Monday, June 30, 2008

The Majority Does Not Rule

News today that Netflix is reversing field and not, repeat not, eliminating the Profiles feature on its website:

Netflix issues mulligan on profiles, won't eliminate the feature

According to Netflix, they intended to delete the feature (which allows members to set up separate profiles for their family members or roommates, with their own movie queue) because only a small percentage of members used the feature. Only it turned out to be a really vocal percentage, and now the company has been forced to backtrack.

Aside from my own personal pleasure in the announcement -- I love the profiles feature, since it allows my wife and me to split our four DVDs down the middle -- I think there's a lesson to be learned here. Namely, not all community members are created equal. I'm sure it's true that only a small percentage of Netflix subscribers used the feature, but it looks like that was the most passionate, dedicated members of the Netflix community. They were the core of that community, and you piss off your core at your own peril.

When it comes to site functionality, the majority does not rule; if something is used regularly by your most engaged, passionate, dedicated members, don't even think of eliminating it, or you'll see those engaged members migrate to your competitor.

Thursday, June 26, 2008

Product-Centric or Community-Centric Design?

The link o' the day this time comes from Search Insider, via a very interesting article on companies that put the product first:

Thank God for Product-Centric Leaders

The full article is well worth a read, but in a nutshell, the author talks about companies -- Apple, Disney, Google -- that based their success on being obsessive about the product they developed. Steve Jobs didn't lead Apple in the development of the iPhone because market research indicated that there was an investment opportunity. He did it because his cell phone sucked and he wanted to design the phone that he wanted to use.

What makes this point interesting is it signals a way in which you can satisfy the customer without listening to him (or her). Basically, you just find your inner customer and design something that you want -- badly.
"The other thing about product-centric leaders is that they don’t have to do extensive customer research. They may, and many do, but they already have a gut instinct for what their customers want, because they are their own customer."
This insight relates very directly to something I've experienced again and again when developing websites: we always talk about building this or that for the customer, but we don't really know what the customer wants. We could bring in focus groups, but that's expensive and time-consuming and hard to justify to management, so it hardly ever happens (or if it does happen, it's too late in the cycle to make much of a difference). So we throw things at the wall and hope that something sticks, and as often as not we end up with a mess of half-baked ideas without any really powerful organizing theme.

What's forgotten in all of this is something that the article cited above really stresses: there's always a customer inside you. You know what you want in a website. So why not design it to scratch the itch that you personally feel? It's a simple question, but one that in my experience almost never gets asked. I have asked the question. I've been in design meetings and put the question to the people around me: "What do you look for in a website? What would make you visit ours?" And I tend to get blank looks. It's not a question people are used to asking, and not one they're ready to answer.

I think we should all take a play from the Steve Jobs playbook. Design sites that satisfy ourselves first and foremost. When you start planning, sit down and record the three to five things that it will take to turn you into a member of your own community. Build to that personal vision. If there's something on your web feature list that doesn't speak to you individually, ask yourself whether you might be better off without it.

I have a simple rule: sites that are designed in order to meet your company's needs tend to suck, and they fail. I used to say that sites that focus on the customer avoid that problem. But now I'm willing to allow that maybe sites that are designed to appeal solely to members of the core team have the potential to be best of all. They will be sites built around a focused vision, with no extra pieces. That's a pretty good recipe for success.

Wednesday, June 25, 2008

The Dangerous World of Web 2.0

Today's link is to an article on Alertbox, Jakob Nielsen's forum for making dogmatic pronouncements about good website content and design:

Web 2.0 Can Be Dangerous

The section on community comes in the middle of an article full of crotchety old man denunciations of what kids nowadays are into. Don't get me wrong, Nielsen makes some good points; if you're being led around by the nose by the latest technological fad, your site will suck, no doubt about it. Good design and content strategy are key, and ultimately it doesn't matter whether it's wrapped up in Ajax or plain ol' HTML. Still, there are a number of points where Nielsen gives in to Angry Old Man syndrome and throws the baby out with the bathwater. For instance, his reason why you should avoid user-generated content:
"On the Web, most people are bozos and not worth listening to."
Nielsen makes a common mistake here: he presumes that people need to be worth listening to. In fact, there are very few circumstances in which the quality of your community's discourse actually makes much of a difference. If you're trying to put together the next Wikipedia, for instance, you'll need to provide some sort of a quality-assurance mechanism. And if your user comments could somehow put your company at risk, through legal liability or other means, then you'd need to do some quality checks. But otherwise? Mostly irrelevant.

Consider forums. A successful forum is one in which people make lots of comments. Doesn't matter if their comments are the stupidest thing you've ever seen, an active forum is a good forum. Same goes for community blogs; with the single exception of comment spam, you don't really care if the blog posts and comments are full of crap, just so long as they're full.

Or consider user-generated content, like videos. Have you seen most videos on YouTube? They suck. No reason at all to watch them. But there's so damn many of them up there that quality is there to be found on nearly any topic, and so the community thrives. I can guarantee that there are just as many bozos posting to YouTube as Nielsen imagines, but that is quite irrelevant to the actual success of the site.

So bring on the bozos. As long as they're actively engaged with other bozos on your site, you have a community that is engaged and drawing repeat traffic.

Tuesday, June 24, 2008

Will Social Networks Ever Make Money?

This morning I came across a two-part article on Technology Review:

It's an interesting piece that focuses on the disparity between all the attention and venture capital focused on social networks, and the uncomfortable truth that (to date) no one has developed a business model to turn a profit on those networks. Definitely worth a read.

The crux of the problem is that advertising simply doesn't suit a social network audience. Google ads are profitable because people often search Google for a product they're interested in. They go there with the specific intention of learning more about Product X, so when ads pop up for Product X, they're more than happy to click on them. The situation is completely different for social networks, though. No one goes to Facebook to buy a television, and they're not on MySpace to do anything other than connect with friends and maybe cruise through a few random pages. Ads in that context are an intrusion, and marketers are only going to make it worse. Consider this ominous quote:

"The trouble," says Goldstein, "is we're putting ads up in front of users, where they can ignore them. We've got to get them between users."
Yes, good thinking -- put your ads between users, thereby effectively blocking the one activity that they come to social networks in order to pursue. I couldn't think of a better or faster way to shut a social network down than to force its members to click through an ad every time they want to make a social connection.

There may be no solution to this problem. Certain human activities are simply not connected closely enough with a purchasing decision for the activity to be monetized. This, of course, could pose a very serious problem for online communities that need to find some way to offset the cost of their bandwidth, servers, and company payroll. One possibility is to charge a subscription fee, but very few sites that throw up a pay wall will attract enough members to make their community experience worthwhile. Another possibility is to take a lesson from the music industry and try to think of ways that free content in one space can attract money somewhere else. Are you more likely to buy a used car from someone on a social network who's listed as a friend of a friend (of a friend)? If so, maybe a Craigslist tie-in is the way to go. Most fruitful are likely to be products or offerings that improve the social experience itself -- greeting cards, for instance, or gifts.

There are reasons for optimism. I don't read magazines or watch television because I'm looking for something to buy, but advertisers have long seen value in placing ads in those media. Perhaps social networks are simply too young, and the money will come later. Still, it's a tricky scenario that likely will not be solved anytime soon. When the next Internet bubble pops (in six months to a year, if the rumors I'm hearing turn out to be accurate), expect more than a few trendy social networks to be among the leading candidates to be this generation's

Tuesday, June 17, 2008

Three Trends in Social Networking

Another useful post from what's quickly becoming my favorite blog: This time they're reporting on a presentation by a Google rep on trends in social networking. The full post is here; I'll be going through the content point by point below.

Discovery is becoming social. Searching on Google is good, but having your friends help you find what you're looking for is better.
This dovetails with what I've seen on gaming sites: when someone has a question, they don't come to the website to find a document they can read, they go to the site to find someone they can talk to. That's why they usually end up in the forums, when there might be high-quality editorial content sitting unused elsewhere on the site.

It's worth remembering: many of your visitors come to your site because they have a problem, and their preference is to connect with another person to find a solution. So enabling that connection is the fastest way to reward the visit and gain a repeat visitor.

How we share is changing. People under-share because they don't want to appear self-important. But your friends really do want to know what you're up to. Facebook and FriendFeed let your friends discover what you're doing on their terms, and encourage more sharing, since you don't have to get in your friends' faces every time you update.

When you start to think of your web community as composed of friends, it changes the way you view web features. How do friends connect with each other? How do they want to interact? Do you provide them with the opportunity to connect in those ways, or will they have to go somewhere else to fill that need? I can think of dozens of "community" websites, but very few that allow me to connect with other members as if they were actually my friends.

So, as a mental exercise, imagine a scenario in which you and a friend of yours would need to keep in contact by way of the web. What would you need? Messaging, certainly, so you can keep in touch. And some sort of status report, so your friend could check up on you and see what you've been doing (those parts of yourself that you're willing to share, at least). There should be some means of scheduling get-togethers, whether online or off. And then a means of sharing items, be they photos or whatever. It's a rare site that enables more than one or two of those activities. How does your site compare?

Social sites? No, social Web. The idea of a site built around user content (like Epinions) is old-school. Today, users expect all sites to be social. They expect that if you're on a commerce site that you know your friends are also on, you can see what your friends bought there and if they liked it. Social is a feature, not a destination.

This is an idea that's good on paper, but very difficult to realize. There's a simple reason: compare the number of friends you have to the number of websites you visit. That's not a very good ratio, is it? Let's limit the discussion to just Amazon: when you're browsing books there, how likely is it that a "friend" -- no matter how tenuous the connection -- is likely to have browsed the same books and left recommendations behind? Even worse, think of all the books and movies you've bought and rented over the years; are there any titles in that list that you don't want your friends to know about (guilty pleasures, porn, or what have you)? Some people are compulsive recommenders (in The Tipping Point, they're called "mavens"); the rest of us will be too lazy and occasionally unwilling to share that part of ourselves.

So while, in theory, I like the idea of "social" being a feature, the reason that there are social destinations is because they're like the corner bar, which you visit because you know your friends are there. It's natural for Google to want community to be a feature rather than a destination -- they're currently promoting a technology for cross-site community features -- but the reality isn't there yet and possibly might never be there.

Saturday, June 14, 2008

What's the Online Community Tipping Point?

Recently I've been reading The Tipping Point, which is a book that poses a lot more questions than answers, but the questions are interesting. One of the topics the book touches on is the size of communities; it cites some research and historical examples that identify 150 as the optimal size of a community, be that community a constitutional democracy or a private company. Get much larger than 150, and your unified community starts to break down into sub-communities, and shared values begin to suffer.

That poses an interesting question: what is the optimal size of an online community? My first guess is that it would have to be significantly different than 150, since of course no one lives permanently in an online community (except, perhaps, for the more obsessive WOW players). The true number might be larger or smaller than 150, depending on how you interpret the impact of that asynchronicity. The repeat visitor figures on the sites I currently work with indicate that the average community member -- defined as someone who comes back to the site repeatedly -- probably doesn't visit more than a couple times a week. My own experience -- being a borderline obsessive-compulsive -- is that I would visit a community I actively identified with five, maybe six times a week. So let's assume a range, where there are a few core users who come by almost daily, and a much larger number of people who visit once or twice a week. What in that dynamic might allow us to identify an online community's tipping point?

I don't have answers. I'm not even sure there are answers to be had. But there do seem to be some interesting avenues of investigation:
  1. The Tipping Point tells us that the importance of community size is communication and shared experience. So, in online terms, you can ask how many pages of copy -- articles, forum posts, what have you -- a reader must take in to arrive at a shared experience with the other members of the site, and then you can estimate how long it would take to read that and compare that to the amount of time visitors are actually spending on the site. 
  2. Clearly a community can be both too big and too small. If we're talking about forums, for instance, too few users give the impression of a wasteland; there's not enough getting posted to reward the visitor for coming to the site. Too many community members, though, and you have a blizzard of posts that no one could possibly be expected to read in their spare time, and (as in the communities cited in the book) you'll see you community fracture into two or more. Now, you might not care -- if all you boss cares about is page impressions, a fractured community is as good or better than a unified one -- but it's worth wondering whether the cohesiveness of an online community can contribute to its long-term health.
  3. There is one feature of online communities (well, more than one, but bear with me) that sharply distinguishes it from offline communities, and that's the fact that upwards of 90% of its members are casual, part-time participants. Online communities are more like a big-city bus terminal than they are like a city, or even a company; they're full of many people hurrying by on their way to something else. What impact does this style of drive-by community participation have on the community experience? There certainly must be a degree of alienation that results, but how to quantify it?
More work is required. It's a potentially crucial question, though. Community is formed when people come together, and common sense tells us that too many people and too few are both bad. So somewhere in between is the sweet spot, the ideal number when things just click. That number exists. But how do we find it?

Monday, June 9, 2008

Measuring Community

Building a community site is just the beginning. Once it's up and running, how do you measure your success? You could take the easy way out and just go by a simple measure like number of registered members, but that's the wrong way to go. What you really need to know are things like: How engaged are my members? Are they getting more involved with the site over time, or less? Are they attracted to the web features that I spent the most time (money) on, or are they instead lining up to use the features that come standard in my CMS?

In short, you need a strategy for measuring your community, and that measurement needs to have historical dimensions so you can track trends. My manner of approaching this task has changed over time; this is where I currently stand.

1) Bounce rate. This is the measure of visitors who come to your site, take one look at it, go, "Bleah!" and leave. Ideally this should be the number of visitors that leave within 10 seconds of hitting your site. Depending on your analytics package, this figure may or may not be available. On my current post we're using Omniture, where the best I can do is measure the number of visits that take less than one minute. Given that the sites I work with are pretty content-intensive, a visit of less than a minute is a pretty definitive bounce.

A variant bounce-rate measure would be single-page visits, though you should look a little deeper here and see which page they're visiting. It may be that you have a single page on your site that meets the needs of a number of your returning visitors, in which case it's not a classic bounce. On the other hand, if this single page is your home page, it's a good bet that you're failing to engage with your content, design, or information architecture.

2) Forum posts. Not a sexy number, since this is your members doing the community thing on their own, without engaging with any of your fabulous (and possibly expensive) original content. But it's real community juice, and you want to know if posts and threads are going up or going down over time.

3) Time spent per visit / Pages per visit. These are two sides of the same coin (so it should be perplexing if one is trending up while the other is trending down). Of the two, I tend to prefer the latter, because while pages per visit is skewed by search engine spiders that crawl every page on your site, my personal bias is that time per visit is skewed more by visitors who load your site then walk away, leaving the browser open without exiting the site. I have no data to prove that, mind you, it's just my lurking suspicion.

Either way, you want these measures to rise over time. More time and more pages per visit mean that what you're building is keeping people around.

4) Visit number. I'm fond of this metric. It shows you how many of your visitors are new to the site vs. returning. Community sites should draw in more and more repeat visitors over time. Of course, this is easily skewed as well; anyone who clears their cookies will be registered as a first-time visitor the next time they visit. That's why you shouldn't look at absolute numbers, rather look on whether the numbers are trending up or down. Down is bad.

5) Success events. This metric isn't special to community sites; pretty much every serious site should come complete with an idea of what you want your visitors to do when they visit. When they do those things, it's a success event. Define them, measure them, graph them over time. Trending up is good.

And last, and somewhat least, you've got:

60 Competitive stats. Odds are that you have competitors: sites that make you grind your teeth, those evil bastards who are stealing your visitors and buying Google adwords that rightfully belong to you. and a few other sites will graph traffic to any URL you enter, though their data certainly isn't of uniform quality. So come up with your two or three top competitors and see how they measure up against your site! You'll never know how trustworthy the data is, but it certainly makes for interesting reading.

Site analytics is where, in my experience, most community sites fall flat. All the energy goes into features; after they're live, no one really wants to follow through and see what is performing well. Maybe we're all just afraid that we're secretly failing, and pulling down those numbers will expose us for the frauds that we are. But if you don't have those numbers you never have the chance to learn from your successes and failures. Gather the numbers, report them monthly, and get better at what you do.

Messaging is the Core Community App

I learn new things all the time on my job. That's one of the things I like about working on the web: there's never a final, determined way to do things. The best practices of today will seem like naive self-delusions tomorrow. If you don't keep learning and discovering, you will fall hopelessly behind.

One of the things I've learned -- or that I think I've learned -- in the past year is that messaging is the core community application. It took me a while to arrive at this point of view. For a long time I believed content was king: produce excellent material and people will come to read it, and out of that assembled audience your community will form. It was a nice idea, but traffic logs exposed the lie. While I was creating excellent content, my intended readers were whizzing past it on their way to something else.

Usually their destination was the discussion forums. That perplexed me. Why, I wondered, would they bypass well-written articles and choose instead to read poorly-written, poorly-spelled, ungrammatical, and often rude, juvenile, and factually incorrect posts in the forums? Couldn't they see that the better content was available elsewhere on the site? And, of course, the answer is they could see that. They just didn't really care.

Eventually I realized that my site's visitors weren't heading to the forums because they wanted to read. They headed there because they wanted to connect with other people. They wanted an active experience, not a passive one. And the way you actively connect with another person is by communicating with them. The sites I worked on didn't provide email or private messaging, so the forum experience was the only means to talk directly with their peers.

Of course, lots of other people have figured this out. MySpace and Facebook aren't built around their excellent content, they're built around the many ways in which they allow people to message one another (for what is SuperPoke but a somewhat annoying way of sending a message?). Twitter has become a hot startup because it managed to bring messaging to the little moments in life that aren't worth email, or even IM, but still are interesting to people who share a connection with you. I may have been late arriving to the party, but that's because -- as an editor and writer -- I had a lot of preconceptions to overcome.

I'm not dissing content here. People are still consuming plenty of passive content, and there are a number of blogs that I read even though I have little interest in posting a comment or emailing the author. Passive experiences still can reward the time you put into them. But, even so, blogs have become so powerful so quickly because the content is just the beginning. A blog post ideally is a jumping-off point to an ongoing community discussion. That's what's addictive: the sense that, even if you choose not to respond and engage actively in the discussion, you can.

I've been an editor too long to stop thinking first and foremost in terms of content. But I've learned also that content by itself is a sterile dead-end. It needs to be combined with messaging for community to perform; it needs to be forwarded, shared, commented upon, and criticized. Because messaging is the core community app.

Make One Friend, Build a Community

I spend a lot of time thinking about the difference between a community site and a regular, ol' commercial site. Where's the line in the sand that separates the one from the other? What are the necessary ingredients in community soup?

I think this is a useful way of thinking about it: a community website is a place where you can make a friend.

At first blush, that seems too simple, but that's actually the idea's strength. It cuts past individual features -- forums, messaging, user ratings, user-generated content -- and, instead, asks the core question: can your visitors use your site to expand their social circle?

Note that this isn't a quality rating: the fact that you enable your visitors to make friends by way of your site doesn't mean that your site is well-designed. You might have a painful sign-in procedure, a bewildering IA, or some other problem that punishes visitors and makes them wish they were somewhere else even as they manage on the margins to connect with other visitors. We'll deal with the question of what makes a good community site some other day; now the question is more basic.

So let's think for a bit about what web features allow visitors to make friends and, in the process, create your community for you.

  1. Forums. This is easy. Threaded communications and persistent identities are the quickest way for people to interact over time and get to know one another (for better or worse). There's a tension here, though: the easier you make it for your members to contact each other and transition their relationship from online to offline, the more likely community is to form, but the more you expose your company to potential liability if some of those offline contacts go wrong.

  2. Messaging. This might be the core community function (more on this in a later post). Friends contact each other; they have private conversations that are not intended to be shared with the group. Enable that via your site (email links, IM functionality, private drop-boxes, etc.) and you enable community.

  3. Shared calendars and other event-planning tools. Online community is, by its nature, a very limited thing as long as it stays online. It doesn't really come into its own until online and offline communities merge into one another. Evite is not a community site per se, but it enables community by making it easier for people to get together. See above about potential liability, though.

  4. Friends lists. Another obvious one. MySpace managed to turn this feature into a competitive sport: its members compete with one another to assemble larger and larger bodies of "friends." That's a neat trick: get your members to compete with each other to build your community, and you might be the next MySpace (though hopefully with a better page layout).

  5. Personal blogs. This is a hard one to get past management. It's OK to give your members a voice, but hosting those voices in a way that might seem to give official sanction to what Joe Random User is saying? Dangerous! Or so the lawyers think. If you can make the argument for it, though, personal blogs can be a very powerful community tool. Blogs draw in traffic, they link to each other in ways that spread the message more powerfully than anything else you're doing, and the kids love 'em. You just need to make sure that they're happy enough with the product or service you provide that they'll be evangelists rather than an unruly mob.

Items that don't enable friendship include the following (and note again that I'm not saying these are bad features, or that you shouldn't include them in your site; rather they simply are not core community features):

  1. User-generated content. There is nothing about UGC per se that builds a community. It's a great way to generate traffic, but you won't have a community until you do something with that traffic (i.e. give a way for visitors to your site to connect with each other).

  2. Marketing. This is a message that the suits don't want to hear. They see your community site as a marketing platform, and why shouldn't they? After all, it's likely that someone's marketing budget is paying your bills. But marketing not only doesn't build community, it has the power (if used unwisely) to destroy community. Use a subtle hand, or see your members abandon a site that they rightly understand to be more about the company than it is about them.

  3. Editorial content. It took me years to realize this. I've spent a decade developing content for community sites, and I still believe that you can build a better site with content than you can without it. However, I've come to the reluctant realization that content doesn't build community. Traffic doesn't lie, and editorial content performs very poorly next to the traffic to true community elements such as forums. Good content can still enhance community, but it's frosting on the cake, not the cake itself.

  4. Licensed content, such as audio or video. If you've got the budget to license audio or video from major media companies, go for it. You'll get plenty of traffic (along with a hefty bandwidth bill). But, as with UGC, you'll need to do something with that traffic to turn it into a community. A million isolated visitors staring at a streaming video are not a community -- not until they're watching those videos in a chat room or otherwise connecting with each other.

  5. Corporate blogs. See above, re: marketing. Corporate blogs are a trifle better, assuming you enable comments. But top-down messaging are antithetical to community. Build features that allow communication to flow from the bottom up (or from the bottom in any direction) and you're on a better tack.

So there you have it: if your site allows a visitor to make one friend -- a real friend, with real potential for taking the relationship offline -- you have a community. If you don't, you might have a good site, but you'll never have a community site.

You Don't Build Community, You Enable It

I've spent the better part of my career attempting to do the impossible: I was trying to build community.

That might seem an odd message on a web community blog, but bear with me. It is certainly possible for a community to form around a website. The question is, who is the responsible party?

When I was a history major in college, there was a lot of hand-wringing in class about the "great man" theory of history. This is the understanding of the world by which history is formed by the actions of great and famous men: men like Alexander, Napoleon, Lincoln, Churchill, and JFK: the sort of men who you've probably seen in statue form. The great man theory has taken a lot of abuse of late, as historians have begun to realize how powerful social movements as a whole can be. Could Hitler have had the effect that he did if Germany as a society was not already ready for his message? And could Kennedy have put a man on the moon (posthumously) if the ground was not already prepared by generations of scientists and amateur tinkerers working in obscurity?

In short, we give these great men more credit than they probably deserve. They influence, but they do not create. And I would like to propose a similar understanding of web community: that the "great content" theory of community is largely a fallacy.

It's an enticing notion: build great content, and then spin a community out of the audience that forms around that content. By that theory, good content should create community, and great content should form community even faster. If community doesn't form, there must be something wrong with your content; go back to the drawing board and try to come up with something more compelling. When you hold to this point of view, you always feel like you're right on the verge of figuring it all out. You were close last time; this time you're sure to put content out there that will be like a powerful magnet, drawing in readers and convincing them to return to your site again and again. And if it doesn't happen the next time, surely you'll learn lessons that will make it happen the time after that. Or the time after that.

Except it doesn't work. And the reason it doesn't work is something I've touched on before: most of your members don't care about your content. Sure, they might read something and enjoy it, and if you put tons of time and money into a dazzling multimedia display, they might even forward a link to their friends. But the amount of resources you spend on content never seems to bring a proportional increase in repeat traffic. Meanwhile, blog posts breed comment threads that head off on tangents and digressions that carry on for weeks or months, and forum posts encourage repeat traffic that puts to shame anything else on the site. What's the problem?

No problem at all; it's just a misunderstanding. You don't build community. You enable it. The only person who can build community on your site is a member. He or she can make a connection with another member; you cannot do that for them. If a community is to be formed on your website, they will do it, not you.

So what do you do? Enable it. Take some of the time and resources that you might have spent on content and spend it on features instead. Give them a messaging system, so they can contact their friends. Give them a friends list or an address book, so messaging is a snap. Give them a calendar tool, so they can use your site to schedule get-togethers. Provide widgets that allow them to track their friends and keep in touch. Allow them to upload content and share it with their friends. Give them what they need in order to make a connection. Enable community rather than seeking to build it.

That's not the end of your job, of course. Since you're not a member of your community, it's a near-certainty that you don't understand what your members want. Call some in for user testing. Track their movements through the site and measure the traffic on your tools. If the calendar tool involves multiple steps, what's the abandonment rate before an event is scheduled? If an invite is sent out, can you track whether or not the event went off successfully? Measure and refine and tweak continually, as your members discover what they want to do over time.

If you work hard and maintain your focus over the long term, you might enable community on your site. You likely won't ever build it, but that's OK. There are thousands of people out there who are ready to do that for you.

Five Tips for Effective Social Design

I came across this a while ago. It's short -- just the summary of a conference presentation -- but it's a goldmine of insight into web community building:

Let's look at those five tips, one by one.

1. Personal value precludes network value: To make a strong social site, you've got to start by making a good personal site. If the features you offer don't serve a solo user, it's unlikely your users will stick around long enough to become social.

This is an easy trap to fall into when you're designing community features: you build out something that will be great among a group of friends, without ever addressing the question of why that group will be coming to your site in the first place. "If you build it, they will come" -- all well and good, but you need to reward visitors on their very first visit, or there won't be a second visit, and the odds are excellent that initially they're coming as individuals, not as members of a group. So as much time and money as you've put into that excellent event-planning tool, with all the social networking gizmos you've plugged into the back-end, your home page had better attract clicks to content and tools of value to individuals, or those individuals won't be around long enough to become members of your community.

The key here: think about customer lifecycle. People will come to your site in every moment within that lifecycle. Some will be vaguely interested in your site/product and looking to learn more, others will be engaged and looking for specifics, some will be committed members looking to get full value from what your site offers, and then there will be the post-users who are bored/burned out on the experience and need some reason to come back into the fold. Look at your home page design: is there something on that page for every one of these people? If not, you probably need to revise your design, or figure out some way to engineer different entry points for the various points in the customer lifecycle.

2. Tie behavior to identity: What you do on the site should describe you more than what you say about yourself in your profile.

Historically, one of the primary obstacles to online communities is the way people tend to behave when they're online. If you've spent more than 10 minutes on Usenet or some other online forum, you know what I'm talking about: people online tend to be jerks. Flamewars break out because otherwise reasonable people, when cloaked in online anonymity, check their social conventions at the door and act like jerks. Racism, homophobia, sexism, obscenity, professions of extremist political beliefs, immature behavior: it's all to be found online, and sometimes it's so prevalent that it's hard to locate the wheat amid all that chaff.

The challenge for those of us who would like to build online communities, then, is to find some way to protect our users from mistreatment at the hands of other community members. Moderation is one option, but that can get expensive, and it can introduce a time-lag in the experience that substantially degrades the member's experience. (Not to mention the outcry that will certainly result as soon as a member is banned or a forum post is rejected for a reason that the community finds to be unjust or otherwise insufficient.)

Or you can take another route: lessen the anonymity. Allow your users to select any username they like -- and by all means don't make it mandatory for them to disclose real-world information about themselves -- but build your site in such a way that their actions persist and follow them around the site. Wikipedia vandals are outed by the fact that their edits can be uncovered -- a fact which has proven embarrassing for a number of companies and politicians caught falsifying their own entries. YouTube uploads are associated with the user who uploaded them. Ebay sellers or buyers who tend to piss people off will find their ratings a serious drag on their commercial success. When your actions follow you around, you'll be more careful about the way you present yourself.

In a nutshell: community depends on trust. Give your members a means by which they can tell whether or not they can trust the next guy, and they'll be more open to joining your community.

3. Give recognition. When your users contribute to the site, applaud them for it. Make those contributions publically visible. But make it temporary: once a user is recognized as a top contributor, let them fall off the map if they don't stay active.

This blends into something that's become my mantra of late: you can't build community, you can only enable it. You can design a terrific site full of wonderful features, but if no one uses them, you have a ghost town. Community doesn't exist until you have a multitude of people acting on their own and choosing to interact with others. You don't build community, in other words -- your members do, using the tools you've given them. Their contribution is critical, absolutely indispensable, so how could you not recognize and celebrate that?

Digg is the poster child here; its members compete with one another to fill the site with the most engaging content, because they crave the recognition they receive in doing so. But the same principle works on a much smaller scale. Think of a forum. Let's say that, when you look at a post, you can also see how often the person who posted it has contributed to the site. Towards the bottom of the post you can rate it on a scale of zero to five stars; now the poster has a cumulative rating, based on how well his or her posts have been rated. Those two features are all it takes to celebrate the most active and well-regarded members of your forum community. They do something nice for you, and you do something nice for them in return. (Or, if they choose to vandalize your site, the same system works to stigmatize them.) It's all good.

4. Show causation: If you're going to ask people to participate, make it clear what participating does for them.

This might be the #1 failing of the corporate site: it puts its own interests ahead of the user. Companies build community sites because they expect, in the long term, those sites to produce higher sales or greater customer satisfaction or some other measurable outcome. And, of course, there's nothing wrong with that; most companies are not in the business of doing stuff for free. But in my experience that situation lends itself to a company-first mindset: we build our sites and our features to address company needs, not user needs. And that is why so many corporate sites fail.

It only takes a moment to understand why: most customers couldn't care less about your company. They want to know what your site does for them. If your site doesn't cater to their needs, they'll find another site that does. Simple as that. So if, when you think about community, all you see is dollar signs, you're in trouble. Ask, instead, what your online community provides to its members. If you're not giving them reason enough to come back time and time again, you're already dead.

The moral: start first with the question, "Why will users come to my site rather than all the others out there?" Give them a reason to come, and then give them many reasons to come back. Then, once visiting their site has become a part of their lives, you can think about marketing to them.

5. Leverage reciprocity: People contribute to social sites because they want to see what other people say about their contributions. Make it easy for people to leave feedback, compliments, and awards.

We all want to be appreciated. When we contribute something, we want to be thanked for it. When we create something, we want people to like it. No matter how thick your skin might be, there's a part of you that wants to be loved and appreciated and embraced. In fact, that's a pretty good picture of community right there: it's the group of people whose opinions matter to you and who value your opinion in return.

The good news is this is dead easy. At the bottom of every forum post, there should be a mechanism for rating its contents. You should always be able to rate user-generated content (and flag it if the content is objectionable), and allow people to add comments as well. Create online friends lists that allow your users to signal personal connections. When you give people a means of adding approval or disapproval to the actions and contributions of your community's members, you create a very strong incentive for your members to make frequent positive contributions. And that, in a nutshell, is the picture of a thriving community.

So there you have it: five tips that might not make your community site the best thing since sliced bread, but I can guarantee that they will improve your site significantly.

Thursday, June 5, 2008

Waiting for the Crash

The analysts are doing what they do best: analyzing stuff. This morning, the news was about a study predicting that use of virtual worlds will explode over the next ten years:

22 percent of global broadband users will register for one or more virtual
worlds over the next 10 years. This will expand the virtual world market to one
billion registrants, with roughly an eight billion dollar services opportunity.

"Despite a multitude of challenges, virtual worlds present a unique marketing opportunity to target a highly sought demographic, and virtual worlds should be part of a company's marketing portfolio," according to Harvey Cohen, President of Strategy Analytics.

Wow! Who wouldn't want a piece of that? Of course, as always with things of this nature, it's easier to predict the trend than it is to make it real.

The first question to ask, naturally, is whether the study has any validity. My suspicion, where all such future-trend-predicting are concerned, is a familiar misuse of statistics in which short-term developments are extended without any adjustment into the future. This is the fallacy that led one sports television commentator, back when I was but a lad, to predict that women would soon be running marathons faster than men because their times had been improving so rapidly. Lost on this commentator was the fact that, at the time, women had been competing in the marathon for only a few years, and their records were relatively soft. Women today are running the marathon faster than ever, but men are still a little faster. That's what happens with long-term trends; rarely do they indefinitely follow a straight line.

So is the virtual worlds prediction founded on a similar fallacy? A handy chart gives us some indication:

There's good news and bad news in that chart. The prediction concerning registrants as a percental of overall broadband users shows the percentage leveling off at around 22% of all users. However, the orange bars hold the bad news: "active users" will allegedly increase arithmetically, without a break, for the next ten years and beyond. Five years from now, this is likely to seem naive.

It's a very seductive idea, though. Who doesn't enjoy the prospect of a future that comes with a fully-realized virtual world? I think we're all still waiting for Snow Crash, the seminal sci fi from Neal Stephenson that portrayed an online world so compelling that the book's protagonists clearly prefer virtuality to the real thing.

Invent a virtual world like the one in Snow Crash, and you should be able to print money. It would be the ultimate online community. But let's assume that you managed to do such a thing; would anyone want to use it?

Think about it: Snow Crash shows a world in which people go online to hang out. It's pretty much the same for Second Life or even World of Warcraft: that's where online friends go to congregate. And who has the time for that sort of thing? Between work, the daily commute, eating, and sleeping, it's a challenge for me to find quality time to spend with my wife. I can barely imagine the changes that would have to go down in my life before I started hanging out in virtual worlds.

I used to have the time for that sort of thing when I was a graduate student. I would certainly have done such things when I was a teenager, except such things didn't exist when I was a teenager. And that's the key: virtual worlds appeal exclusively to those who have time for them. That means kids, teenagers, and kids right out of college. The rest of us are too busy leading our lives to get involved.

So before the your eyes fill up with dollar signs in contemplation of virtual worlds, consider the challenges. Your market will be functionally limited to people between the ages of 6 and 24. Those customers will certainly subscribe to more than one virtual world, since if there's any one characteristic that most young people share, it's a fickle temperament. They will not be there to shop, they'll be there to meet with friends, and they'll see your advertisements and merchandising as an unwanted intrusion. And when the winds shift--as they certainly will--and something new comes down the pike, they'll be gone. Remember Friendster? No one remembers Friendster. Some day it's likely that no one will remember MySpace, Facebook, or Second Life. This is what happens to companies that cater to a fickle, trend-driven user base.

Virtual worlds can and probably will get bigger and more prominent. But we'll still be waiting for Snow Crash for a long, long time.

Wednesday, June 4, 2008

Sony takes the shotgun approach to online gaming community

Sony's down with web community. They think it's a good way to promote the video games they've developed for their PS3. Only one problem: they don't know what "community" is:
  • LittleBigPlanet - While they're "not decided yet" on exactly how to integrate the game into a community website, if any mechanic seems ripe for such a project, rankings and stats for LBP's user-generated content seem like a sure bet.
  • Killzone 2 - There already is a site for Killzone: Liberation, but adding "rankings and statistics" for Killzone 2 is the plan.
  • SOCOM: Confrontation - They plan on launching a community site - think "My SOCOM" - on There's nothing there now, and is taken.
  • Motorstorm: Pacific Rim - The Motorstorm representative wasn't very interested in providing any details, but she told us that they would have a "community website" as well.
  • Resistance 2 - Insomniac has perhaps the most ambitious plans for community content, dedicating an entire development team to the task. They're "overhauling", adding things along the lines of "facebook-like features."

Do you see Sony's overarching strategy for online community there? No? That's because, quite obviously, there is none. The reason these sites sound like the random shots in the dark is because they're treating "community" as a buzzword rather than a product.

I don't want to single Sony out for what is a common problem in the video game industry: they all want community websites, but they haven't actually defined community or how to develop it. Some things are easy, like forums: give people a way to talk to each other, and you have one style of community. And maybe statistical leaderboards will also promote community, on the premise that people will come to see their rankings and stay to brag about it to their online friends (presumably in the forums?). Finally, if you want to be ambitious, try to build some social networking into the site, because that's what Facebook enables and we all know that Facebook rocks (see previous post on the $50 billion valuation). But be prepared for disappointment, because quality web features are not easy to build, and the web designers you're working with probably don't have the skills or experience to do that well.

The bottom line in this market is we can count the number of truly great video game community websites on the fingers of one hand: there's, and then a bunch of sites that wish they were Building another isn't easy, as I'm sure Bungie will tell you; after all, that community has been many years in the making. Even if a developer managed to match's features, there's no guarantee that a passionate, engaged community would form around those features. Projects that require lots of money and time yet have uncertain outcomes are not easy to get approved, least of all in the gaming world where a title that was released six months ago is already yesterday's news. Hence the predictable outcome: Sony's new sites won't amount to much, and no one will really notice because by the time they've failed we'll all be playing different games anyway.

This is a situation that's ripe for improvement. If you build a site without a vision or a strategy, you're not likely to succeed (or know whether you've succeeded once you're done). But how many people truly have that vision? "Community" is an easy word to toss around, but a difficult reality to create.

So, here's my quick prescription for community:

  1. Start with a vision: "People will come to our site -- repeatedly -- because of X." Hash it out, print it out, hang it on the wall where everyone can see it.
  2. Add in features that clearly and explicitly support that vision: "Our site will enable community in the following way." Anything that doesn't support the vision should be left out of the site plan, lest it clutter up your UI with stuff that doesn't contribute to success.
  3. Plan for post-launch analytics: "We expect our community to do the following things, and this is when and how we'll measure those outcomes."
  4. Finally, assume that you won't get it right on the first try; after all, Bungie didn't.

Verse the First

Web community: what is it?

At first blush, that's a simple question. After all, isn't the entire Web 2.0 experience supposed to be about community? The hot sites are social sites: MySpace, Facebook, even Google now that it's pushing its OpenSocial idea. The new web is about sharing, and sharing defines community, which is why YouTube traffic is growing while solitary experiences such as television-viewing are shrinking. Community is easy: just boot up a browser and take a look.

Except...what is community, anyway? When I think of community, I think of family and friends. The word has a connotations of intimacy for me. MySpace is more like being in a New York subway terminal: lots of people, lots of sound and color; it's exhilerating and alienating at the same time. I've spent more time randomly clicking through strangers' profiles on MySpace than I have actually connecting with someone who's part of my social circle. It's a very busy space populated by lots of people, but is it community?

Then there's the question of money. Facebook is hot; its traffic grows while MySpace's remains flat. It's been valued (in some quarters) at $50 billion. It's hard to get your head around that number, especially when you consider that Facebook has never once shown a profitable quarter. If Facebook is the most successful web community out there, what does it say that they still haven't figured out how to make any money?

This blog will concern itself with such questions. I don't have answers. I don't have readers or Twitter followers. I'm just one guy who's paid to foster online communities and who would like to do his job better. This blog will be the record of my explorations of web community in its many manifestations.

Welcome aboard.