Meet the Exec Pushing Google to Fix Its Dismal Diversity

Meet the Exec Pushing Google to Fix Its Dismal Diversity

GOOGLE ISN’T GETTING very far. This week, the search giant released its latest diversity report detailing the demographics of its 53,600 full-time employees at the end of 2014. There’s been little progress despite a year’s worth of initiatives. Women still hold roughly one-fifth of the company’s tech jobs and leadership positions. African-Americans make up 2 percent of the company and Hispanics 3 percent, both unchanged from last year.

“Early indications show promise, but we know that with an organization our size, year-on-year growth and meaningful change is going to take time,” said Nancy Lee, Google’s head of diversity strategy, after the numbers were made public.

Lee, who sat down with WIRED before Google released its latest diversity figures, was instrumental in persuading the company to release such data for the first time last year. Other big tech companies, including Apple, Facebook, and Twitter, followed suit. The provocative move is one of many tactics Lee has undertaken in an effort to break through the tech industry’s deep biases.

Such strategies are sorely needed. It’s no secret that tech has a diversity problem. You see it at conferences, in headshots of company executives, and indeed, in the employee demographic reports that have started trickling out of Silicon Valley, the first of which came from Google. As appearances suggest, the workforce is still overwhelmingly white and male.

More recently, the groundswell of public discussion has put increasing pressure on companies to make diversity a priority—and not just for altruistic reasons. Over and over again, research has shown diverse teams are more creative and perform better financially. As technology continues creating products that change the way we interact with the world, the thinking goes, we need people with diverse views working in the field. Otherwise, tech products could end up only serving a narrow population with a narrow set of concerns.

Google’s diversity strategy faces especially close scrutiny. Not only has the search giant long made a name for itself as a company unafraid to push the boundaries of tech, it’s one of the few in the world that has a wealth of people and data to run innovative experiments. If Lee’s efforts are successful, they’re sure to resound through the industry. The question is whether they’ll be enough.

Google’s Out-of-Control Growth
When Lee was 16, growing up in a single-parent home in Sacramento, California, she took a job at Chuck E. Cheese, assigning birthday hostesses to children’s parties. During one party, a parent approached Lee to complain about the hostess Lee had picked and ask that she be removed because she was an older woman with an accent the children had difficulty understanding. The hostess broke down in tears, and Lee tried to defend her. The manager stepped in to grant the woman’s request while telling Lee, “The customer is always right.” Lee was suspended.

The experience stayed with her. Lee eventually earned a law degree at the University of California at Berkeley, then went to work for a San Francisco law firm, where she practiced employment law. In 2006, she brought that specialty to Google, when the company, still less than a decade old, had about 5,000 employees and was in aggresive growth mode.

Lee and her boss, Laszlo Bock, Google’s senior vice president of “people operations,” talked about building “fairness functions” at Google. In 2013, Lee took on her new role leading Google’s diversity efforts, where she set about compiling the company’s diversity statistics.

“My sense was that we were never going to have an honest conversation about what we needed to do until we knew where we actually were,” Lee says.

Where Google was, it turned out, wasn’t so great. Part of the problem, Lee says, was that in its early days, Google was expanding much too quickly. As the company experienced exponential growth, its ways of doing things didn’t keep up. “I think because we were growing so fast, we just hadn’t been that deliberate about what we actually looked like as a workforce,” Lee says.

But just knowing wasn’t enough. Lee wanted to share the information publicly, hoping that shining a light on the data would spur the industry to look at the issue honestly. After working on a communications plan for about a year, Google in May 2014 became the first major Silicon Valley tech company to release its diversity statistics. As suspected, they weren’t pretty: globally, 30 percent of employees were women and only 17 percent of its technical employees were women. Of Google’s US employees in 2014, 61 percent were white, 30 percent Asian, 2 percent African American, and 3 percent Hispanic.

More Work to Be Done
A year later, those numbers have hardly budged, despite what Google said was $115 million spent on diversity initiatives. This year, Lee says, Google plans to spend $150 million more.

Among its efforts, Google is embedding engineers at historically black colleges to help rethink computer science curriculums and recruit interns. It’s creating tools for kids to learn to code and working to help minority-owned and women-owned small business owners build a strong online presence. Inside the company, Google runs “bias-busting workshops” and has rolled out a program called Diversity Core, which allows employees to set aside time at work devoted to diversity projects within Google and in local communities.

The company is also changing how it recruits at the university level so more women and minorities can find a path into the company. Historically, Google has recruited from a pool of about 100 colleges—mostly top-tier schools—but in 2014, the company more than doubled that number.

“When we looked at the research, the demographics of diversity of student bodies at some of the most elite colleges where we tend to recruit was more homogenous than the general college population,” Lee says. “So we started targeting more schools.”

Now, almost 20 percent of the company’s hires out of university come from these new schools, including non-Ivy League standouts such as Alabama A&M, and the University of Missouri-Columbia.

And because this is Google, the company is taking a data-driven approach to diversity and bias. Lee says runs analyses on, for example, the distribution of salaries at the company to ensure there are no glaring gaps among employees of the same level. This pairs nicely with some of Google’s more low-tech approaches, Lee says, including a calibration meeting among managers before evaluating an employee, so that people are made aware of and primed for the possibility of biases—the halo effect, for example, or the recency bias—creeping in during the process.

An Evidence-Based Approach
Joan Williams, a professor at University of California, Hastings College of the Law who specializes in women’s issues in the workplace, says that many of Google’s initiatives are laudable. But she thinks it could still be doing more—strategies that, in fact, sound uniquely well-suited to Google.

“Instead of scattershot training about lots of different forms of bias, or imagining scenarios of things that could be happening, they could be doing things to analytically figure out what the problems are within the company, and address those specific patterns,” says Williams.

Google, she says, could ask employees if they think bias is occurring, then put a “bias interruptor” in place. After a period of time, it could measure whether its metric of bias has changed. For instance, in some work environments, women have observed that they are more likely to be asked to do the “office housework”—taking notes, planning events, and other tasks that prevented them from meaningfully participation in certain activities. One gentle bias interruptor could be to simply check in with women if they feel something of that sort is happening, Williams says. If so, she says, a company could make such a survey public to discourage the behavior.

“Google is in an extremely good position to lead on these issues because not only is it a large company, it has a lot of money,” Williams says. “It can commit itself to an evidence-based approach to mitigating bias.”