Dylan's Desk: How to improve diversity without compromising on excellence

Silicon Valley often prides itself on being a meritocracy, where people advance solely because of their talent.

Yet it has an obvious diversity problem. Big companies like Google, Twitter, and Facebook regularly release diversity statistics that show clearly that women, blacks, and Hispanics are underrepresented. Venture capital firms are populated largely by white and Asian men, and the companies that get funding from VCs are disproportionately similar.

And, yes, tech journalism has a similar skew: Too many of us are white guys.

It might be that there’s a smaller pool of talented engineers, entrepreneurs, and tech journalists among women and minorities. That’s why the argument is often framed in terms of meritocracy vs. diversity or excellence vs. affirmative action.

But it’s also quite possible that there’s something wrong with the recruiting process and that the lack of diversity is actually getting in the way of hiring the best people.

Orchestrating excellence

This is where a lesson from classical music might come in handy.

Orchestras in the US used to be 95 percent to 100 percent male and zero to 5 percent female.

But after instituting blind auditions, with the applicants performing their music behind a screen so they can only be heard, not seen, that ratio changed dramatically.

According to one study, the number of women in top orchestras rose from less than 5% to 25% after those orchestras implemented blind auditions starting in the 1970s and 1980s. One quarter to one half of that change, the study found, is attributable to the blind auditions, which force auditors to focus on what they’re actually hearing, not what they see.

It’s not just music, either: Double-blind reviews of scientific papers have increased the number of female authors in professional publications, according to a story in the San Francisco Chronicle earlier this year.

Recently, people have started suggesting that Silicon Valley needs to make a similar change. Startup guru Eric Ries made a similar experiment by removing names, gender, and ethnicity from résumés.

And Google has made its own efforts to tackle unconscious bias. Even well-meaning people sometimes skew their judgements unconsciously, because of shortcuts our brains have internalized over a long period of time.

The fact that Microsoft CEO Satya Nadella can so easily put his foot in his mouth with an ill-considered offhand comment shows that unconscious bias is real. It’s not that he meant to say anything patronizing or belittling to women; he probably didn’t even think about it.

Despite the best efforts of many well-meaning people in the tech industry (and beyond), women still earn less, get their startups funded less, and find their way into the ranks of VC firms less than men do. And that’s not even considering other aspects of the tech diversity problem: race, sexual orientation, religion, or political leanings.

People just naturally tend to gravitate towards people they are comfortable with, and that creates a self-reinforcing circle of sameness — unless we take deliberate steps to break out of it.

Focusing on excellence

My goal is to widen the pool of applicants but without compromising on quality. I want to hire the best possible writers and reporters.

I was inspired by this guest post on VentureBeat about how one startup used anonymous applications to recruit job applicants. So I thought we’d try an experiment of our own: Conduct blind auditions for job openings at VentureBeat.

We currently have several openings for ambitious, motivated tech journalists. (Plus we’re looking for a copy editor.) I created an online application that I’m asking people to fill out if they’re interested.

We’ll take all the applications, reformat them from an a relatively difficult to read Google Spreadsheet into a clean-looking Google Doc (using this handy script), remove the names, and evaluate them that way.

For URLs of published clips, I use Instapaper to reformat the articles into a neutral, readable design. For sharing with other editors, I paste them into Google docs, minus their bylines and publication names.

It might feel a bit impersonal, but this gives us the ability to do an initial review of all applicants based on the quality of their content, not their name, gender, race/ethnicity, or how well-designed their portfolio site is.

Will this make a difference? I don’t know. Ultimately my goal is to make hiring decisions based on excellence and ability.

Anything that lets us focus on the quality of an applicant’s writing and reporting — not his or her background, name, or face — seems like a good start.