I had heard several people say Design for Community by Derek Powazek is a great book and was published well “before it’s time.” As someone who works on several sites that I very much think of as community sites, I picked it up and gave it a read. Published in 2001, the book is just over 13 years old now. Ancient history for a typical tech book. It is a tech book in that it talks about specific websites and specific technologies, and those things do feel ancient (the screenshots are hilarious). But as a design book (and kind of an anthropology book), it has aged quite gracefully.
Since I read the book with a highlighter marker in hand, I thought I’d share quotes from it and think about how the ideas from the book have come to pass all these years later.
The word “community” is dangerous. It’s the kind of word that shows up in taglines and quarterly reports to make stockholders feel warm and fuzzy.
I’ve loosely used “CSS-Tricks: A Web Design Community” in the past (like on T-Shirts) and it does feel a bit empty. I don’t do myself any favors by reminding people that it’s “just me” around here and downplaying the site as essentially just a personal blog. It’s grown up quite a bit over the years and I’m going to try to keep growing it. I think I’m going to stop specifically referring to it as a community though and let it just be that if it truly is one.
If you’re successful, your members will start calling it a community on their own.
The book starts off with some bad reasons to add community features to your site, including a funny conversation with a CEO that wants to add a live chat room to their site.
“Well, if someone comes in and starts talking trash about the company, you’ll wanted them kicked out, right?”
“Right!” he said.
“Good,” I said. “Then it’s settled. We’ll have moderators in there 24/7. Who wants the night shift?”
And that ended that.
Completely open/anonymous/live chat is pretty much gone on the web. It will probably stay that way. The second you put up something like that it’s absolutely full of bad behavior. Even here on CSS-Tricks, a live chat demo I did once was full of awful things within hours. Imagine AT&T putting a live chat room on their homepage.
Chat is still around, in the form of private, invite-only chat. For instance, realtime customer support like Media Temple does or team/group chat like Google Hangouts or Slack. In the case of IRC, some channels have moderators. In the case of chat apps like Tlk.io, there are no big public rooms (there is no directory to find them) so rooms are kind of share-only and self-selecting.
When users post to your site, it isn’t a gift from them to you; it’s the beginning of a very real relationship between you and the user.
But they were the victims of their own success. Users began posting comments and questions so frequently that the comments only lasted on the homepage of the community area for a brief time. The more popular they became, the faster the posts scrolled away before getting responses.
An on-going problem today. So much so that Facebook continually experiments with what to show people when you visit. It’s not as simple as the most recent, it’s a (presumably) complex algorithm. Some forums feature things like “hot threads” or send digest emails of bustling activity. But most of them, even the forums here, just list threads by most recent. Definitely not a “solved” problem.
I’m always a fan of manual curation. If you have the resources, you pick the best stuff to highlight.
In a story about a digital photography site successfully adding community forums:
I’m certain that had Askey not primed the pump with high quality content from the beginning, the quality (and quantity, for that matter) of the forums would not have been nearly as high.
Certainly the case here on CSS-Tricks. You can’t just toss up a brand new website, call it “Wood Carving Forums”, and expect wood carvers to come flocking in. Even if you do “prime the pump” it’s no guarantee. The only guarantee is that things are going to start slow.
I often had (pipe) dreams of researching video games that were only going to come out years from now and setting up nice forums for the site so when they did launch things would pick up and maybe become the active/default place to talk about that game. But my lack of passion for said game would surely have shone through. And with no other content to support the forums, it was almost surely a fail.
Asynchronous communication gives the participants more time to craft elegant responses. … You may find that you want to employ a variety of synchronous and asynchronous communication methods for different tasks.
Posting = Asynchronous = Slower = More thoughtful
Chat = Synchronous = Faster = More off-the-cuff
They both have strong points and don’t need to be mutually exclusive. For instance, CodePen supports comment threads on all Pens and Posts. That’s asynchronous and you’ll sometimes see essentially mini posts there with code samples and lengthy explanations (Although I admit, they are, for some reason, prone to very short comments.) But in Professor Mode and Collab Mode, we offer the participants of that live chat. Messages come in real time with no page refresh and are visually called out.
Facebook obviously sees the value in both, offering both posts/photos which are slower and more considered and Messenger which is live chat.
Sometimes the line is blurred, like in GitHub issues where issues are essentially comment threads, but new ones come in in real time.
… the tone of the content you give your users is replicated and amplified a thousand times in the responses it generates.
No question. Publish an enthusiastic post, get enthusiastic responses. But more than the tone of an individual post, the tone comes from the copy everywhere on the site. From microcopy to documentation. From headlines to emails.
Mailchimp knows this well.
Personally I see it happen on the most molecular level. If I post something with a pissy tone in a comment thread, I get pissy responses. More and more I see that positivity spreads a little, negativity/anger spreads a lot, and positivity can often stop a negative domino effect dead in its tracks.
If the community is constantly reminded that the leaders are all real people, everyone will stay a whole lot friendlier.
I’m a big believer in this. I like to write like I’m just some dude talking to you. That’s how I think; that’s how I feel. That creates tone. That tone is casual. That is amplified in the communities.
It’s imperative that you remember that everything you type serves as an example of what is permissible in the forum.
No one is above the law! Especially the king!
And everything a user posts serves as an example as well. At a minimum you have to moderate threads, and ideally give moderation tools to users as well. Comment threads can get incredibly bad, like as bad as bad can possibly be. I also think of the insane threads we see these days on Reddit / Hacker News / YouTube. Awash with personal attacks, racism, etc. The longer it’s like that, the more “OK” it seems to others.
Find the proactive, positive members and get them involved in the production of your content.
I’m constantly trying to do this and wishing I was better at doing this. Easily could be a full time job.
Speaking of a full-time job, from an interview in the book with Matt Haughey:
Don’t underestimate the commitment required. Done right, a community site will take a lot of your time, and the payoffs, in whatever form you set for yourself as goals, may not come for a very long time.
Years, at a minimum, in my experience.
… pick something you’re passionate about, devote the necessary time to building a site around it, stick with it for as long as it takes.
… songwriting and web design are really the same thing. Each of them involves setting a beat and then improvising around it.
Nice sentiment. I’m admittedly fond of music analogies though. I like that idea that no matter how much of a plan you have, the real designing starts after you’ve started building and even after you’ve launched.
… tie your community features to your content as much as possible, visually and architecturally.
Having comment threads specifically for bits of content that are someplace entirely different on the site is just weird, and not something we’re used to seeing these days. Perhaps those days are rightfully gone.
Whenever I’m reading a site with a beautiful design and then I link away to a bargain basement area for the discussion tools, I feel like I’ve just been sent to the kid’s table.
I know I like discussion about content as close to the content as I possibly can. Even things like Disqus (while I generally like Disqus) sometimes feel a little more disconnected with the content than I would like (with the often different look and feel than the content above it). I wonder if the Medium technique of commenting on a particular paragraph is something that will catch on. Not sure how I feel about that yet. Clever, but distracting? Too close?
… the first thing the user is greeted with is other people’s answers to the question.
This has largely become a standard. Post > Comments > Comment Form. You have to at least scroll past some comments (which hopefully set the proper tone) before you get a chance to comment yourself. Maybe that’s what is weird about Disqus. Although they really minimize the comment form until you click in there which is nice.
Later, Derek talks about “burying the post button” in which he admits is counter-intuitive, but leads to higher quality content. I’m still afraid to do that one. I feel like I’d just constantly get called out for bad UX.
… don’t be annoyed when your users do something totally unpredictable. They’re doing you a favor – they’re teaching you about your site.
So valuable. As long as it’s positive, you accommodate, not force into another direction. That forcing isn’t going to go well.
… if your users sense that they’re being manipulated, they’ll resent you for it.
Rules are all well and good, but if they’re not communicated to the user in a clear way, they might as well not exist.
Not even that the rules exist at all and are posted somewhere, but that they are right in front of users right when they matter. Derek’s message on his site is a classic, one that I’ve riffed on in various ways over the years, and that I still see riffed on by others:
This is my personal site. It’s like my living room. I hope you’ll come in, have a seat, and be cool. If you’re not and you post something off-topic, mean, or just plain stupid, I will delete it and kick you out. I reserve the right to delete any post for any reason. I also reserve the right to believe that people are essentially good, in spite of overwhelming evidence to the contrary.
It’s like a compliment sandwich in paragraph form: start positive, mention the negative clearly, end positive.
He also points out some fun microcopy from another site:
If you’re being a jerk, we have the right to de-activate your account and put gum in your hair.
We try and do the plain-talking thing on CodePen in the midst of legalese. And it wasn’t our idea. It’s a trend I’d like to see more of.
Unfortunately, in my experience, the people who are very good at designing websites are usually not the same kind of people who are good at hosting and moderating them.
I hope that’s not true in the very personal case of me, but I take the point. I often see small companies hiring community people among their first hires and large companies having entire dedicated teams. “Social Media Person” surely wasn’t even a job when Derek wrote this book.
… whenever I posted something official, it was prefaced by “MODERATOR” and the color was changed slightly so that the community would know it’s official.
These days, WordPress applies a class name to the comments based on if it’s by a site administrator or the author of the post or whatever, which serves as a styling hook for just this reason. I’m sure most commenting systems have a take on this. The idea has caught on for sure.
While it may be important to give users whatever they want when you’re trying to sell them a widget, in a community setting, barriers to entry are a necessary part of creating a successful community space.
I think of big sites switching over to Facebook for comment threads. It’s harder to be anonymous there (although not impossible) but also I suspect a lot of people just live logged into Facebook so it’s extra obnoxious to switch over to your “anonymous” account. It’s not a perfect solution though, as I’ve heard YouTube had worse comments after forcing you to comment via Google Plus.
I think it’s fairly obvious that the bigger the barrier to entry, the higher quality content you’ll get. For instance, to this day you have to be invited to Dribbble. Certainly over the years not having an open signup has meant there is higher quality work being posted there. But also less of it.
Imagine a site you have to submit an application and be approved before you can comment there. That site would be lucky to get a handful of good applicants, I’d bet (in general). The idea of limiting what users can post, I suspect, is frightening to companies. Less activity is scary (nobody cares about us! It looks dead around here!) even though more activity is probably what you should be scared of (moderating and community-building is hard work).
If you’re trying to foster prolonged conversations over time, requiring accounts is the way to go.
If, on the other hand, you’re primarily trying to foster a one-time connection, then requiring accounts is an unnecessary hassle to impose on the user, and you’ll wind up turning too many people off to justify it.
Just on this site alone there are at least four types of barriers to entry:
- Content – people will never come here if this kind of content isn’t of interest and they don’t seek it out. (good, this is a site by web nerds for web nerds)
- Blog comments – Registration isn’t required, but leaving a name and email address is. (quick, anybody can do it)
- Forums – Registration is required. (makes it easy to come back and be a part of ongoing discussions)
- The Lodge – comment threads on those videos require a paid account. (less conversation, easier to manage, good for business)
The barriers to entry can be decided upon on a per-activity basis. Writing or reading.
Email connects people in an immediate, personal way. So it’s not surprising that some of the most powerful community experiences can be found in your inbox.
In this chapter Derek talked a lot about email-powered communities. Mailing lists / newsgroups kinda things. Nothing public on the web, which gives that style of communication an “ephemeralness”. I don’t see a lot of this anymore. Perhaps a slight resurgence in email newsletters, but that’s more one-way.
I also think of things like the W3C mailing lists, which take place over email, but are archived on the web. To me it feels archaic and exclusionary (people still do this?!).
Email is alive and well though, and probably just about the same as it was 13 years ago, if not relied upon even more heavily. Their connection to communities being more relegated to notifications. I don’t go on Vimeo every day, for example, but thanks to email notifications of new comments, I can still feel connected there.
I’m still reading it. Savoring it, really. The last chapter is titled “Killing Your Community” and you just KNOW that has some juicy relevance to today’s world.
Derek’s current project is Exposure, which I’ve used and quite like. It’s community-like, in that you can browse around other non-anonymous people’s photo-stories, but interaction with them is very limited. I wonder what they have in store for community on the site, or if less-is-more here. What lessons from the past are driving decisions here? Interesting to consider.
Ethan Marcotte just wrote about an old article with similar value. Are there any “old” tech books or articles that you still savor?