UncomMarket logoUncomMarket
Community Playbook #2: The Two Forces That Make People Stay
Back to Blog
How It WorksMarch 27, 2026

Community Playbook #2: The Two Forces That Make People Stay

Social psychology says there are exactly two reasons people feel attached to a group. Understanding the difference changes how you design your community.

In Part 1, we talked about why most online communities die: people don't feel attached. They visit, they poke around, and they leave.

But attachment isn't random. Social psychologists have been studying group dynamics for decades, and they've identified two distinct forces that make people feel connected to a group. Getting these right is the difference between a community that grows and one that slowly fades out.

Force #1: Group Identity

In 1971, psychologist Henri Tajfel ran a now-famous experiment. He assigned people to groups based on completely meaningless criteria. Literally just random labels. And something strange happened: people immediately started favoring their own group over the other one.

No shared history. No real differences. Just a label. That was enough to create a sense of "us."

This is group identity in action. It's the feeling of belonging to something bigger than yourself. You don't need to know every other member personally. You just need to feel like "I'm one of them."

Researchers found five specific triggers that strengthen this feeling:

Categorization. Simply assigning people to a named group activates identity. Give the group a name, a logo, maybe a short description, and people start identifying with it. In one experiment, researchers used group names and uniforms, and attachment shot up. The name doesn't even have to be meaningful. In the MovieLens study we'll cover in Part 3, they used random animal names like "Tiger" and "Eagle," and it still worked.

Group information. When you show people what their group is doing, who's in it, and what the group is about, they start to see themselves as part of that collective. The trick is to present people as group members first, individuals second.

Homogeneity. People identify more strongly with groups where members seem similar to each other. "We all like the same kind of movies" or "We all care about the same cause." Emphasizing what members have in common strengthens the group's pull.

Performance analytics on a laptop screen

Competition with other groups. Nothing bonds a group faster than a shared rival. Wikipedia leans into this when it compares itself to Encyclopedia Britannica. Sports teams have known this forever. When there's an "out-group" to measure against, the "in-group" gets tighter.

Repeated exposure. This one comes from a 1968 finding by psychologist Robert Zajonc: the more you see something, the more you tend to like it. It's called the mere exposure effect, and it applies to groups too. If members regularly see updates about their group's activity, they gradually develop a stronger connection to it. Wikipedia's Community Portal works exactly this way, constantly showing members what other Wikipedians are working on.

Force #2: Interpersonal Bonds

The second path to attachment is completely different. Instead of feeling connected to a group as a whole, people feel connected to specific individuals within it.

Think about why you keep going back to that one group chat. It's probably not because of the group's mission statement. It's because of the people in it. You know them. You like talking to them. You'd notice if they were gone.

Four triggers strengthen interpersonal bonds:

Personal information and self-disclosure. When people share things about themselves, it opens the door to connection. Profiles with real names, photos, and personal details shift attention from "the group" to "the people in the group." Research by Collins and Miller (1994) found that self-disclosure is one of the strongest predictors of interpersonal liking.

Similarity between individuals. We're drawn to people who are like us. If a community can show you that another member shares your taste in movies, books, or whatever the community is about, you're more likely to feel a connection to that person. This has been replicated in dozens of studies since Newcomb's work in 1961.

Repeated exposure to the same people. The mere exposure effect works for individuals too. Facebook's News Feed was built on this principle: by repeatedly showing you what your friends are doing, it strengthens your connection to them. The key word is "repeatedly." A one-time encounter doesn't do much. Seeing the same person's activity over and over does.

Direct communication. Talking to someone one-on-one is the most powerful way to build a bond. Frequency matters here. The more two people interact, the more they tend to like each other. This has been known since Festinger's research in 1950.

A diverse group collaborates and discusses together

They're Not Mutually Exclusive

A community can use both forces at the same time. The GNOME open source project describes itself as "a worldwide community of volunteers who hack, translate, design, QA, and generally have fun together." That sentence hits both identity ("worldwide community of volunteers") and bonds ("have fun together").

Members of GNOME join sub-groups focused on specific projects (identity), attend in-person events to "meet old friends" (bonds), and contribute to a shared mission of building a Linux desktop environment (identity again).

Most successful communities activate both forces, even if one is stronger than the other.

Attachment Spreads

There's one more thing worth knowing. When someone feels attached to a group within a community, or to a specific person in it, that positive feeling tends to spread to the community as a whole. Psychologists call it affect generalization. It's similar to the halo effect: if you like one part of something, you start liking the whole thing more.

This means you don't have to make every member feel attached to the entire community directly. If you can make them feel attached to a sub-group, or to even one other person, that attachment will bleed over.

But does all this theory actually hold up when you test it? A team of researchers ran a six-month experiment to find out. That's what we'll cover next.


Community Playbook series: ← Part 1 · Part 3: One Feature That Doubled User Visits →