Achieving a diverse workforce is a goal that has received increasing attention from the tech industry in recent years, though the problem of homogeneity persists. Despite that, the business case for diversity in tech employment has been consistently reinforced. A December 2023 report from McKinsey found that despite a rapidly changing business environment the relationship between leadership diversity and company performance is the strongest it has ever been.

In the Diversity Matters series, McKinsey tracked the diversity of leadership teams in terms of gender and ethnicity and the financial performance of those companies. In the 2023 report, companies in the top quartile of leadership teams with women representation were nearly 40 percent more likely to financially outperform teams with fewer women in place. When McKinsey began the research in 2015, the gap was only 15 percent. When ethnic diversity was considered, companies in the top quartile in terms of diversity of leadership were just as likely (39 percent) to outperform firms that had the lowest diversity numbers. Both forms of diversity in executive teams appear to show an increased likelihood of above-average profitability, the research shows.

If the business case for diverse teams and leadership is so sound, how do we get more representation in tech, in more leadership positions, from a wider variety of humans?

Encouraging Diverse Founders

According to Dr. Alexis Wichowski, part of the answer will turn on who decides what gets built. Wichowski, a professor of professional practice at Columbia University noted during a keynote address at the recent Data Universe event in New York City that the foundational principles guiding the evolution of technology included democratic principles like free and unfettered access to computers. While that implied that unfettered access would also be available to those who develop tech,  that wasn’t what actually happened.

She explained that the founding cohort of the web era companies that dominate the landscape is a very homogenous group. The “Gods of Tech,” as she refers to them, were young, white, affluent men who were solving problems and they directed their prodigious energy and intelligence toward the problems they saw and cared about. So, the solutions that received investment were most often limited to the lived experience of a homogenous, wealthy group. In other words, the diversity of founders matters.

A more diverse group of founders will make more decisions based on more experiences. Opening investment to a wider variety of founders solving for a wider variety of problems affecting a wider variety of people could also attract a more diverse workforce. And, the older, whiter cohort that built the industry, Wichowski says, will be instrumental in making connections and driving investment in a new era.

“Think about trying to figure out ways to connect the experienced startup founders, maybe who've tried a few times and have learned the hard way what doesn't work and are ready to launch their next venture,” she advised the Data Universe audience. “And connect them with people who have different kinds of perspectives in life. So, for all of us out there who are not 26 years old, uh, don't give up hope. There's a lot that you've gained from your experience that you can hopefully apply to the next generation of tech that we build.”

Building More Ethical Alogrithms

At the same event, a panel discussion on ethical algorithms also touched on diversity within data science. Renee Cummings, a professor at the University of Virginia’s School of Data Science, agreed with Wichowski that the diversity of the data scientists creating the algorithms affects the impacts those technologies have on the people using them.

“I think it's critical to consider who can participate in the creation of AI algorithms,” Cummings said during the Data Universe discussion. “A much more diverse group of people are asking, ‘How could I become a stakeholder?’ Because we know one of the most critical things to the advancement of this technology in an ethical way is building stakeholder capacity—stakeholders who are educated about the technology and who are able to bring the requisite level of assessment and evaluation to the ways in which this technology is being deployed.”

The organizations that are more inclusive are the ones that are asking the right questions about how AI is impacting users, she says. More diversity in the ranks of the data scientists developing AI algorithms results in more consideration of those impacts during development, not after. A more diverse team will consider things like possible long-term impacts or ways the product could cause crisis earlier in the process.

And, while some headway has been made around including more gender and ethnic diversity, Cummings noted that “reimagining” around diversity is necessary.

“One of the things that we continue to see in our data sets is the lack of representation when we're thinking of individuals with disability or when we're thinking about neurodiversity,” she said. “Representation is so critical to the ways in which we need to think about data and think about AI and in particular, generative AI.”

‘Move Swiftly’

As the demographics of college students continue a generations-long shift toward more representation of diverse genders, ethnicities and more, the diversity of the tech world will continue to slowly grow from the bottom up.

But the lessons Wichowski and Cummings were teaching the Data Universe audiences show that the benefits derived from a more diverse workforce will truly accelerate and be realized when that shift reaches the ranks of decision makers.

“While we do have diversity, in some respects, we really have got to move swiftly and at scale to ensure that we make this technology inclusive,” Cummings concluded. “We're realizing the impact on society is so extraordinary that, if we don’t, we can be running after huge rewards, but also creating significant risks that we may not be able to deal with.”