Ask HN: How are online communities established?

I have built a couple communities over the years and currently run a community with more than 20 million monthly users.

I'm going to assume the kind of community you're referring to is one where the primary focus is social networking / discussion in the vein of Hacker News, Reddit, etc.

The way you build a community is to start with an existing community.

What I mean by that is you have to find an existing pool of users who are interested in what you offer and bring many of them in at once. From there you can focus on slower organic growth. Examples: YouTube and MySpace both began as dating sites. YouTube focused on getting people to upload introduction videos of themselves. Once they had accumulated a number of people who were willing to film themselves they pivoted over to content creation. Similarly, MySpace was a very crude dating site that allowed people to customize their page. Brad Greenspan was a serial investor who bought up a large number of tiny dating sites. He cannibalized all the revenue from those sites to promote MySpace as a "free" dating site. They had millions of users coming from other sites.

reddit was promoted heavily on Hacker News and focused on a tech crowd at first. Paul Graham threw his endorsement behind the platform a few times and that also helped interest people in checking it out.

For my successful community sites, I'll just mention one experience - I had built a Q&A site from scratch. It was finally done one Sunday night and I decided to go to bed. There were eleven posts on the site, mostly from myself, but also a couple from friends I'd asked to test the site for me. It was obvious there was no community there. I bought a single ad - for $10 - at $0.01 CPC on a Quiz site, then went to bed. My intention was not to launch the site, but to throw a little real traffic at it and see if any bugs cropped up that neither I nor my friends had found. When I got to work in the morning and had finished catching up on my emails I decided to check the site and see if anyone had posted. There were over 100 posts and people were using the site exactly as intended. Not only that but there was about $0.50 in ad revenue already, meaning my monthly run rate was net-positive from day one. I had hit upon a great fit between people who were already interested in asking and answering questions (the quiz site I advertised on) and my product.

In each case I'm familiar with, the formula was to find an existing audience that had an interest in the product in question and bring them on-board ASAP. If the internet were brand-new and no community websites yet existed I would build one by building a non-community website that provided a useful service, and building a community around that once I have captive eyeballs. In other words, to belabor the point: I would start with an existing community.

Shared common interest. If you're building it for one community segment, odds are that you have connections that would find use in engaging with that community. So I'd start sharing with the ones who don't mind starting the community from nothing.

Like if it's a community about a type of car, you'll want some initial owners or prospective owners (if it hasn't released yet). Maybe some DIY guides for basic maintenance, news section for OEM and aftermarket releases, etc.

This way newer visitors find already-established and useful content and stick around to discuss and add to it.

I’m trying to do this right now at conferacity.com, and it’s really hard. But I hope people will continue to try to experiment with new discussion communities online, not just based on the topic, but also on the governing rules and structure. It seems unlikely that 50 years from now, we’ll look back and say that the existing formats for online discussion were the best we could do, even though the existing formats are all clearly providing value to users in their current form.

An area I’m particularly interested in is, given that anonymous forums require moderation, but “moderation doesn’t scale”, what would happen if you only allowed moderators to contribute to the discussion. Or rather, what if you only allowed people who have the required attitudes, ability and emotional intelligence to be an effective moderator to contribute to the discussion?

You’d obviously grow much slower if you’re restricting contributor growth to a small subset of users. And maybe that constraint on growth would mean you never become a viable online community. But if you could get to some threshold size, you might find that the quality of discussion is sufficiently different that a lot of people might want to read such discussions without necessarily taking part directly.

Ideally, the contributor community would make decisions about what those required qualities are, and how to assess new applicants. But we already have communities like this offline - this is exactly how academic communities behave. Existing “contributors” (professors with phd’s in the subject) decide who they accept to their community, based on evolving standards decided by that community. It’s also interesting that this, at least from my outsider perspective, is also how YC operates - they have a particular ethos, norms of behavior etc, they admit people who they think are consistent with those norms, and then enforce those norms: https://www.ycombinator.com/ethics/

I think there’s an interesting analogy with professional sports. Baseball is a super accessible game, all you need is a stick, a ball and a field, so almost anyone in the world can play. But the way to get high quality baseball that people want to pay to watch is not by allowing anyone onto the field, but by restricting the players on the field to those with the required skills.

Regardless of how my particular “experiment” works out, I think it would be great to see more innovation in this area.

Reddit was seeded by sockpuppets.

It was a Dave Matthews Band fan community and I shut it down in 2015 due to server costs, (which the community mostly covered Via PayPal but I had to do fundraisers) looming security issues (drupal and vbulletin integrations that became hard to get security updates for and my own passing interest in the band.

But not because there wasn’t community.

I had a solid moderation and community admin staff and people writing content for the blog.

To your point yes. Any community, band fans to burning man needs new blood constantly or it will die. People move on.

There’s a lot of talk about this in the DOTA 2 community—-fear it will not stay at the top of esports if it doesn’t do more to make itself welcoming to noobs.