A friend recently recounted one of the first days of her career, which happened to be at a web technology firm. Taking your initial steps as a professional should be a time filled with anticipation, excitement, and inspiration — yet this was anything but.

I was dismayed to hear that not only was she the solo female member of the staff, but it also quickly transpired she was chosen based on looks alone and not her considerable knowledge and talent. The employer hadn’t even read her CV.

I can’t say anything so blatant has happened to me, but I have on many occasions been the only woman in a meeting room.

Given the well-known male dominance of data science, these things will probably surprise no one. That doesn’t, however, make an obvious gender gap right. Without a more inclusive culture, my profession — which is more vital than ever in a post-pandemic, tech-driven world — will remain uncomfortably lopsided.

Time to stem the confidence crisis

Demand for data science jobs is predicted to grow rapidly, yet just 26% of women occupy industry positions in the U.K. (source: Better Buys, 2021). Lack of representation occurs early on, with just 35% of women studying higher education STEM subjects (source: stemwomen.co.uk, 2021).

Meanwhile, a 2012 survey conducted by OECD among 15-year-old students revealed females feel discouraged to study STEM subjects and identified a ‘confidence gap’: 41% of girls surveyed compared to fewer than a quarter (24%) of boys agreed with the statement: ‘I’m not good at mathematics.’

Looking back, this makes sense: I was the only female in my A-level math class. And I reckon things aren’t much better today.

Of those who do embark upon STEM careers, more than half (53%) of women quit companies after 10 years, compared to just 31% of men (source: Stych, 2019).

Lifestyle choices play a fundamental part in this, with fertility and parenting impeding women’s career development. The gender pay gap leaves females feeling frustrated and undervalued, resulting in attrition.

While the U.K. gender pay gap for full-time workers has decreased from 9% to 7.4% since 2019, it still exists (source: ONC, 2020). These factors are arguably dominant in all professions — it’s simply intensified in data fields due to the smaller volume of women at the outset.

Lack of female mentorship is also a factor, with 30% of women in STEM roles feeling isolated due to lack of mentorship provision (source: McKinsey, 2015). I’ve been lucky enough to be exposed to several amazing female line managers. They’ve all given me guidance and support. But I know I’m in the minority.

Mounting evidence also suggests women’s viewpoints and experiences are being omitted from development of data technologies. Instead, they are predominantly based on a ‘universal male’ perspective.

Because of this, there are concerns that algorithms will widen gender gaps, with data only telling half the story, rendering machine learning systems less robust, and even invalid.

There are plenty of good reasons for tackling the gender inequality in data science. A more level playing field:

• encourages collaboration, creativity and innovation

• enhances staff retention by improving employee morale and reducing churn rates

• ultimately improves competitiveness and increases revenue

Redressing the balance

So, what can we do about it? Here’s a summary of my thoughts:

• Address the issue at source — the government must continue to invest in improving gender diversity of STEM subjects within schools and universities. There’s already been a 50% increase in the number of women accepted onto STEM-related degrees between 2011 and 2020 (gov.uk, 2021), but more must be done.

• Initiatives aimed at women — entities like The Alan Turing Institute connect women within data science, generating a like-minded community that shares valuable resources to develop data careers. Creating partnerships with such organizations shows businesses genuinely care about achieving gender diversity. 

• Tackle inherent bias — people’s perceptions around data science are undoubtedly affected by cultural stereotypes. I must admit, I believed in the prevailing image of a ‘nerdy’ male data scientist poring over spreadsheets on a screen all day. However much you love a pie chart or Venn diagram, there’s much more to a data career than the usual media portrayal.

• Overhaul talent management — recruitment processes should be non-biased and anonymised. In addition, junior female data scientists should be linked with senior mentors who have the experience to provide encouragement at critical points in their careers.

Ultimately, our collective goal should be to shine a spotlight on the innovative, exciting, and transformational power of data, and its ability to solve real-life problems every single day.

At RAPP, we’re closing gender gaps by fostering an open company culture that champions diversity and is backed by strong commitment from senior management. 

Internal initiatives, such as DISCO (Diversity and Inclusion Steering Committee), promote an intersectional mindset via education, raising awareness and celebrating diversity. Flexible working practices are in place, which is instrumental in helping both men and women with young families stay in post and flourish.

Drastic changes are happening to address the gender imbalance within data science. It’s noticeable that more women are coming into the industry, but it will take time for this to filter through — especially to senior roles.

The gender gap is a serious issue. If ignored, it will compromise economic and sustainable growth. It’s predicted that improving gender equality within STEM industries could improve global GDP by $12tn over the next four years (source: McKinsey, 2015). We have much to do to get there, but it’s possible if we all pull in the same direction.