The exclusion and under-representation of women in technology (tech) innovation, design and development is resulting in digital tech products that appear skewed towards the male experience and are not representative of a diversified society.
Having come about under strong male leadership, many of the foremost tech and digital innovations now commonplace in our daily lives have been designed by men, for men – although not necessarily intentionally.
Despite phenomenal growth in the global tech industry, women are inexplicably under-represented in the cohort of successful tech entrepreneurs. Despite constituting 49.6% of the global population, only around 25% of tech jobs globally are held by women.
This is mirrored by the top five tech companies in the US – Google, Apple, Facebook, Amazon, and Microsoft – of which only 25% of employees are female. Alarmingly, a study conducted by PWC puts this figure at 19% within the world’s top 10 global tech firms.
“The inclusion of more women in tech and the achievement of gender equality will result in products and solutions that are designed to be more inclusive and are able to help a wider audience than before,” explains Lara Du Plessis, head of Ecosystems at Finch Technologies.
Data from the Kauffman Foundation clearly shows that investing in women in leadership positions has tangible benefits for stakeholders.
Moreover, private technology companies led by women appear to be more capital-efficient, achieving a 35% higher return on investment (ROI), and, when venture-backed, have 12% higher revenue than start-ups run by men.
Including more women in tech and making the sector more gender-equal can thus lead to an increasing number of thriving businesses.
“From a profitability perspective, women play a fundamental role in the development of the tech sector, and their inclusion and involvement in business decisions result in increased ROI,” Du Plessis says.
The gender data gap and our increasing reliance on AI
Adding further fuel to the fire is the gender gap relating to available big data, which is increasingly used to inform decision-making in tech innovation and development.
As data starts to play an ever-increasing role in shaping how the world develops, the transparency, validity and representativeness of the data on which we are so reliant have become critical. Those collecting data decide what and how to collect, and this can have important consequences for women.
"Equal representation is vital not only to the development of the data and AI used to make decisions but also within the teams that make those decisions in every organization.
“Women represent 50% of the world's population and if female customers are not represented where decisions are being made, you’ll fail to reach them in a way that resonates with and serves them," notes Thea Sokolowski, head of Marketing and Communications at Stitch.
Many organisations make decisions based on AI machine learning, in which algorithms learn from vast amounts of data to find patterns and make predictions.
A consequence of the core function of machine learning is that when you feed it biased data, its ability to be biased improves. This can lead to AI systems reinforcing and intensifying existing harmful gender prejudices.
As such, businesses should make a conscious effort to collect sex-disaggregated data at the inception of every tech innovation process.
Deep-rooted gender biases that exist in the data we produce are the result of the lives of men being used to represent those of humans overall – not as a deliberate strategy, but rather the product of a method of thinking that has prevailed for many years.
Being aware of the gender data gap, and no longer taking each statistic and ‘fact’ at face value will ensure that the data used to build tech products and services is free of gender biases and discrimination.
Historical examples of male-centric design
The design of seatbelts has a basis in male-centric design, which renders the level of safety provided for men higher than that for women.
When initially introduced in the 1950s, car crash test dummies were modelled on the measurements of the 50th percentile male, assuming that the 50th percentile male would be representative of the 50th percentile human.
While over time, female crash test dummies have been introduced, these are merely scaled-down versions of the original male dummy. The result is that women are at greater risk of dying or being seriously injured in a car accident than their male counterparts.
A second alarming example is that ‘typical’ warning signs of deadly health conditions are actually common symptoms for males.
According to a Harvard study, women are less likely to survive their first heart attack than men due to the fact that the symptoms differ between men and women and we are predominantly taught to observe for the male-specific symptoms.
The historical lack of female representation in the design of biological and medical education is largely to blame for outcomes of this kind.
Many studies mentioned in the ‘limitations’ sections of their papers that the female bodies were deemed “too complicated and unpredictable” to include.
Voice recognition software is predisposed to respond to requests, follow orders and accurately comprehend instructions from a male voice rather than a female one.
Speech recognition is a form of AI that has been shown to perform less effectively for women than for men.
It is largely accepted that accuracy for male voices is still higher by around 10%. These disparities exist because of the way in which data analysis, databases, and machine learning have been structured.
Changing the narrative of women in tech
Gender equity – the process of being fair to women and men by providing both genders with what they need to succeed – is valuable in itself.
To ensure such fairness in the tech industry, strategies and measures must be put in place to compensate for women’s historical and social disadvantages:
– Enhancing access to educational opportunities in the Science, Technology, Engineering and Maths (STEM) space for girls is a major driver of increased female representation in tech. According to a PWC study, In South Africa, far fewer women than men graduate with STEM-related degrees, creating a far smaller female talent pool.
“In the tech world, we often get labelled as ‘good female developers’ and not just good developers. This can imply that we don't measure up to our male counterparts.
We need to get to the point where it is the norm and accepted that female and male developers operate at the same level. We should get recognised for our skills regardless of our gender,” asserts Sara Owen, team lead and Software Developer at Electrum.
– The second building block for changing the narrative around women in tech is the presence of female role models (and current lack of) in tech careers.
The more powerful are the female role models in the tech space, the more young girls will be inspired to take up STEM subjects and pursue careers in historically male-dominated fields.
Until corrected, male leaders in tech can also serve as role models for women in the industry, actively encouraging increased women participation and inclusion.
“For gender equality in the tech space to have a meaningful and lasting social impact, business leaders in tech must recognise that diversity is not merely a tick-box item, but rather a golden key to designing products that work for everyone,” says Christopher Ball, co-founder of Finch Technologies.
IOL BUSINESS