Want to see your candidates’ true colours?

Why and how to hire blind

By Angela Bradbury on 8th December 2017



Here are some of the things I hear fellow founders and hiring managers say on a regular basis about hiring:

“Interviewing is a real time drain and I don’t really know what questions to ask”

“We just don’t get diverse people applying”

“One of our new joiners interviewed really well, but I’m not sure they’re going to work out”

“We need new hires to be a culture fit”

“We can’t let diversity targets get in the way of hiring the best candidates”

“Our best candidates come from our team’s personal networks”

Sound familiar?

The way most of us go about recruiting for our teams is haphazard, biased and less effective than we would like.  Not all of the factors at play are within our control, but there are a surprising number of steps each of us can do to make the process better for ourselves, our teams, the candidates and our industries more broadly.  Hopefully I’ll convince you of a few of them.

I often hear people complain about the ‘pipeline issue’ - that, for example, it’s difficult to get women applying to software development roles because there aren’t enough girls who learn to code in the first place.  I’m not going to pretend that that’s not a factor; the recent BBC documentary No More Boys And Girls investigated how children today are still being fed depressingly gendered ideas about what careers are appropriate for men and women.  Another BBC documentary, Will Britain Ever Have a Black Prime Minister?, shone a light on the obstacles facing black Britons in rising to positions of power and influence, from school to university to a graduate job and beyond.  Large organisations have a responsibility to tackle these issues where they can, and many do, for example Airbus does a lot of educational outreach in girls’ schools.  But for smaller companies, that simply don’t have the budget to move the needle here, what can you do?

The first place to look is your job ads.  A report by Hewlett Packard showed that women are less likely to apply for jobs that they are not 100% qualified for.  When writing a job ad, it’s tempting to write a wish list for a candidate’s attributes and experience as if they’re hard requirements, although you’d happily interview a candidate with half of that list.  Don’t do that.  Your candidate pool will be disproportionately male, so that’s who you’ll interview, and most likely hire.

Second, look at your criteria for putting a candidate through from CV screen to interview.  Are there hard lines you are drawing on qualifications, previous roles, or number of years of experience?  Or are you simply going on what you or your recruiter like the look of?  Discrimination is rife at this stage; numerous studies using the “correspondence method” (where pairs of fictitious resumes are sent in application to job ads, and the rate of callbacks for interviews are compared) have shown that white-sounding names receive 50% more callbacks than African-American sounding names in the US, for example.  If it’s practical to do so, find a way to remove names, photos, and any other information that is irrelevant to your advertised position before the CV screen, as many organisations in the UK are now doing.  That might go as far as educational institutions, for example the law firm Clifford Chance has removed references to applicants' universities, to overcome what it felt was a perceived bias towards those from Oxford and Cambridge.  Solutions do exist for this, for example Blendoor.

The other thing you can do is to decide on hiring criteria before advertising the post, and stick to them, so that hiring managers can’t justify their biases after the fact.  A study by Yale researchers on police chief recruitment found that if a male applicant had more street smarts and the female applicant had more formal education, evaluators decided that street smarts were the most important trait, and if the names were reversed, evaluators decided that formal education was the most important trait.  By formalising a requirement for say, a certain level of formal education and/or a certain number of years of practical experience in advance, you can reduce this tendency.

OK, so then what happens when you get to the interview stage?  All other things being equal, it has been shown that hiring managers prefer people who look like themselves, in appearance, background and leisure interests.  That goes some way to explaining why interviews are such a poor predictor of performance.  In addition, interviewers often ask random questions, that they use to build up an idea of the candidate’s character in a way that can undermine more useful information.  This is often even encouraged by companies, for example Google looks for “Googliness” in new hires, a “mashup of passion and drive that’s hard to define but easy to spot”.  Perhaps not completely coincidentally, the company has faced calls to make improvements in the diversity of its workforce.  And, of course, there are the occupations where being socially able is not particularly necessary or common amongst high performers, and therefore interviews discriminate against neurodiverse candidates.

There is an alternative to unstructured interviews, and that’s auditions.  An orchestra would never interview a potential new player; they audition them.  Of course, it’s not perfect; an audition can’t fully replicate what someone is like to work with, or how well they can perform over long periods of time, but it gives a much better idea of their suitability for the position compared to a 30-minute chat.  In the same way, you can probably find a way to create a work sample test, for example a code challenge, written assignment, e-tray exercise or role play.  Again, solutions exist to help with this, for example GapJumpers.

This provides an opportunity to keep the application process blind to non-relevant information like candidate name even further.  The Boston Symphony Orchestra pioneered blind auditions in the 1950s, where musicians audition behind screens, and a study found that the likelihood of a woman being hired increased by 25-46%.  Confident that they would be treated fairly, female musicians started applying in greater numbers.  So if you can keep your application process blind until after your ‘audition’, there’s the potential to dramatically improve the number and diversity of successful candidates.

The case for every company to do something about this is clear, as testified to by the quotes at the top of this post.  The costs of recruitment are high, particularly in fields where there is a shortage of talent, and the costs of mis-hiring are even greater.  The potential benefits are huge: every 1% increase in diversity correlates to a 9% increase in sales revenue, and companies with diverse boards have, on average, 14% higher turnover than their non-diverse counterparts.  

So, to start hiring better today, do these three things:

  1. Decide on hiring criteria in advance, and stick to them

  2. Name-blind your application process

  3. Audition rather than interviewing your candidates