Picture: Shutterstock

Wednesday 21st January 2015

I predict a riot

Could predictive analytics be the next big HR asset?

Every person fears something different. Spiders. Heights. The ever-accelerating tumble towards death. But what every rational person should fear is typing in half a sentence into Google and gaining another terrifying insight into the mass human psyche.

That is early predictive analytics at work. What people have typed previously being used to predict what you are searching for now. In theory it’s a time saver, a spellchecker, and a relief to see that you’re not the only one.

In the beginning it was clunky and silly, but now it works so seamlessly you hardly notice. You probably haven’t seen the actual first page of Google results since 2009 – what you see is tailored – your bespoke search results. For a baker, “man digs into crust” might yield videos of enthusiastic bread-eating. For a geologist, it might yield something about a journey to the centre of the Earth.

What matters is that with 2015 sweeping in, the time has come for another look at analytics and their relevance to HR. What can predictive analytics do for you, and what are the advantages over its standard counterparts?

As Big Data gets bigger (IBM says that 90% of all data was produced between 2012-2014), it also gains further potential uses. Hidden within all that data lies many holy grails of HR – the identification of perfect-fit candidates, forecasts of employee retention, even the precognition of office conflict in spooky Minority Report style.

The problem is that likewise, it becomes even more inaccessible to standard analytics. In the words of Robert Skidelsky in the New Statesman:

The net of delusion is being cast ever wider, as we are bombarded with more and more information masquerading as knowledge, more and more material for the calculus, which far outruns our ability to sift it into truth and falsehood.

The only avenue is to analyse smarter, which is where predictive analytics comes in. Predictive algorithms can find patterns in data too subtle for standard analysis, separating the wheat from the chaff, and leaving the conclusions and insights to the people.

Its success in other more numerate fields is well measured, and its efficacy in HR is already on the horizon, if not at the front door.

For example, Black Hills Corp, a two-thousand employee energy firm, used basic predictive analytics to calculate that they faced a huge talent shortfall due to imminent retirements. This insight allowed them time to create action plans in order to prevent a turnover disaster. Their forecasts showed a staggering loss of 8,063 years of experience over five years.

In the future however, the sky’s the limit. For example, predictive analytics could use hundreds of thousands, even millions of data points gathered from social media and companies, to analyse ideal seating arrangement based on shared hobbies and interests. Music taste and where you grew up might be used to correlate likelihood to resign.

Certainly many high profile suites such as Oracle, Workday and Visier are offering subtler and more advanced predictive capabilities than ever. And in a twist, since many suites are now cloud-based, the more companies that use them the more accurate they become, leading to a beneficial feedback loop.

This means that the window of when to adopt it in order to get a commercial advantage is even smaller than usual. It also means that smaller companies, for whom analytics would normally be cost-ineffective, could benefit from the much larger data samples taken from larger firms.

Jump too early and you risk the curse of the early-adopter, too late and you’re left twisting in the wind. The adage used to be you might as well look into your crystal ball, but perhaps nowadays it would be more appropriate to let Google predict for you:

About the author

Jerome Langford

Jerome is a graduate in Philosophy from St Andrews, who alternately spends time writing about HR and staring wistfully out of windows, thinking about life’s bigger questions: Why are we here? How much lunch is too much lunch? What do you mean exactly by ‘final warning’?