The first story Ellen Pao told was about a law firm partner who stood outside a female lawyer’s office everyday after lunch, licking an ice cream cone while staring at her. Management’s response to these instances of sexism were invariably dismissive, e.g. “you should be flattered he’s interested in you,” or “it’s just one weird guy, but overall it’s not that bad, right?” The result is an environment that views sexism as the product of just one bad apple, but it’s instead the product of a bad system: one which condones and encourages “boys club” behavior. Staffing decisions were made at all-male steak dinners and strip club visits, which reifies the structural exclusion present within VC and tech.

It’s now clear that tech has a diversity problem, given the wide range of stories over the last few years: Amazon, Susan Fowler at Uber, Leslie Miley at Twitter, SoFi recently, and almost certainly more we haven’t heard. Having these voices speak out make the structural problems impossible to ignore. Recognizing that tech companies exist in (and perpetuate) an ecosystem of exclusion is an important step to take in ensuring that workplaces are safe places for all employees.

The latter half of the talk covered strategies of inclusion. In business school we learned the Kotter Model for organizational change, which I think is compatible with most of what Pao talked about. A few highlights:

• At Reddit, it was useful to have the support of Yishan Wong (Reddit’s previous CEO), who had the respect of the engineering staff, when instituting initial reforms. This is similar to recruiting a guiding coalition who can offer support, tailor messages, and create buy-in within the community. It’s often not worth it to work with a CEO/executive who needs to be convinced that diversity and harassment-free workplaces are important.
• Just as we have metrics to measure business goals, we need good metrics for diversity goals. These metrics go beyond just “number of women hired in the company, relative to peers” to encompass hiring, retention and promotion, as well as measuring performance across all functional groups and all minority groups. Airbnb had a good case study where the data science team examined the recruiting funnel to see where women were entering and dropping out, and made changes so that half of new data science hires are women.
• Finally, we talked basically about intersectionality. How do we create an environment that is welcoming to all people, regardless of age, nationality, orientation, and more? To do so we need to focus on total inclusion, seeing people as people rather than numbers to hit in a spreadsheet. Within the Kotter model, this would be the last step of instituting lasting change.

Notes from Ellen Pao, 27 September 2017.