Increasing diversity in recruiting – backed by behavioral science

Diversity is moving away from being just a buzzword. We now have proof that it has the power to push a business forward. There is ample evidence that diverse companies perform better financially, innovate more, have lower turnover and more engaged employees. More and more companies are interested in recruiting diverse talent – and often, they find it challenging. One of the industry sectors that have been notoriously bad at diversity is tech – which is why we were thrilled to be called by a tech company earlier this year to help them design a hiring process that would increase diversity.

The company, a scale-up, was facing a recruiting round that would essentially double the number of employees. The starting point was not great – by early 2019 the company had zero women. Our task was to design the recruiting process from scratch with the goal of leveling the playing field and increasing diversity. The objective we set for ourselves was that at least 30% of the new hires would be women. The result? At the end of the hiring round, that number was 50% – well exceeding our target. Let’s look at what we learned along the way.

Learning #1: Good intentions are not enough

Our base assumption is that people have good intentions, also when it comes to diversity and inclusion. In our experience, few CEOs or hiring managers believe that women make for less competent workers and actively try to make it harder for them to get hired. Certainly we knew that this was not the case with our client. All the people we talked to – from management to the engineers, expressed both awareness to the benefits of diversity, and a genuine motivation that the company would hire more women.

So if everyone is pro-diversity, why hasn’t it happened so far? Behavioral science gives us a clue to the answer. We all have our good intentions, but are equally vulnerable to the same kind of biases, which make us sometimes miss a talented candidate or positively rate a less talented one. There are no shortage of biases that are active in the recruiting process: halo effect (one positive aspect in a candidate influences our whole judgement of them – “he went to this great school!”) and affinity bias (we like people who are similar to us – “she plays volleyball and so do I!”) are just two examples. And research tells us that it is downright impossible to override these biases, despite our good intentions. So via these hidden mechanisms we end up making the same kind of faulty judgement calls we promised ourselves we wouldn’t do, which hinders us from achieving our goal: recruiting the diverse workforce we would like to see.

Learning #2: Don’t try to change attitudes – nudge instead

So much of companies’ energy (and money) to increase diversity and inclusion goes into diversity trainings. These are meant to increase awareness of the ways in which our brains are wired in ways that make it tougher for women and other minorities. Sounds great. The problem? They don’t work. Since we have already established that we have good intentions, it’s not our attitudes that need changing. It’s our behavior. Our hearts are already in the right place, but our mind is vulnerable to biases, as we have shown. Context matters a lot, so we need to carefully adhere to the design of the environment where decision takes place in order to nudge ourselves in the right direction. This means designing recruitment processes, rather than working to change the attitudes of recruiting managers. The way to do this is through a technique that has been called “behavioral design” or “nudging”. Basically, taking everything we know about the biases people are subjected to, and using that to design an environment that promotes better decision-making.

Learning #3: The first step? Make it easy

One of the first rules of nudging? Make it easy. In this case – you want more diverse candidates to be hired by your company? Make sure it is easy for diverse candidates to apply. For us this presented a challenge. Our client being a tech company, many of the positions we were looking to fill were of software developers. To recruit developers, there is one platform which rules above all – where all developers are and where all companies recruit. Unfortunately, it turned out this platform is not super inclusive – only 20% of its members are women. It became clear that if we want to reach women in our recruiting, we need to look beyond it. So we did. We searched and found recruiting platforms that target specifically women in tech, like this, this, this, and this. We wanted as many women to be exposed to our job ads so that they can apply. This clearly required more effort than just putting the ads on the one platform that everyone uses. But we knew that as a first step, we need to make it easier for diverse candidates to reach us.

Learning #4: Framing matters

The framing effect is a known quirk in behavioral economics which tells us that the way we present information leads to changes in how the information is perceived. In other words, our choice of words matters. This is true also in the context of recruiting. Research shows that women and men respond differently to different words in job ads. For example, words such as competitive, dominant and leader attract more male candidates, while female candidates respond better to words such as support, understand and interpersonal. Using too few words that appeal female candidates is just one of the ways that companies might make their job ad seem less appealing to women, without even realizing it. We revised all our job ads in this spirit, and even ran them by a software tool that gave us feedback on gendered language using AI. Words matter, so paying attention to the words we chose was important.

Learning #5: Be honest. Does everyone get the same chance to apply?

As a hiring manager faced with the question above you might say – of course, everyone who sees my job ad can apply. But will they? It is a known fact that while men apply for a job when they meet only 60% of the qualifications, women won’t apply unless they meet 100% of them. So having a grocery lists of “musts” might be actively discouraging women from applying to your job. What is important is that this is not about lowering the bar. Often hiring managers would put a long list of demands in job ads, but during the hiring process would find a candidate who is brilliant despite of not ticking all the boxes. That candidate would get a chance to move on to the next stage and might even get hired. But that candidate is very unlikely to be a woman – the woman, who might be as brilliant but does not answer all of the requirements in the ad, well, she never applied. So although we think everyone has the same chance to apply to our job, they don’t. With this in mind, we severely cut into our list of “musts” and divided it into two: “you have” – which was a short list of the qualifications, experience, or knowledge the candidate really MUST possess, and “we’d be extra impressed if you also have” – which was our list of extras. It would be great for the candidate to have them, but we would review them even if they didn’t. This way, we level the playing field.

Learning #6: Interviews are not the place to let your creativity loose

Many companies like to say they recruit for a “culture fit”. Our client was no different. So in their pool of interview questions you could find questions about programming languages and search algorithms, as well as the question “what is your favorite sci-fi book?”. Let’s ask ourselves first – what do we mean when we say “culture fit”? There is no problem with wanting to recruit candidates who are collaborative, trustworthy, creative – or whatever it is the “company culture” stands for. The problem starts when “culture fit” becomes a code name for “a person like us”. Particularly when “us” is not a highly diverse group of people. Knowing a person’s favorite sci-fi book might make you like them more (remember the affinity bias?) but it won’t give you any indication about their competence, so you’re better off without this question altogether.

Don’t take our word for it – Lazlo Bock, former SVP of People Operation at Google agrees with us that coming up with creative interview questions might be fun for the hiring manager – but the chances that it would improve the quality of your recruiting are slim to none. In order to overcome the biases we mentioned earlier, what you want to do is conduct structured interviews instead: define in advance the list of criteria you want to recruit for – such as leadership, problem solving, or a technical mastery in a specific area. Then match them with behavioral questions about the candidate’s experience – those give you the highest predictability with job performance. It is very important that all candidates get the same questions so that everyone gets a fair chance to impress (or unimpress) you regarding their level of skill in each area.

Learning #7: Test. Learn. Adapt

Which out of all the learnings in this article is relevant for your company? Which is the most likely to make a difference for your recruiting? You won’t know unless you try. Testing is a cornerstone of behavioral science, particularly using the golden standard method of RCTs – Randomized Controlled Trials. As a small company, our client didn’t have enough scale in the recruiting process to conduct true RCTs to see which interventions worked best. Very few companies are large enough to do that. But we did not let this fact stop us. In an iterative process, we kept trying, testing, learning and adjusting – our ads, our choice of platforms, our interviews.

Knowing what works to increase diversity may not be clear at the outset, but one thing is: doing what you’ve always done is unlikely to produce a different result. Remember that your good intentions are not enough. If you look around in your company and all you see are white males in their 20s, ask yourself what factors in the recruiting process may be contributing to this. The bad news is that if you don’t address them, you are likely to keep recruiting white males in their 20s. The good news is that just a few small adjustments in your process might go a long way. Give it a try! And if you do, make sure to let us know how it goes.

Want to learn more about how to use behavioral science to build more effective organizations? Check out our online course “Designing nudges”, or our complementary guide for applied behavioral science. Prefer to listen? I talk about this project, and more, in a Behavioral Grooves podcast which first aired in July 2019.

—

Nurit Nobel

Nurit.nobel@impactually.se

+46 76 191 71 34

I want to learn more!

We hear you. If you’ve already read our posts about  what behavioral economics is, how it can be used in practice, and how to apply it in organizations, then here’s a list of books, TED talks and other online resources. You are also welcome to sign up to our newsletter, where we give you relevant news and links on behavioral science.

I’m convinced that behavioral science can do wonders for me. Now what?

Contact us and let’s talk about how we can help you get going.