Tracy Chou is a software engineer at Pinterest, a job she apparently unabashedly loves. "It’s the first place, in school or professionally, that I’ve not been aware or made aware of my gender, ever, in any situation," she's written. "I don’t feel like a female engineer." But to Chou, in order to figure out exactly what to do about the limited number of women in tech, women in tech need numbers. So she's looking for them.
On Monday, Chou wrote a post on Medium entitled "Where are the numbers?", praising the annual Grace Hopper Celebration. The GHC is named after Grace Hopper, a computer scientist who helped shape the field at a time when there were even fewer women in it than there are now. GHC brings women in tech together for a series of conferences once a year and, since its inception in the 90s, has slowly boasted rising attendance. Currently, the number of women graduating from college with computer science degrees hovers around 13%, according to the Computing Research Association. But Chou says that, in her experience, the actual number of women in the field "are far lower than anybody is willing to admit" and that most companies have "some way of hiding or muddling the data on women actually in engineering roles":
This means nobody is having honest conversations about the issue. While companies do talk about their initiatives to make the work environment more female-friendly, or to encourage more women to go into or stay in computing, there’s no way of judging whether they’re successful or worth mimicking, because there are no success metrics attached to any of them.
But it’s hard for companies to admit that they’re working through any issues when they want to paint the rosy picture for recruiting. The competition for tech talent is fierce, and companies are inevitably pitted against each other fighting for good candidates, even more so for good female candidates because of their paucity. While it’s ok to say in a generic way that the industry has a long way to go, companies always need to at least intimate that they’re doing better than average.
Chou's tried to do just that, using Github to crowd source user-submitted data on the gender breakdowns of the engineering teams at various companies. She's then moving that information in a Google spreadsheet. So far, there's information from companies like Mozilla, Dropbox, Rent the Runway and Reddit. The chart's categories have also prompted conversation about whether there should be information included about people who identify as genderqueer or trans. Right now, Rent the Runway, Highfive and Medium are "winning" the gender race; they all have engineering teams that are over 20% female.
I emailed Chou to ask her what she'd like to do with the information when she gets more of it. "Once we have the data, the goal is for the community to draw actionable insights from it, and then to act on those insights," she said, adding that she'd like the companies that aren't doing well to take lessons from the companies that do seem to have hired a number of women:
One idea is to look for patterns in the companies that are benchmarking significantly above average, and see if they have specific initiatives or processes worth emulating at other places. Early data seems to suggest that smaller companies are doing better than larger companies, which is a little counter-intuitive, and also worth investigating.
Chou's potential plan to involve companies sounds like it has a good chance of working, given that it has less to do with shaming them into actually being honest about their demographics, and more to do with figuring out what they could be doing better. Right now, there are multiple companies listed that have not a single woman on their engineering teams, and others don't have numbers that are a whole lot better looking. (For instance, Dropbox has only 9 women out of 143 total engineers.)
Chou also noted that though the total percentage right now is actually pretty positive (of the companies counted, women make up 10% of the engineers, just slightly less than the number graduating from school with a degree in their chosen field), those numbers "are likely skewing high due to self-selection bias." It's interesting to see these numbers roll in but more interesting still will be to see what kind of change this evolving project sparks – if it even does.
Image via Evening Standard/Getty