The data science team over at home-sharing platform Airbnb was able to increase the overall ratio of female employees on its team from 15% to 30% in 2015, with 47% of its new hires being women. That didn’t happen by accident. After scaling its team in terms of hyper-growth over five years, Airbnb’s data science team found that only 10% of the new hires were women. That’s when it became clear that the team was not being mindful of diversity, Airbnb Head of Data Science Riley Newman told TechCrunch.
So, coming into 2015, Newman and Airbnb Data Science Manager Grewal started thinking more about gender diversity in data science. It was just one dimension of many aspects of diversity Airbnb especially wanted to become more mindful about, Newman said.
The team did this by analyzing the equal employment opportunity commission data in Airbnb’s recruiting tool, Greenhouse, to better understand the diversity of the team’s applicants. In that process, the team noticed that they were getting a lot of female applicants, but they weren’t converting those applicants into actual hires.
“That said to us, unlike in other fields, it wasn’t that there are no women applying for the role. We have a significant number of women coming into our pipeline,” Newman said. “We wanted to correct for an imbalance that occurred.”
So, the team redefined the interview process to make it more gender-blind by removing the names of people on projects, as well as making it more welcoming to women by ensuring women were in the room during in-person presentations.
“On site, typically in past, because the team was predominantly male, a woman would come in to present in front of a panel of men,” Newman said. “That can be a really intimidating environment.”
The data science team also started a series of talks called “Small Talks, Big Data” to inspire women outside of Airbnb to apply for data science jobs. That’s because they noticed that there weren’t as many women in data meetups as there were women in tech meetups.
Companywide, Airbnb is 46.3% female, 63% white, 7.1% Latino/a and 2.9% black in the U.S., according to its latest diversity report. The data science team is 6% Latino/a and a whopping 0% black. While the data science team is proud of the progress its made with women, it’s not happy with the overall numbers and recognize that they’re nowhere near good enough. This year, the data science team will focus more on achieving racial diversity on its team.
“In 2016, [race] is going to become another area of emphasis for us,” Newman said. “The things we’re looking into right now is whether the same tactics will have the same effects. We certainly want to bring as much of the playbook that we have made over the last year to bear on some of these issues.”
Source: http://feedproxy.google.com/~r/Techcrunch/~3/luA0xXpAu2M/
