By Thomas Lee San Francisco Chronicle.
It has been one of the greatest paradoxes in Silicon Valley that a region known for its progressivism and almost messianic belief in the power of code to cure just about any social ill can't seem to hack a solution to a social problem that it sees every day: the lack of women and minorities among its ranks.
Not that the issue is going to go away anytime soon. The media has recently latched onto the idea that the tech industry needs more diversity, publishing story after story on the hardships and outright sexism of the valley.
And affirmative action is again a national talking point after a divided U.S. Supreme Court this week ruled that states can bar public universities from raced-based admissions.
Enter Entelo, a software startup backed by Menlo Ventures and Battery Ventures. Next week, the company plans to start a service that allows companies to identify potential candidates who are female, Latino, African American or military veterans, among other groups, by mining data from social media sites like Facebook, Twitter and LinkedIn, or profession-specific networks like GitHub, a popular destination for engineers.
Race and gender are touchy subjects -- especially in the business world. But CEO Jon Bischke says Entelo's business clients came up with the idea. Considering the startup's current customer list, it's not hard to see why. ESPN, Google, Facebook and Salesforce are all making meaningful efforts to recruit a diverse workforce.
"The one thing you hear the most on Silicon Valley from the fast-growing technology companies is that they want to hire female engineers," Bischke said. "They know that most diverse teams are more effective, more productive, and the right thing to do. But they struggle with it because more engineers are men, men who know other engineers are men, and so the teams are lopsided."
According to a recent study by the Center for Talent Innovation research firm in New York, 32 percent of women in science, engineering and technology say they are likely to quit their jobs within a year partly because of "isolation."
Women in those fields "no longer find themselves the sole female on a team or at a site," the report said. "Yet they still feel excluded from 'buddy networks' among their peers and lack female role models."
As a general rule, I find that the best candidates are not the ones waving resumes outside your window screaming "Hire me!"
Instead, companies should recruit successful people who already enjoy their jobs -- though tracking them down can be difficult precisely for this reason.
Which is why companies often ask employees to recommend their pals. But since we tend to primarily associate with people from similar socioeconomic backgrounds, it's reasonable to assume that a tech firm's white male engineers often recommend other white male engineers.
Entelo has created an algorithm that determines the probability that potential candidates come from a certain background by analyzing the data they upload to social media sites. Bischke won't say what exact factors influence the results, but cited some intuitive examples like membership in certain professional organizations or whether they attended women's colleges, like Barnard, or historically black universities, like Howard.
This sounds kind of "well, duh!" but the process is not so simple. Despite the gobs of information heaped onto social networks, most people don't self identify on the Internet, Bischke says.
"It's a great question. I think part of it is the social networks are not even structured to ask those questions," Bischke said. "We don't want to know that information."
Ironically, race and gender are sensitive issues even in the era of personal branding. Based on your Facebook page, I'm more likely to know that you prefer half and half in your morning latte than your ethnicity.
Bischke said Entelo might some day use image-recognition software to help identify candidates, but that technology is still a ways off -- and frankly, kind of creepy.
To prevent running afoul of employment law or helping companies with major ethical shortcomings, Entelo designed the software in a way that prevents companies from excluding specific groups from searches.
If it works as advertised, Entelo could become a vital hiring tool. I'm frequently one of only a few Asian Americans in the newsroom, even though there are plenty of talented minority journalists out there. But even as the country becomes increasingly diverse, the percentage of minorities in newsrooms has fallen.
Often managers, mostly white and male, tell me they would love more diversity but just can't find any qualified minority talent.
Here's hoping some code can help solve that problem.