Women in Technology - A Behavioural Approach
A randomised controlled Experiment on How to promote careers in technology.
Women hold 32.4% of FTSE 100 board positions, a proportion that has grown steadily over the past decade. In STEM-focused organisations however, leadership representation drops to 5% with little evidence of a shift on the horizon. Women make up just 15% of undergraduates in tech and 23% of the STEM workforce.
The imbalance is a concern on multiple fronts:
The digital skills gap is widening - a failure to attract more women to a career in technology will have serious economic consequences.
In the future, careers in technology will carry a salary premium and greater protection against automation - under-representation within the field is likely to increase inequality outside it.
Everyday technologies, from digital assistants to AI algorithms, are being inadvertently programmed with gender (and racial) biases. Experts suggest that increasing women’s involvement in the design process would lower these risks.
With five times more young men than women reporting technology as their ideal career choice, these problems will not just disappear over time. We applaud the efforts being made to reverse these trends in the education system, but we also want to highlight the simple improvements organisations could make today. It’s in this spirit of ‘thinking small’, that we turn our attention to a single, practical question: what’s the best way to promote technology careers to women entering the workplace?
Thinking Small - How can employers better promote a career in technology?
Behavioural design is ruthlessly practical, improving precise components of an experience or decision-making process. Our focus here is at the start of the pipeline - what’s the best way to promote technology careers to women entering the workplace?
One well-trodden way to answer this question is through self-report measures, or to put it simply, to ask people what they want. We started our research by replicating this approach. We asked 450 young people aged 18-23 why they would consider a career in technology and we split their answers into three themes:
Prosocial - to help people and solve social problems.
Self-interest - to increase personal reward and career opportunities.
Communal - to work in a close community and be supported by a tight knit team.
Why would I be interested in a career in technology? (%)
These results are hardly groundbreaking. Search ‘what millennials want from work’, on Google and you’ll find hundreds of surveys that report similar findings. Here is a typical excerpt from a report from PwC:
“Overall, 40% of students say an important factor in their career choice is feeling that the work they do helps make the world a better place. But this figure rises to 50% among females and falls to 31% among males. In contrast, males are more likely to cite salary as an important factor – with 44% of males saying this, versus 32% of females”
The popularity of this type of research has a significant influence on how employers attract new talent. If you want to explore this, go to a few graduate career sites to see whether they pitch their job opportunities based on individualism or pro-sociality. Taken as read, the insights above suggest the more they lean towards the latter, the more women they’ll attract. Indeed, that is exactly the claim PwC go on to make:
“Given females’ greater desire to have a positive impact, more understanding of transformational effects like these would encourage more of them to get involved in technology as a career”
Moving from words to action - a Randomised Controlled Experiment
This ‘common-sense’ connection between self-report and behaviour may be intuitive, but it’s not always accurate. We often turn to David Ogilvy for a snappy quote: “People don't think what they feel, don't say what they think and don't do what they say”.
Our way to test the relationship between words (self-report) and action (behaviour) was simple. We randomly allocated our 450 young people into one of three treatment groups. Each received one of the below headlines, supported by a circa 100 word description about why a career in technology would support this particular aim:
Prosocial - “Help people live happier and healthier lives”.
Self-Interest - “Go further than you dreamed”.
Communal - “Join a community that work together”.
We were interested in which of these messages would encourage more young people, and particularly young women, to spend time exploring a technology website with career opportunities from top employers. In short, would the way we promote technology careers change their behaviour? You might predict that we couldn’t influence preferences with something so simple, or that our results would follow a similar pattern to our self-report measures. But this isn’t what we found at all…
Explore a career in technology - click rates (%)
Although the differences in the prosocial and self-interest messages seem to have reversed slightly, they are statistically insignificant - they do not have a gendered effect. This null finding casts doubt on the assumption that reframing technology careers as purposeful will attract more young people in general (another influential insight from self-report research), or to women as a subgroup. From a behavioural perspective, we find no evidence for either claim.
The only statistically significant difference was in the communal treatment, which has a clear gendered effect. To those familiar with the literature, the direction of this finding might come as surprise: communal values and behaviours are often found to be stronger in women. However, we do not suggest our findings contradict this research. In fact, we would like to explore whether the opposite may be true.
How can we explain the gender difference in response to communal framing?
When we asked participants what encouraged them to pursue a career in technology, only a small fraction stated that they had communal motivations. Yet, this framing lead to the highest click rate in men and the lowest in women.
Communal framing emphasises support, collaboration and teamwork - values that are stereotypically associated with women. However, communality is also part of what makes the technology industry so masculine, spawning words like “Brogrammer” and “Brotopia”. We wonder whether our communal messaging made this divide more salient to our audience: i.e. men belong in this community and women do not.
Research on communal messaging is in its infancy in the context of career promotion. We suspect in part, because the self-report research that dominates the field would deem it unworthy of further exploration. However, our experiment demonstrates a significant influence on the behaviour of men and women that would not be found with traditional methods. It also hints we have much more to learn:
We hope to pursue a number of different avenues of research from here. For our next experiment, we wish to test whether employers who have been successful in changing their culture to be more inclusive, may be able to encourage more women to apply by promoting their uniquely communal and supportive environment. This hypothesis relies on the assumption that women weren’t unwilling to explore tech careers because a communal message was unimportant, but because it was SO important. We look forward to putting that idea to the test.
We understand this nuanced conclusion may feel unsatisfying, so we’ll close with another: Tech employers who want to create a more inclusive culture should put experiments at the heart of their quest for progress. In the context above, we imagine brand, reputation and industry will interact with our motives in different ways, so an experimental approach is the only way you’ll find out whether you’re putting your best foot forward as an employer. Imagine how much we would learn if just a fraction of research budgets were spent in this way?
You may also be speculating whether these findings replicate in similar contexts in your workplace. So are we! Are you inadvertently promoting your internal careers, learning opportunities and leadership roles in ways that appeal more to groups that are over-represented? Again, we don’t have the answers. But we know how to find them…
A huge shout-out to the brilliant team at Gorilla, whose industry-leading experiment builder technology provided the platform for this research. Credit also to WOCinTech for the image we used in our thumbnail.
As always, if you found this article valuable, please consider sharing it with your network. If you want to commission an experiment or explore the potential of behavioural science in your workplace, you can get in touch below…