Why Do “Women in Tech” Communities thrive, while Women in Tech don’t?

It feels like every week I am invited to another new community for women in technology. I am a fan of these efforts and participate.

But…

• Diversity numbers at major companies don’t move, even with referral incentives.
• We keep ignoring the data we have on “what works” for diversity and inclusion. Many prominent communities still seem to have low participation by women.
• When I go to general tech events, I’m still told that no women are participating because they “aren’t interested” or they “aren’t qualified” (and people blame a whole lot of things).

In the best case, “Women in Tech” groups are a leading indicator. They’ll allow women to speak up, gain confidence, and then gain visibility and impact in other communities. Eventually (soon) more women will thrive across the board. This is what everyone likes to say.

But what if it isn’t true?

In the worst case, these communities won’t change anything. People will say women aren’t at their events because they prefer the “women-only” events and communities. Plus, other underrepresented groups (like men of color) will continue to not be served by existing groups, and also won’t be served by these new women in tech groups. It’s not an intersectional approach to this issue.

Based on what I’ve observed so far, that’s what I fear.

Why?

I see three major distinctions between “Women in Tech” communities and other communities that I think cause the difference in success. (By success I mean the rapid growth, wide participation, recognition of the groups). Of the three, none are easily transferrable to our existing groups in technology.

“Women in Tech” communities won’t cause a waterfall of changes for women in tech in other places. We’ll need to work on those separately. I don’t think it’s hopeless, but it’s an uphill battle. I’m not the only one who’s noticing this (and it’s probably not only tech).

Why are the “Women in Tech” efforts thriving, but actual women in tech aren’t?

(1) They have enough women early on.

When I lived in NYC, I ran a Product Management breakfast with my friend Steven. Early on, we managed with a spreadsheet and invited 10 people each week (inviting more if people were busy and couldn’t attend). We hand-curated the list to get people we thought would like each other. I also happened to know a lot of women. The breakfast group remains close to gender parity at nearly every breakfast. I got curious about this, and I think it might be because of the early balance.

The first 104 Product Breakfast invitees:

There were no unknowns in this group because I had access to the Mailchimp list.

Separately, I found out that my friend Amelia was doing a project for HBS about Product Hunt. She was curious about what Product Hunt would look like if the community had more women. She built her own (closed) community to observe qualitative differences. She’d also collected a lot of data to see who was posting products and how products targeting men vs. women fared.

I looked at the data she’d collected from Product Hunt through a different lens. I figured out as many of the first 100 members as possible, and looked at them by gender.

For the first 101 Product Hunt Members by User Number:

Some of the first 100 members became inactive and I was unable to see their name to do gender research. I’d love to know a little bit more about that to see if more women were likely to fall off as well. (I feel comfortable using Product Hunt as a comparison because it’s much nicer than most places on the internet. I doubt people left because they were abused/attacked as might be the case in other places. Also, I was a relatively early member of their community, #842).

At Product Hunt, this is possibly less than 10%, but also possibly up to 25%. These ratios held consistent from the first ten members for both Product Hunt and Product Breakfast.

We know that people tend to know others who are similar to them. When does your community already model steady state? The first two people? The first ten? The first hundred?

My first hypothesis is that women end up in most general communities “too late.” By the point they’re invited, they already feel like a token member, an outsider, or too different. Women in Tech communities don’t struggle with this. Women are by definition the very first members.

(2) Current “gender-neutral” communities aren’t built with women’s needs in mind.

If an early user base/community is mostly built mostly by and for men, they may not see what needs to be built to make the platform accessible for women. This happens with products (hey look, Apple has a period tracker finally), but can also with communities.

A great example of this in a community platform is how Safety & Abuse are handled on Twitter.

This made waves recently after an outpouring of hate toward Leslie Jones, Ghostbuster costar. Jessica Valenti, journalist, just left the platform after a rape/death threat towards her five year old daughter. While abuse may happen to many users, the type of vitriol and risk that accompanies trolling can be worse for women.

That said, this isn’t new. Charlie Warzel recently published a piece on Buzzfeed chronicling the full history of abuse. He brings forward a similar conclusion:

Looking back on Twitter’s early years, multiple former senior employees cite Twitter’s disproportionately white, male leadership — a frequent, factual critique of Silicon Valley’s biggest and most influential tech companies — as creating an environment where building tools to combat harassment was a secondary concern. “The original sin is a homogenous leadership,” one former senior employee told BuzzFeed News. “This is part of what exacerbated the abuse problem for sure — because they were often tone-deaf to the concern of users in the outside world, meaning women and people of color.”

“Getting it” also needs to happen at an individual scale. Jason Toff recently shared a thoughtful set of tweets about how his perspective on Twitter abuse changed when his wife was impacted.

Rationally seeing the additional needs of a subset of the community may not be enough. In addition to rationally understanding community challenges, there needs to be empathy — and too often that is only built through personal experience.

Unfortunately, until we figure that out, these problems can make it impossible to for certain segments of the population to join the community. Another good example is in Slack, as Sara Mauskopf recently brought up.

Slack is built as an enterprise tool, but has expanded greatly for community platforms. I left a prominent PM community after a member asked me if I was a recruiter. While one case isn’t a big deal, it signaled to me it might happen more, and I was tired of that implication. Sara’s suggestion of a tool is one way to solve this problem (and a good one).

That works for products, but many communities aren’t within products. A community may “think” it’s neutral, but not successfully be dealing with the behavior of bad apples who disproportionally affect a subset of the group (like women).

Communities need to establish ways to prevent that behavior — such as through norms, or rules. One such example is a Code of Conduct at a conference (that link will take you to way more background information from Ashe Dryden). Many of these came about after specific complaints from women for behavior at conferences. Yet, there will still protests from other attendees who claimed it wasn’t an issue.

Spaces that are designed specifically for women take into account women’s needs by default — that’s who is making the community, and who is participating. Other spaces must also think about ramifications that will disproportionately impact women if they want to great a community with more women.

(3) Women may have higher churn rates in communities.

This builds heavily on the first two. If we don’t have women included from the beginning of communities, and we don’t give them a safe space to exist in, they are likely to leave. This impacts individual communities the same way it impacts our industry.

If women aren’t included in the beginning, they are less likely to be a “core” part of the community. If you’re in the core of a community, it’s a much bigger decision to leave. That makes you less likely to churn. If you come in late, the discussion may feel crowded, or like there isn’t anything new to add. You leaving doesn’t feel like as big of a loss, and community members are less apt to convince you to stay or make changes.

If a community doesn’t work for you, you aren’t going to stay. You’ll churn out even if people technically tried to “include” you. In any community there will be bad apples, and the community needs to find a way to make that not disproportionately impact certain members.

Women may also be more likely to churn because they may be expected to perform additional emotional labor in communities. Unfamiliar with emotional labor? I recommend reading Hochschild’s work (here is a quick summary). This particularly may happen if a group felt a vacuum of a certain type of labor beforehand, or has unreasonable expectations for how women should participate. This goes beyond taking notes, it’s also about facilitation and making social dynamics work.

Uh-oh, now what?

This all sounds a little hopeless. We had one good sign: all these new women in tech groups! Lots of women that can be hired, speak, or be included in tech! But it might not matter.

It does matter, but we need to let Women in Tech communities be the starting point, not the ending point.

1. We need to accept that “Women in Tech” communities are just a starting point. Large companies shouldn’t only be looking at how to fund Women in Tech initiatives — they should be looking at inclusion initiatives within other organizations.
2. Women in Tech orgs (as they are able) should try to have broader impact. If all the knowledge is only shared internally, that doesn’t help elevate the members. I love the way theli.st handles this — in addition to being a network, the community is also a “visibility platform” and helps the women to have more impact beyond the group.
3. We’ll save ourselves a lot of work if we do a better job of thinking about diversity from the beginning. It needs to be something we think about in our first ten community members (or first ten hires). Given how quickly things in tech change, this might be higher impact than fixing the current behemoths. If you’re starting a group, please try to think about this from the beginning.
4. We need to recognize that there are experts on this topic, and studies that show us good paths. This isn’t a topic where we ¯\_(ツ)_/¯ and wring our hands and say we don’t know what will work. We do know some things, and we need to use those. If you don’t know, do the research or hire an expert.
5. We need to start keeping track of what we’ve tried: what works and doesn’t, and share that. We need to learn more. I’d love to see more data and stories on the experience of creating inclusive communities. Took a community that didn’t have women early on and added them (successfully) later? Share what helped you make that leap. Had a community that started gender balanced and then regressed? Do a post-mortem on why and how you could have prevented it.