconnected via the cloud so it’s consistent across the board, is intelligent enough to keep learning, is agile enough to refocus. We also need tech to be consistent enough to be a Watson to our Jeopardy. But that’s exactly what predictive analytics offers: the ability to take the past and make sense of it in terms of common factors and key relationships, and to use that information not just to model and predict the future, but to make sound and insightful recommendations.

It may seem like a glaring paradox, but in data lies the future of human resources and talent management. So yes, we do need to change the culture — to one that relies on data. And then we can see clearly.

A version of this was first posted on Forbes.