Addressing gender balance in online communities

This seems like a suggestion worth considering but would need to be baked into core. Of course, you need not ask everyone there gender to get these statistics. In plugin territory, you could set up a system to serve questionnaires to a random sample of your users. This could pick up on gender as well as a number of other useful stats.

1 Like

In our goat keepers community, it happens so that MOST active users are women. The impression is that 99% of them are. That’s what I think impacts perception of the balance: not registration statistics, but activity statistics.

It has even become natural in our community for people to say “hey girls need your help”, because most of the time it will be women who will respond and give advice.

That might be specific to the industry our forum serves. Nonetheless, our community is one great example of a totally gender-imbalanced community.

Does anyone struggle because of this? None that I know of. Girls are happy that they prevail in the community. Men are also happy because there is so much attention to them from girls. Haha.

Jokes aside, I think it might have been different if we had had male-prevailing community, but I don’t know for sure and just guessing. Not trying to say anything particular, just sharing our experience in case it’s useful to anyone.


I think it’s important to clearly delineate a few things here, to focus on the practical impact of this question and avoid politicising it to much (not that anyone here has, but I’m guessing some people would be reticent to wade into these waters due to the political / social climate(s) around issues such as this).

  1. Gender identity, not identity in general. Online gender identity, and diversity more broadly, is a subset of the larger issue of online identity. Within the larger issue of online identity, the issue of anonymity can be gender neutral. You may want to not use a real name or profile picture for reasons unrelated to your identity as a person, gender or otherwise. As has been pointed out, I don’t think we want tackle the whole real name / anonymity debate in this context. We should focus on the issue of gender and diversity.

  2. Substance over form. As @meglio alluded to there are differences between both: 1) registration and activity (i.e. being a real part of a community); and 2) perceptions of diversity and statistics about diversity.

  3. #notallcommunities. And, as @meglio also pointed out, this issue affects different communities differently. Many tech focused communities feel quite ‘male’ dominant. His community feels (is) female dominant.

Given that, I think we’re dealing with two issues here:

  1. Knowledge. Site admins and mods may not know if diversity is an issue (assuming they care about it). On this front I also think @barryvan’s suggestion is a good one. v1 of this is actually possible right now. You could add a custom user field in the admin panel, require it at sign up, not show it on the user profile and add a description telling the user that this information is kept private. As to the broader questionnaire @jcoates referred to, you could do that with the Custom Wizard plugin (albeit not the part about sending it to a randomised subset).

    The knowledge issue is itself a subset of a bigger issue: What does my community look like? Figuring out ways to have statistics about the actual people involved in your online community baked into Discourse would be in-line with Discourse’s overall approach of helping people build civilised online communities.

  2. Culture. As has been pointed out, it often takes initiative by leaders of a community create an inclusive culture. There may be a role for the software itself in this too though. In fact Discourse is explicitly designed to play a role in helping to construct a ‘civilised’ community. On this front, perhaps we could review these aspects of that structure with an eye to gender inclusiveness, in particular the wording and content of the default community guidelines, education messages and other text. I’m not saying they are necessarily lacking in that respect, I just haven’t read them through that lens before.

Discourse isn’t going to ‘solve’ gender and social issues in society(s) at large. But within Discourse’s remit of facilitating civilised discussion online, it can give specific tools and advice to people who want to:

  1. Know more about their communities, particularly how diverse they are.

  2. Make their communities more inclusive for non-dominant groups.

At the end of the day we may find that Discourse itself is already set up to do this (e.g. the custom user field at login). And if it is, maybe Discourse should mention that somewhere in a tasteful way on its homepage :slight_smile:


Just brainstorming: Maybe when a user reaches TL 2 we could ask them to optionally fill in more details, like full name, gender and birthdate. This could be done with discobot, a wizard module or just a plaintext request by PM.

Of course it would be more interesting to have this data from the beginning and for everyone, but it’s an additional barrier to entry upon signup and I strongly dislike that. Why should I trust my personal data to a system that I’ve had no interaction with yet.

Trust Level should be a two-way thing. At TL2 the system trusts you more, and you may feel comfortable trusting the system (and by extension, the admins) back with a little bit more information about you.

I also like this because it runs counter to the industry standard of “collect as much data as you possibly can up front”, which users have somewhat unknowingly allowed. Discourse is different: it actually takes its sweet time getting to know you better.


There are two things here and both have my mind spinning

  1. Collecting demographic information on members
  2. Finding ways for communities to be aware of potential bias in content / membership that is either endemic in the population or occurring online … and giving managers the tools to counter it

On the first point I have had a dream for a while which probably deserves a separate thread, but where admins could ask a series of questions (either all together or as a series) to collect preferences and provide data that can be made public or used to create groups. I will expand on this elsewhere, but it seems that completing occasional 1 or 2 question surveys could help to build up useful profiles, … and gender (in all its beautiful, myriad expressions) could be one of them.

On the second, we can only be aware of bias (in our content, culture, participation) if we can measure it. In my Discourse community, we deal with this by tracking gender using an excel calculator sheet, but I am also exploring the option to tag members in Community Analytics. We split our analysis of engagement by gender and can see issues with participation levels as well as engagement.

This in turn leads us to try and create attractive content for this audience, but I’ve struggled to keep it on-track because the nature of the bias is not that male members are not interested, but that they see the topic in a certain way (which I believe is different from the female audience). I’ve tried limiting their participation by excluding the most active male participants from early-stage discussions, but that is not easy to achieve with the discourse security functions (see my conversation on exclusions rather than inclusion).

We have also created a ‘mentor’ concept where we on-board female members in a slightly different way and encourage existing female participants to interact with them and mentor and support them, but this is in its early stages.

Interestingly I am thinking of moving away from real names because I am finding that female members (and some ethnic minority members) are less willing to be identified as such, and therefore by trying to protect them we are actually making them a more obvious target. This doesn’t help or make sense. Thankfully, we have not had any issues yet (and I don’t really expect them), and I want to keep it that way!


Yup, that’s my experience. I don’t want to be forced to use my full name and expose my gender and I know many other women feel the same way. (For clarity, I don’t feel strongly about it these days but I did when I was younger and believe it is my right to choose.)


To add another dimension to this already complex topic: to what extent or in what way does the users’ gender identities actually matter? I would say, it matters only as far as they are expressed in the forum discussions. In other words: the actual distribution of gender identities among forum members doesn’t actually matter as long as what is being said on the forum (or how it is said) doesn’t negatively affect a specific gender group.

I suppose that is what @angus meant by culture, except that I perhaps wouldn’t make the distinction culture (how we talk) vs knowledge (what our identities actually are) but rather what we know about how we speak vs what we know about member identities. And then, I’d say, the latter becomes much less important (as much as we like measuring things in order to grasp them) because what matters is how we talk, say, about intersex people, regardless of whether the forum even has a single member identifying as intersex. So, I guess what I’m saying is that this discussion should focus on how to promote gender diversity in forum content rather than gender diversity among forum members (though the latter can be a tool to promote the former). As has been mentioned before: gender diversity is one element of civil discourse.


While I agree with what you say here, the initial discussion was about finding ways to demonstrate to people that visit a community for the first time that there are ‘others like them here’ and I’m not sure content can always do that quickly enough. That said, that wouldn’t always be addressed by gender.


I agree in general, this is not about box-ticking. However, if you see that your member profile is split by gender differently to the population, and/or that within the forum, engagement levels differ substantially by gender, then you have a “gender balance issue” that has nothing to do with how we speak about gender itself.

My topic is wine.

Men like to compare, rate, achieve, and “win” by discussing wine in a certain way.
I believe that Women (and some men too) find this off-putting because they want to share, learn, or “be social” with wine in a non-competitive way.

It is nothing about how people of any gender are treated, it is just how they see the world by default, and maintaining a healthy balance of topics and viewpoints can be tricky :slight_smile:


Good point! I tend to see forum participation as being driven by common interests and in that view, it doesn’t matter whether or how many “like me” are part of the community. Participating is rather a way of finding out. But I have learned that this view is somewhat lacking. So it’s good that you remind me that relationships count.

I fully agree. What I meant by “gender diversity in forum content” and “how we talk” was not just “how we talk about gender” but how we talk about anything, because certain social groups talk differently about things and if one way of talking dominates certain groups are excluded (which, of course, further amplifies the dominant discourse).

Bu I suppose you are right: I probably overstated my point. I don’t know whether that is characteristic of male talk or just a way of talking on forums (too much nuance can kill a conversation), I don’t know…

1 Like

Just so you know, there are communities where it works vice-versa, like ours: Everyone wants to mention others by name, and everyone asks everyone else to fill their name in their profiles. People treat mentioning by name as kind of respect. I even have to teach people to ADDITIONALLY mention by @username to keep notifications working.


Achieving nominal gender equality through active discrimination? I understand your intention but I think the method is totally misguided and will lead to worse outcomes.

Social engineering is ugly.

Here’s an analogy. I’m a man who joins a knitting group (let’s say, by way of crude example). It’s female dominated. That doesn’t particularly bother me.

And then one day I find out there’s a social engineer leading the group who has muzzled female members - prevented their participation - in an attempt to make me feel more involved.

Firstly, how sinister.

Secondly, how patronising!

I would want nothing to do with such a group. I came there to knit, and to speak to people about knitting. Not to be part of someone’s macabre real-life sociology project.


It’s more based on activity, really, from my perspective, not gender per se … anyone that is dominating the discussion. See this post from @mcwumbly which has some good guidelines:

So leading a good fight is a skill that can take a long time to develop. To try to accelerate that process, my Stanford colleague (and former HP executive) Debra Dunn and I have been pulling together a list of methods. Here are five of the tricks used by some of the best bosses we’ve observed:

  1. Don’t let the arguing begin during the initial generation of ideas or solutions. Make it safe for people to suggest crazy or controversial ideas. After you have some ideas, then invite people to push back on them.
  2. Bring everyone into the fray. Gently rein in people who talk too much and encourage those who are silent to speak up.
  3. Don’t just listen to people’s words, watch non-verbal behavior. Are they smiling? Really listening? Glaring, smirking, or rolling their eyes? Model constructive non-verbal behavior and coach people who (perhaps unwittingly) interject negative expressions.
  4. Learn people’s quirks. Some have remarkably thick skins; nastiness doesn’t faze them. Others are so thin-skinned that even gentle critiques send them into a rage or a funk.
  5. After the fight is over, do some backstage work. Soothe those who feel personally attacked and whose ideas were shot down. If anyone made personal attacks, call them on it and coach them to do otherwise.

Emma Irwin is an “Open Project & Communities Specialist” at Mozilla and wrote a series of articles about their findings after researching their Diversity & Inclusion efforts. They’re worth a read.

Part 1:

Part 2, Part 3, & Part 4

Some conclusions from the series:

Despite positive sentiment, and optimism we heard a great deal of frustration (and some delivered tears) when people were asked to discuss elements of participatory design that made contributing feel valuable to them. Opportunity, recognition and resources were perceived to be largely dependent on staff and core contributors. Additionally, recognition itself varies wildly across the project to the omission or inflation of achievement and impact on the project. We heard that those best at being seen, are also the loudest and most consistent at seeking recognition — further proof that meritocracy doesn’t exist.

Emerging from this research was a sense that standards for recognition across the project would be incredibly valuable in combating variability, creating visions for success and surfacing the achievements. Minimally standards help people understand where they are going, and the potential of their success; most optimistically standards make contributing a portal for learning and achievement to rival formal education and mentorship programs. Success of diverse groups is almost certainly dependent on getting recognition right.


While I agree those guidelines are useful, they are incomplete. The don’t take into account the different starting points of participants in the discussion.

The point about gender imbalance is that historically, and currently, women and other non-dominant groups have been marginalised (intentionally and un-intentionally) in certain communities. If we assume that all participants in the discussion are equal rational actors, only differentiated by their individual foibles, I think we miss something significant about communities.

I think it’s tempting to imagine that online spaces are neutral and objective in a way the offline world isn’t. We all have the same ‘access’ after all. But they’re still made up of people, and people are affected by culture, groupings and norms just as much online as off.

This doesn’t mean that we need to overwhelmingly focus on non-dominant groups and start posting ‘trigger warnings’ everywhere or something.

Rather, I would add a 6th item to that list:

  1. Be aware that your community/group/team is affected by social norms. Some new members may find those norms a barrier to fully engaging, even though they have valuable contributions to make. Make an extra effort to make them feel included.

There are plenty of very dominant women on my forum!

It’s a bit dangerous to prescribe traits to genders IMO.


I mean it in terms of social power, not in terms of personality. Historically and currently (in various respects), women are a non-dominant social group in the vast majority of societies, i.e. most societies are patriarchal.


I think that the concept of “real names” is problematic, in that this says that the name given to one by others (such as parents or the state) is somehow more real than when one names oneself. Especially when the context is people with marginal identities, and we are essentially saying that your identity is ascribed to you by others.

And when this concerns gender, it buys into the stereotypes of gendered expectations of names. Those expectations are often culturally-specific. There are trans* people who have difficulty with legal recognition. And what exactly are the queer, non-binary, and intersex names?

I can understand wanting to establish some reputational metric, but I think that using a name is not useful for this.


I think that the best way to go about it is to simply not jump to conclusions about people. If you don’t know what they look like, sound like, their ethnicity, their gender - then leave these concepts about them “blank”. My take is not be blind to these things, but rather let people define themselves. People try to guess these things because they are socialized to assume they are relevant, but do you have any trouble not guessing what somebody’s blood type is? Probably not. That is the kind of discipline I aim for.

A related problem which I haven’t seen mentioned here is the problem of users ascribing race, gender, and other identities onto other users in the context of (usually political or social) discussions. I have been on fora where people advocate for marginal people, and seem to relish “exposing” others in some troubling ways. For example “You post like a Hispanic!” would probably get somebody suspended or banned for bigotry. But regulars would routinely dogpile critics by insisting that their views betray them as “white cis males” and they receive no sanctions because of the societal power imbalance between the groups. Still, it functions like tone policing and another troubling ascribing of identity - especially when you are indeed a marginal person with dissenting views. I treat this as another instance of reasons to avoid guessing about people’s identities.

As for a user-disclosed gender feature, I recommend an actual string field instead of checking a box. Most online services I have seen offer only M and F, a few offer M, F, and “other”, or “rather not say”. Better to let people define it openly themselves.


I agree this is the only way to go.
Additionally a second string field for the users preferred pronouns might be a good idea.