Future of Web Apps - Metafilter.comMatt Haughey talks about online communitiesMatt Haughey, the creator and of MetaFilter, the community news site (think of it a bit like a massive blog slash slashdot slash early prototype for Digg) who's talking about creating and running communities.
Blimey, this lad's tall.
Any site or application you create should include a social component. This is essential now for Web 2.0, but the potential for disaster is huge. But a successful community great for everyone involved, the creators and the members.
The lifespan of a community follows a classic growth curve sees an early spike where everyone tries it out for 10 seconds and never comes back. Maybe you'll keep 10% of people, and then over time it starts growing... and then there are three outcomes.
1. It sky rockets
2. It flattens out
3. It declines (it's almost always the person who's running it lets it die)
Be a third place. Following the idea that you have to be something outside of people's work and outside of their home: be something that people want to join in every single day and enjoy. That's the high-level goal. But it's not as simple as throwing some software together and a domain. A lot of thought has to go into it.
Going back to first principles, it's about having a compelling idea. The more clever and original it is, the better it will go. Ideas are cheap. If people tell you it's impossible, that probably means it's a good idea.
It's now 2007 and we've got 15 years of prior art on online communities, and five years of social applications. Pluck the best bits you've seen elsewhere. Eat your own dogfood, fulfil your own need, scratch your own itch [insert more cliches here]. It's obvious when someone's trying to make a buck or trying to be in the right place at the right time. Build for yourself first.
How do you encourage users and outsiders? Highlight the best bits. Reward your best users - it makes them feel good and recognition makes your site more interesting to casual readers, too. Lurkers get the benefit too. Your best contributors can be moderators.
Sometimes communities run into trouble when they get too controlling. Get out of the way when necessary, let things run their course within reason. Building in flexibility is a good thing. Don't always go with the constraining option - look at MySpace, where people stick HTML everywhere (even if it looks horrible).
Build out on those edge cases. When you see something good and interesting, make it part of your app. Smart tech companies watch what hackers are doing and tweak it. Car companies watch what customisers are doing and four years later it's a feature.
How to keep it going: stick with guidelines over rules. Rules lack the fliexibility, they put you in a position where you know something's wrong technically, but you know they did it for a good reason. Rules don't allow people to do innovative things. It's important to keep your emotions out of all decisions; if you don't have thick skin you're screwed. You have to learn to roll with it.
Tailor to your community norms - don't surprise everyone with a new rule out of left field. It will cause outrage. Don't anger your commnuity over something you just felt like doing.
It's a balancing act. I'd describe running metafilter as running a razor's edge between chaos and happiness. Here are some things that can push you over the edge. Ownership is a huge issue: the more people get invested in your community, there's a grey area between whether it's yours or theirs. Are you going to make your decisions top down or bottom up? keep these things in mind: it's not always black and white.
Happiness on the part of users is totally fleeting. Having a day of downtime loses all that good will. A lot of things get a reputation for being flaky just because they have a few issues.
Every community suffers a revolt eventually. People may say they're going to another site, no community's immune. Digg, Facebook, Flickr, they've all had them. Take them as an opportunity to learn.
Customer service. I never thought I'd be talking about this, but you start a site and you have service and support issues - even if it's free, even if people can do everything themselves. it's not just about coding, it's the sad truth that you'll spend more time on customer service than coding. It took me years to notice this: I thought I'd have eight hours a day to code, but instead I had about 20 minutes.
Hire or take volunteers as early as you can.
Metrics can really ease the workload. Users can report abuse, they become your eyes - if 27 people don't like something, there may be a problem with it. If you don't have something like that, if you have a support forum where one area has 50 posts a day then concentrate on that.
How do you avoid the eventual disasters? Be transparent. Do support public forums where everyone can see every decision. Explain everything; people feel like you're someone to be trusted. Have a place to talk about the site or app. If you don't have an outlet for discussion, you'll find people complaining about the site all over the place and then complain that you didn't see it and think you're unresponsive. Have a dedicated place
This also gives you a good venue for collaborating on new features. Explain why you're doing something. Over-explaining is always better than under-explaining. Acknowledge your mistakes - don't play the blame game, you'll just lose goodwill. If you see your community turning, engage them directly - email them, IM them, you can take the wind out of their sails. I fessed up to huge wholesale giant mistakes.
One thing that comes up more often than it should is that you will have legal issues. It's important to realise that you live somewhere and you made this app that's used by the world. What laws apply to you and your users can be different: find a well-versed lawyer who understands the rights of website operators and talk to them about what specific laws are most going to apply to you. Understand your liability: you don't want to lose your house over some silly web app you built. Remember - lawsuit threats are many, but actual lawsuits are few.
So what's stopping you? Find your balance, run your community on an even keel. Because in the end a successful community can please readers and creators; everyone's a winner.
Ian Forrester from BBC Backstage asks how dying communities can revive themselves. (Wonder why?) Haughey: Sometimes it's because people don't have much time, as the creators. Put time in.
How do you stop the core audience from calcifying? It tends to happen; the old guard will fight the new. I haven't found the magic bullet, and I can't think of any quick tricks. Digg shows you who's recently joined.
How do you find that seed group of users? Really specific communities are harder to get going, but then after the first six months or a year they really get into it. I would find similar user groups that cover that subject, maybe announce it in a mailing list - without stepping on toes - or make it your email sig when you're talking those people. Subtle promotion.
How do you make a small community look attactive to outsiders? I'd start out with good seed content. I'd shown the site to 15 friends who gave me honest opinions. Half of them started posting - it wasn't phoney but I went through a silent slog on my own and eventually people showed up. When you launch with anything, try to highlight the best bits on the front page.
Is there a point when you say now is enough because it takes eight hours a day? I said in the support forums that the more time I spent answering questions there, the less time I'd have for coding.
Are you earning more money now than before? errrrrrr.
Yeah.