Are you ready to use predictive analytics as part of an inclusive hiring process?

Are you ready to use predictive analytics as part of an inclusive hiring process?
Promoted by Are you ready to use predictive analytics as part of an inclusive hiring process?

‘Predictive Analytics’ is still a buzz phrase in recruitment, years after it was first coined. It is a way for recruiters to make more informed hiring decisions. At their most rudimental level, predictive analytics can predict peaks and troughs in hiring demand before they occur- particularly useful for seasonal recruiters or rapidly growing companies. At a more advanced level however, predictive analytics could be used to tell you which kind of candidate is going to be good at a job.

And this is where it gets tricky.

Let’s say you record information on candidate height. Cross reference with some employee productivity stats, you could make inferences like ‘short people are typically good at {insert job title here}’. Now unless your remit is to hire an Olympic level weightlifter (unlikely), you’ve probably been led astray by your predictive analytics. Short accountants may be more productive than tall accountants, but they are not more productive because they are short.

Could you imagine an application form for an accountant that specifies a maximum height? It sounds ridiculous.  While this might seem like a trivial example- of course a candidate’s height has no bearing on their ability to be an accountant- this kind of conclusion could made with more serious implications.

What if you were looking to recruit at FTSE 100 CEO? Looking at the profile of FTSE 100 CEO’s 93% are male. Predictive recruitment algorithms might, therefore, suggest you should only consider male candidates.

Of course, being male of female has no bearing on a candidate’s ability to be a CEO of a FTSE 100 company- and this is just a much harder hitting example of how measuring the wrong stuff can get you the wrong results.

But if you measure the right stuff, then predictive analytics can be used as part of an inclusive hiring process in the same way as any recruiting tool. Here are three questions to ask yourself before you start using predictive analytics.

1. Are you collecting the right kind of data?

There is the temptation to collect as much data as possible. But not all data is relevant, as we’ve seen in the example above.

2. Are you making the right kind of inferences?

Famously, Accountancy firm EY recently removed academic and education details from their trainee application process, preferring to interview candidates based on their performance in online tests. This is a great example of taking data and making the right kind of inferences.

3. Are you removing potential bias?

We’ve mentioned Nameblind applications before, and it’s no coincidence they are now being used by some of the country’s biggest employers to combat recruitment bias. Nameblind applications virtually ensure that a candidate’s application is reviewed on merit alone- and it means that any profiling from predictive analytics is kept to a minimum.


Be the first to comment.