Stories

Why You Need Bias To Be A Great Recruiter

Bias blog header

"Technology has played a role in transforming the world around us, and recruitment has not been immune." - Phil Blaydes, Talentful Founder and Co-CEO

By making candidate information more available, technology brought inclusivity to the table. You might not have Ivy League connections, but you’ll still have a chance. 

The downside of technology is how tools like LinkedIn have lowered the bar for recruitment, removing the art of the craft, and reducing the quality of the hire. You don’t need to be very good at recruitment with these online tools – you can simply send a message to someone without the bigger picture of company culture or understanding what the client needs.

 

Deep technology like artificial intelligence (AI) or machine learning (ML) enables companies to speed the processing of large volumes of data in recruitment and allows recruiters to weed out irrelevant candidates. But to do that, AI and ML use biased parameters. 

So if you have 1,000 people applying for one role which doesn’t require a degree, how do you reduce this number so it’s manageable? Maybe you would use a variable, like education. Do you opt-in or out for degrees? Probably in, and therefore you’ve created a bias.

 

You have to intentionally put bias into machine learning to get any use out of it, so we are consciously putting biases into the system. Parameters by their very nature are biases – we assume people with degrees are more likely to succeed, but that’s not based on science, it’s just trimming down a pool of candidates.

A crucial part of training our talent managers is to understand and identify the biases that are uncomfortable to them. One parameter is to understand our clients and what they like or don’t like, and that’s a bias. It creates consciousness and awareness of your market.

 

Humans come with a lot of biases. You aren’t born with them, they’re created by experience and your environment. At Talentful, we train people to recognize their biases to understand who they are. We have to accept that there are biases, and no one is exempt from having them. The first step is to understand what those biases are – positive or negative. 

You can’t be afraid of your biases; you have to pick the best ones suited to the job (that’s a bias). You need positive biases that serve as countermeasures in your decision-making process. So if you have a bias towards socio-economic diversity by example and your client doesn’t prioritize that and wants another parameter, what do you do in that scenario? Where is bias allowed or not allowed?

 

It is the job of the recruiter to help others challenge their negative biases, not just accept them to fill a vacancy.

Even with technology, people play a crucial role in the process of hiring. People are biased, and so are machines. The human brain is a much bigger and better computer, which takes into account so many more nuances than an AI can. Until a tech company can replicate the human brain, humans will always have a bias – positive or negative.  

The best recruiters will counterbalance that and look at all the factors at play and weigh the nuances needed to choose the best candidate for the client. That’s what makes people the best part of recruitment. 

All stories