What does it take to build an online community?
Talking about the promise of community is one thing, but what are the design principles that make it easy for community members to find one another and to share their interests? Are all community members equal, or are some of them more equal than others? Do community and commerce mix?
Those are questions that occupy the minds of Net-company leaders for whom community is a crucial element of their business strategy — whether their business is a highly specialized B2B site or an entertainment site aimed at teenagers. And Cynthia Typaldos has been asking and answering those questions for years.
Typaldos, 50, is a true Web pioneer. A software developer by training, she earned an AB in chemistry at UC Berkeley and attended graduate school in computer science there as well. (She also received an MBA from MIT's Sloan School.) Later, she held senior-management positions at Data General and Sun Microsystems.
Then she got with the Web program. In January 1995, she launched GolfWeb (www.golfweb.com) , a now-popular site for golfing enthusiasts (and currently part of CBS SportsLine) . An important element behind the success of that site has been the GolfWeb Players Club, an online community whose members pay an annual fee of $39.95 to get discounts on merchandise and to trade tips and war stories about the best courses and the toughest holes.
As she worked to develop the Players Club, Typaldos thought more and more about how online communities evolve and about what it takes for them to thrive. Her research base broadened from the tens of thousands of postings that she studied on GolfWeb to the work of sociologists like Marc Smith, now the resident sociologist at Microsoft. In November 1997, she and Mark Waters, who was then an executive at GolfWeb, founded RealCommunities, a company that designs and builds infrastructure and services for other Web communities. Typaldos's interest in online communities even led her to teach a class at UC Berkeley Extension called "Web Communities for Content, Commerce, and Customer Retention."
Here, in an interview with Fast Company, Typaldos discusses the ideas and design principles that define real communities.
Most Net companies talk about building "community" on their sites. What does it take to make community real?
Communities form around a shared purpose. They're made up of people who come together to do something that they couldn't do alone. Think about a PTA, or a neighborhood association, or a Web-based meeting for systems administrators. Each of those communities exists to accomplish something that's important to people who are involved in an activity — and that they can do only as a group. If you don't provide people with a compelling reason to come together, and if you don't give them the means to accomplish that shared sense of purpose, then your community will fail.
Now, there's a big difference between people who have a shared purpose and people who merely have something in common. You can't build a community around people who own white cars, for example. A couple of years ago, Citibank tried to create a "community" around its checking-account customers. But what's the common purpose or the shared passion among checking-account customers? They're just a random group of people who happen to have checking accounts at Citibank. That fact makes them similar to other account holders, but not in any way that really matters.
How is a Web community different from a real-world community?
It's not. We're not changing human nature; we're just changing the tools of communication. In the class I teach, I ask that very question, and I always hear the same answer: "Online, you can't see other people." But do blind people have communities? Yes, they do, just the same as sighted people do. "But they can hear the intonation of your voice," my students will say. Well, don't deaf people have communities? Of course they do.
Tone of voice, facial expressions, body language — these are tools that people use to communicate, but they aren't fundamental to building community. In fact, most ideas about online communities focus too much on the digital tools that people use to communicate: bulletin boards, chat rooms, email lists. What really make or break a community are issues of trust and identity, clarity of purpose, and boundaries — the same issues that affect real-world communities.
Let's talk about those issues. You mentioned "trust and identity." So much online interaction is relatively anonymous. Does that pose a problem for community building?
You can't build trust or a viable community environment among people who don't have persistent identities. Repeated encounters are what allow people to work collaboratively. But people can have persistent identities that are anonymous — as long as people can recognize one another.
The GolfWeb Players Club is a perfect example. The shared purpose of this community is for members to help one another improve their golf game. So people want to know: How good is your game? Where do you play? What kind of clubs do you use? They don't care about what you do for a living, what kind of car you drive, or how much money you have. In the Players Club, your identity is connected to the nature of your game — not to your name or your demographics. In any case, without that identity, it's hard for other people to judge the nature of your participation.
Now, "identity" is different from "reputation" — another key factor in virtually all Web communities. Identity is based on who you say you are. Reputation is based on what you do and what other people think of what you do. Your reputation provides a context for members of a community to judge the value of your contributions.
On Amazon.com, members can vote on how useful they found another member's book review to be. On eBay, members can rate the quality of their experience with other members, and those opinions get aggregated into a symbol for each member — a star whose color varies, depending on their rating. On the Motley Fool, members get color-coded stars that appear next to their name when they post messages. These "badges of Foolish achievement" indicate the number of messages that members have posted — how much they've contributed to the community.
So reputation matters because it helps members of a community make judgments about you, even before they actually deal with you?
You want to have a good reputation, because that helps you accomplish what you're trying to get done in the community. But making your reputation visible to others is also a great way to persuade you to stay in the community. We are all driven by status symbols. It's hard to leave a town in which you've become somebody important. Imagine if your company said, "We're going to make you a vice president, but only on the condition that nobody else knows about it." Would that feel like much of a promotion?
Smart communities understand this principle. On eBay, the highest status symbol is a shooting star. To get a shooting star, you need 10,000 positive-feedback postings from the community. (You also lose points for any negative-feedback postings.) Let's say you have 9,500 positive-feedback points. Are you going to leave that community and go somewhere else, where you're a nobody? If a community has a way of awarding status that is visible to other members, people will strive to achieve it. And a site's status symbol can be a very powerful piece of intellectual property as well.
What else is critical to community?
Boundaries. If just anyone can join your community, then it's not really a community. So you must set boundaries: Who can join? Who can't? How do people join? Who decides whether they belong?
There are lots of ways to implement boundaries. To become an employee of a company, people have to go through an interview process, and they have to receive an offer. On the Web, you can give potential members a quiz. On a stock-picking site, maybe a member's picks have to perform in the top 10% of the community before he or she can be part of an elite subgroup. On an IT site, maybe members need to have a certain Java certification in order to participate.
One reason for having boundaries is to make sure that the people who join a community actually take part in it. But even if you define your boundaries clearly, you can't expect everyone to participate. I often hear people say, "I want every member of my community to participate." That's not natural. The participation rate is almost always the same: Only 10% to 30% of a site's members are active at any one time.
In the real world, of course, it's easy to identify active community members: They're the ones who show up for community events. At a PTA meeting, it might seem like everyone in the organization is involved. But in reality, maybe only 10% of all members actually attended the meeting in the first place.
The virtual space of the Web is a little different. As a Web-community builder, you have to let people discriminate between active members and passive members. For example, the GolfWeb Players Club, the site brings up everybody who meets the criteria that you enter - without discriminating between who's active and who isn't. So, if only 10% are active, and your search brings up 10 members, chances are that only one member will communicate with you. Your reaction will probably be "Hey, most people here aren't doing anything." But that's true of every community! What you really want to do is focus on those who are active.
So that's a design principle: Differentiate between active members and passive members. Do you have other design principles?
A sense of history is vital for an online community, especially as it grows over time. But forgetfulness is part of history too. It's easy for a Web site to "recall" everything that's ever happened there, but what you end up with is a big mess. The important question for a community is: How do you remember what you need to remember in order to develop a shared history?
There are a few principles that apply here. You can allow members to delete things that they think are no longer relevant. You can have only the most recent posts appear on the site, or you can have events appear in reverse chronological order. You can also let members "age" things - by letting them specify that their posts should be deleted after six months, for example. While member-generated content should be eternal, individual transgressions should have a statute of limitations. Members should be able to redeem themselves: A successful community learns from its mistakes.
The flip side of history is what I call "expression." Every community has a shared sense of itself. But that doesn't matter much if members don't know what's going on. Members need to have an easy way to take the pulse of the community at a glance: What topics are under discussion? Which members are reaching their goals? Who's on the site now? How many people joined today? Who are the top 10 posters? Are we accomplishing what we're here to accomplish?
If you went to a PTA meeting, you'd never have any questions about what was going on. People would either be talking about a topic, or they wouldn't be talking about it. Online, the points of interest aren't as immediately obvious. So you need to tell the community, "Hey, everybody is talking about this."
Can community and commerce mix?
Communities need to provide a way for members to exchange things of value. On eBay, people come together to exchange collectibles. But the items of exchange don't need to be objects. Information, expertise, recommendations, and positive feedback work just as well. But community members must be able to say, "Okay, I'll do this for you if you'll do that for me." In that sense, there's no great philosophical divide between community and commerce.
Are there limits to growth? How big can a community get before it suffers?
Here's how I think about size: How many people are there for whom the shared purpose of a community is essential to their sense of identity?
A real community needs to be one of the five most important things in the lives of its members. That's a pretty serious limitation on size. Let's say that a person is a parent, has a profession, and pursues a hobby. That person probably has time for one other thing in his or her life. So any community must be able to attract a reasonable number of people for whom it can become that fifth thing. That number determines how large your community can grow.
Katharine Mieszkowski (firstname.lastname@example.org), a former Fast Company senior writer, is a senior writer for Salon.com. Contact Cynthia Typaldos by email (email@example.com) , or visit RealCommunities on the Web (www.realcommunities.com).
Sidebar: 12 Steps to Real Community
Cynthia Typaldos, president and CEO of RealCommunities, teaches people how to build online communities — and sells them the software to make communities work. Her "12 Principles of Civilization" will help you find out whether your community is for real.
- Is there a necessary shared purpose that we accomplish together?
- Does each member have an identity? Can we tell who's who, even if members remain anonymous?
- Are we able to share information and ideas that fit our purpose?
- How can we build trust? What tells us that it's safe to deal with other people in the community?
- How do we form reputations? What lets us build status?
- Have we created ways to work together in small groups?
- Is our environment a shared space that is appropriate for our goals?
- Do we know who belongs in our community and who doesn't?
- What's our system of governance? How do we regulate behavior so that it supports our shared values?
- Is there a system of exchange that allows us to trade knowledge, support, goods, services, and ideas?
- Are we able to express our group identity in a timely way? Are we aware of what other members are doing right now?
- Do we have ways to review our history and to track our evolution — and leave behind what's best forgotten?
A version of this article appeared in the September 2000 issue of Fast Company magazine.