Tracy Chou speaking at Inspirefest 2017. Image: Conor McCabe Photography
Tracy Chou speaking at Inspirefest 2017. Image: Conor McCabe Photography

Diversity and the data-driven workforce

10 Nov 2017

Tracy Chou explains how the tech industry can tackle diversity using the same method it applies to other issues: data-driven decision-making.

This week, the World Economic Forum (WEF) – which has been tracking gender parity around the world since 2006 – published its latest Global Gender Gap Report. The results are disheartening, with the WEF reporting that, for the first time in its measurements, the gap had widened.

This annual report tracks four key metrics across 144 countries, with a myriad of subcategories using databases, surveys and calculations from global reports as source material. The data is very clear on the world’s gender diversity problems.

‘Without data, there was no language for describing the magnitude and the pervasiveness of the problem, and it was too easy to dismiss’
– TRACY CHOU

Demanding the data

When Tracy Chou decided to instigate action on diversity in the tech industry, she turned to data. In 2013, while working as a software engineer at Pinterest, she wrote a Medium post that challenged tech employers to reveal the number of women engineers they had on staff, demanding that these figures not be hidden away.

“I had been struck by the hypocrisy of an industry in which the problem of the lack of diversity was an open secret and yet no companies were willing to measure and manage it. An industry in which everything is supposed to be data-driven, and yet there was no data on diversity,” she told the Inspirefest audience in Dublin this summer, explaining the motivation for her step into activism.

“Without data, there was no language for describing the magnitude and the pervasiveness of the problem, and it was too easy to dismiss.”

Voice from within

Though it’s quite simple to draw a line from Chou’s data-led diversity activism to the burst of diversity reports that subsequently issued forth from the leading tech companies in Silicon Valley and beyond, she is quick to acknowledge that she was not the first to give voice to the issue.

“There had been people who had been calling for the release of diversity data in the past but oftentimes those were people that weren’t within the industry,” she explained to Claire O’Connell in the latest episode of Inspirefest: The Podcast.

“I think there’s some cachet I had from being an engineer within the industry, and I think some of my credentials of having grown up and worked in Silicon Valley and having attended Stanford right there in the heart of Silicon Valley helped to give more credibility to my call to action.”

Thus, Chou struck a chord by tuning her message to the key of data – the driving force behind most actions taken in the tech industry.

Building the case for diversity

Chou cites the WEF in her Inspirefest keynote, which you can hear in the podcast episode or check out the full video below. The WEF, Chou tells us, hinges its argument for diversity on three pillars: the business case, the consumer case and the talent case.

The business case for diversity has been thrashed out extensively. In brief, per Chou’s keynote, “diverse teams are more creative, more diligent and thoughtful, and, most importantly for the business case, drive better financial returns” – and there is plentiful research data to back this up.

The consumer case is less discussed but just as pertinent. “We’re building products for everyone. The quality, relevance and impact of these products and services can only be improved by having the people who are building them be demographically representative of the people using them,” said Chou.

‘Some use cases are just not encountered by the people who are currently building tech’
– TRACY CHOU

A prime example of the missteps that come from product teams incapable of taking different types of users into account is Apple HealthKit, billed as the ultimate app for the quantified self-tracking all aspects of personal health – but without a period tracker. A “glaring oversight” noted by Chou.

Fleshing out this idea in the podcast interview, Chou said: “Without all the perspectives of people who are using these products, I think it’s very easy to overlook these different cases. I don’t want to even call them ‘edge cases’ because some of these are just very standard cases, they’re just not encountered by the people who are currently building tech.”

Tackling the talent shortage

Finally, then, there is the talent case – and more figures to put this problem into perspective. This time, Chou cited the [US] Bureau of Labor Statistics’ projection of a shortfall of 1m workers to fill open computer science jobs by 2020. “The tech industry severely underutilises the talent that exists in our economy with its systemic exclusion of certain sectors of the population,” Chou added. “Talent is evenly distributed, opportunity is not.”

Applying an analytical mindset to the data at hand, Chou puts forward new insights on the issue that dispense with some of the rote responses to date. Chiefly, companies will refer to the pipeline issue – which Chou acknowledges – and use this as a reason for their skewed male-to-female hiring ratio. You can’t hire qualified women if they don’t exist, right? And it’s not up to industry to solve academia’s problems.

Chou does not allow the issue to be so easily brushed aside, however. Citing NPR’s exploration into ‘When Women Stopped Coding’, she identifies the peak of women majoring in computer science at university in 1984 and pointedly states: “If there were so many women entering the field in the ’80s, it’s a bit disingenuous to say that the lack of women in leadership is a result of pipeline problems.”

‘When hiring managers complain that they can’t find qualified female candidates, it often means that they’re looking in the wrong places or looking for the wrong things’
– TRACY CHOU

Chou also deftly dismissed the notion that diversity in hiring practices means “lowering the bar” as “sexist and racist”.

“First of all, there is no bar. We’re not doing the high jump here,” she said, reminding us that the measure of a job candidate is never straight up and down, but a multidimensional evaluation of characteristics, some of which are neither good nor bad, just present.

For those companies struggling to find women or other minorities to hire, Chou suggested they look harder. “When hiring managers complain that they can’t find qualified female candidates, it often means that they’re looking in the wrong places or looking for the wrong things. For example, the obsession with hiring software engineers out of Stanford and MIT means that these schools are very over-fished small ponds and it’s hyper-competitive to sign them on – even when there are amazing pools of talent elsewhere. And whether or not someone has a brand name CS degree has very little bearing on their ability to do the job.”

Diversity debt

Chou effectively uses data to define the tech industry’s problem and she also advocates for the use of data in taking action. “You can’t manage what you don’t measure,” she summarised.

In the years since her Medium post ignited a spark, she has become co-founder of Project Include, a movement helping companies to better manage diversity – and not just in terms of gender – from the outset.

“A metaphor that one of my Project Include teammates likes to use is that of debt. Sometimes you need to take on debt just to get going and it’s a trade-off in favour of the short term versus the long term. It’s OK to make that trade-off deliberately but, the longer you wait to pay off your debt, the harder it gets to do so because the debt compounds. This is true of financial debt, technical debt and very much so of diversity debt.

“It is much harder to change an organisation of thousands, or tens of thousands or even more, than it is to set the right cultural norms and organisational processes in a team of five or 10, and this is why we’re so focused on start-ups getting things right from the beginning,” she said during her keynote.

Healing 1,000 paper cuts

Describing her keynote as a “Diversity 101” introductory lecture, Chou concluded by addressing the size of this challenge and the multiplicity of actions required to fix it. She described the struggle of women and minorities in the industry as “death by 1,000 paper cuts” and added that “countering those thousand paper cuts means lots and lots of little interventions”.

In sum, Chou said: “Diversity is important. It’s smart for business, not just a feel-good endeavour. But how we achieve diversity means confronting the deep and uncomfortable reasons for why we don’t have it now. It’s not just a pipeline problem (although, of course, there is a pipeline problem as well). There are structural impediments for women and minorities in this industry. To correct course – which means changing culture and process – will require a lot of hard work.”

Inspirefest is Silicon Republic’s international event connecting sci-tech professionals passionate about the future of STEM. Super Early Bird Tickets for Inspirefest 2018 are on sale now!

Inspirefest: The Podcast is available wherever you get your podcasts. Subscribe now via Inspirefest.com.

Elaine Burke
By Elaine Burke

Elaine Burke was editor of Silicon Republic until 2023, and is now the host of For Tech’s Sake, a co-production from Silicon Republic and The HeadStuff Podcast Network. Elaine joined Silicon Republic in 2011 as a journalist covering gadgets, new media and tech jobs. She later served as managing editor before stepping up as editor in 2019. She comes from a background in publishing and is known for being particularly pernickety when it comes to spelling and grammar – earning her the nickname, Critical Red Pen.

Loading now, one moment please! Loading