The Elephant in the room – why we struggle with diversity in tech?

Code Chiefs
Career, Code Chiefs, Development, Tech

Code Chiefs:

The Elephant in the room – why we struggle with diversity in tech?

What is the biggest problem in the Sydney tech market in 2018? Is it the rapidly changing technical landscape? Or maybe the lack of suitable candidates?

No, the most obvious and sad problem in the Sydney tech community to date is diversity.

Attending the last Code Chiefs event, I saw the problem myself – out of 35 Sydney technology leaders we only had 2 women. Scary, right?

On the 8th of April we held our 3rd Code Chiefs event and decided to discuss the elephant in the room. With two of the biggest advocates of diversity in tech – Roisin Parkes from Gumtree and Dan Draper from Expert360.

Code Chiefs

We had a rather heated discussion on the topic and some very important issues raised. I’ve summarised a few of the key findings for you:


1. The HR manager is not responsible for diversity & inclusion

Roisin makes the fantastic point – The HR team is not responsible for inclusion in your team. It is on you as a technology leader to influence the environment within the team and make people feel included in group discussions. A lot of companies fall into the trap of positive discrimination and make diversity a KPI. This tactic might attract women to the industry, but without the right team structure and support, the chances of them leaving the development team or moving to a less technical role are incredibly high. It is essential for technology managers to ensure that diversity and inclusion is not a % sign on a spreadsheet – it is something that is nurtured and cultivated within the team

2. Diversity works because it makes us feel uncomfortable

As Dan mentioned, homogenous teams feel easier and more comfortable. And comfortable is usually bad for performance. Divergent thinking and multiple perspectives make your team perform better. When culture is stagnant people feel comfortable and, hence, reluctant to change. Creating constructive conflict within a team encourages thinking and brings new ideas. That is why creating a diverse team is not only good for the organisation’s morale, but it actually boosts performance and increases cognitive flexibility.

Code Chiefs

3. Perks of a Dev job are not table tennis, beer pong and paintball

In my three years in recruitment I’ve seen countless adverts highlighting perks of the job like Playstation room, beer pong, etc. As much as I love a good PS4 game, it is apparent that most of the benefits listed in tech job adverts are very skewed to the male population. I rarely saw options to leave your child for day care, or strong maternity leave packages listed as added benefits, and, I have to say, things like this are usually very important for women. Traditional “masculine” advertising is one of the reasons many women don’t even apply to some of the technical roles. Switching the focus of your job advert may just change that.

4.Team structures matter

It might seem very simple, but team structures matter. You might feel tempted recruiting 5 women and putting them into 5 scrum teams full of men and, as stated in point 1, hit that “diversity KPI”, but think about this: Every time you are a sole representative of a minority group in a team, I bet you feel uncomfortable. And this is the most common thing for women in development. They’d be lucky to work alongside another woman. But usually, in an attempt to hit diversity targets, female candidates are placed in homogenous teams that consist only of men. And this can create a lot of stress and tension, forcing women to leave the job.

The idea of diversity is not as simple as the man and woman ratio in the workplace. Roisin and Dan demonstrated this in the examples above. With a few simple steps behind the scenes and from the leaders of your business- you too can have a high performing diverse team. But the change comes from the top.


Code Chiefs

  • Grant W

    It’s quite frustrating to see these low level interpretations of “diversity” and their challenges/solutions make it into the data sphere.. it’s just pure conjecture. We should be digging deep into data sets to explain differences, not making broad generalizations based off of group averages. It kinda amazes me that men and women are compared in this binary manner, with no regard to variance or data distributions – you can’t judge the individual from their cohort. I always find this ironic as diversity movements should be more about empowering the individual and not falsely tying them to some kind of group ideology.

    First, what does diversity even mean? Diversity in skill, gender, height, sexuality, ethnicity, hair colour? It’s just a broad umbrella term that could encompass an almost infinite amount of groups. But in the context of the article above, i’ll take diversity as being used to describe gender diversity (which could be sub-grouped into black females, black youth females, black christian youth females, black christian youth females with divorced parents…. which is where you actually start to find meaningful differences rather than broad groups like men vs women).

    Point 2 really hits the nail on the head with everything wrong with the current objective to increase diversity.

    “Divergent thinking and multiple perspectives make your team perform better.”

    Divergent thinking and multiple perspectives do not necessarily make your team perform better – there’s no evidence to support this statement, certainly not as a constant truth. Further, it also implies that you can only get divergent thinkers or a different perspective from someone outside of your “group” – in this context it is implying that men and women are different and neither group is capable of thinking like the other.

    “When culture is stagnant people feel comfortable and, hence, reluctant to change. Creating constructive conflict within a team encourages thinking and brings new ideas. That is why creating a diverse team is not only good for the organisation’s morale, but it actually boosts performance and increases cognitive flexibility.”

    I’m not even sure how you can determine if a culture is stagnant and why a stagnant culture wouldn’t include constructive conflict. It sounds you’re trying to describe bringing in an outside perspective or skill, but this doesn’t necessarily have to be from a different identity group. It’s conflating the idea of diversity in skills with diversity in gender and diversity in gender doesn’t have a lot of empirical evidence to suggest it boosts performance. The idea that “diversity” is just an all round boost to performance seems performance seems counter-intuitive to human history, where homogeneous groups in which individuals hold common values perform better – which is exclusive of gender, ethnicity, age and race.

    This isn’t to be obtuse – it’s literally what the data we have so far is telling us. Grouping humans into males vs. females is too broad to make conclusions like the ones above. The variance in individuals is too high – some females will add to a work place, some won’t. It’s the same for men, some will, some won’t. Ignoring the differences between individuals in a group is a good way make bad interpretations. It just sends out a horrible message that you’re locked into your group.

    Point 3 again shows how divisive this thinking is – table tennis isn’t a masculine or male only perk. Plenty of men don’t like table tennis and plenty of women do. I’m also sure plenty of companies do offer “female” benefits you mentioned above too. Using one example of a table tennis / ps4 perk to extrapolate information about an entire gender/population is just poor practice. Maybe there are more companies that offer these “male” perks over “female” perks, but I wonder how different the talent pool is? If the talent pool is 80% male, the firms are only reacting to the market and trying to appeal to the majority. I’m not suggesting this is a socially optimal outcome, I’m suggesting they are just acting as rational actors.

    Traditional “masculine” advertising – what does this mean? Why would a woman, who presumably has the self-awareness to realize that the data industry is skewed towards men (which is multivariate problem with gender playing a relatively small role), not apply to an advert that adds table tennis to their perks? I don’t like table tennis, but “fun perks” like that would never sway my decision on a job – if anything you read between the lines and see that as the company trying to promote a more casual and fun environment. It just seems to demonize the word masculine and only attribute it to male environments.

    You’d also think if there was this glut of female talent caused by “masculine” adverts, that any company that placed “feminine” adverts would excel. Why has no one explored this obvious market?

    Point 4 is just again pure conjecture. It doesn’t even align with the premise laid out in point 2 – that diversity drives performance. Here, apparently adding diversity (1 female into a 5 male group) led to worse performance. I wouldn’t agree with either interpretation as if you played this scenario over 100000x times, you’d get a wide range of results, suggesting it’s not just gender variables at play.

    But anyway, I can’t tempt myself and bite on pieces like this. Not to disparage the author, but to try counter this toxic ideology that seems to be rearing its head in society about identity politics that group identity is more important than individual identity. There are differences between groups, there’s plenty of empirical evidence to show that men and women AS GROUPS are different when comparing average. This doesn’t mean differences are deficiencies, this doesn’t mean that all men are better than all women at *A* or vice verse…it just means there are average group differences, but given the distribution and variance you can’t judge an individual by its cohort. Anyone who works in data should know this – that’s why I take concern with pieces like this.

    If you actually want to help females enter the data workplace, you should be empowering them by empowering the individual. Make data driven decisions (preferably not based off of anecdotal and group averages). If there are discriminatory barriers – confront them specifically and not just make broad meaningless generalizations that are impossible to action. If there are structural barriers in law – challenge them (e.g. equal maternity benefits).