Behavioral Differences Between Four Types of Online Networks

There are many different types of online networks and each encourages different user behavior. For the purpose of this discussion, I would like to create 4 different categories of online networks: centered-public, centered-private, decentered-public, and decentered-private. “Centered” means everyone is looking at the same thing. “Decentered” means everyone is looking at a different thing. Facebook is a good example of a decentered network; when I login, what I see is different from what you see when you login. “Public” means anyone can view the content, and “private” means only selected/invited users can view the content.

Here are some examples for each type:

  • Centered Public: discussion forums, blogs, eBay, Craig’s List
  • Centered Private: Groupware, Basecamp, private email list
  • Decentered Public: Twitter
  • Decentered Private: Facebook

eBay is an online network where everyone is looking at the same content and layout, so it’s “centered”. And, it is “public” in the sense that anyone can view the content. Some contents are reserved for registered users, but even then, anyone can register at the site. Nobody at eBay needs to approve your registration. This is important to note; registration does not automatically make a site “private”. If anyone can register and view the same content, the site is public.

Basecamp is an online project management tool. It is a collaborative application that hosts many private groups. One group does not talk to another. It also provides a way to create sub-groups within a group where each project can have its own members whose content is protected from other members. With private-centered network, registration is not usually open to public. A group manager usually registers and invites the members. Within each private network, all its members see the same content, so it is centered.

Twitter is a micro-blogging site with social networking capabilities. After you login, you see your own page which is different from everyone else’s, which makes it decentered. The biggest difference between Facebook and Twitter is that the latter is public. There is a way to make your content private, but Twitter’s main purpose is to share information publicly, and this is where Twitter thrives.

Facebook is a social networking site to promote personal networking online. It is private in the sense that only your “friends” can look at your status updates or “Wall”. Although there is a way to make your content public, Facebook primarily encourages personal and private conversations. The site is decentered like Twitter where my home page looks very different from your home page.

We humans need to maintain our relationships within social groups as well as our own personal groups of friends. In the former, each person belongs to various groups with different centers. In the latter, each person has her own group where she is the center. The 4 different types of online networks above satisfy these needs. For this reason, decentered networks are conducive to more personal networking whereas centered networks are conducive to achieving shared goals. Although it is trendy now to add social networking features to every site, depending on what you want to achieve on your site, you need to be careful about which type of network to adopt.

Whether our communication is private or public makes a big difference in the content itself. When we know that we are writing for a small group of close friends or colleagues, we tend to use more specific examples that our friends know but are not necessarily known by the general public. We also tend to be more daring or direct because we feel safer. We might also share highly speculative ideas or opinions with our close friends, but not with the public. Our contents published to the general public are likely to be misunderstood by some groups of people. We need to take certain precautions when publishing to the general public. These differences are so critical that even one wrong word can trigger a series of problems. For this reason, communication to the general public usually takes a lot more effort, as its consequences are much wider and greater. When communicating, we need to know who our audience is, and every piece of communication has a certain degree of trust that must be assumed for it.

Much of the content on Facebook, for instance, is for private consumption only. Facebook is designed in such a way that we feel comfortable sharing more personal matters. If Facebook were to suddenly announce that they were shifting their model to decentered-public, like Twitter, I’m sure there would be a huge uproar of protest. In fact, I doubt it would be possible for them to do so legally. So many people would panic about all the contents that they’ve already posted on the site.

Interestingly, Facebook seems to be shifting towards that model gradually. I believe it was prompted by the success of Twitter. They probably felt threatened by them, and decided to add features that are more public (such as Facebook Fan Pages, or the “Everyone” button for status update). I find this a bit problematic. The more public features that they add, the less secure people would feel about sharing any personal information. For instance, what happens if you comment on someone’s status update which is published to “everyone”? Does your comment also become publicly viewable? The answer to this question is irrelevant to the negative effect it would have on the strength of Facebook because most people would simply err on the safe side, which means they would gradually stop sharing their personal thoughts, opinions, and information. This can eventually make Facebook less useful. If everything is assumed public, why not just use Twitter? We love Facebook because it allows us to share content privately with our friends and families. If the line between private and public gets blurred, it would create a market for more strictly personal social networking sites, and we may migrate over to them en masse, as we did from Friendster to Facebook. If that happens Facebook would be a master of none; it would be a mediocre way to reach the general public as well as our private friends.

Decentered networks do not encourage collaboration because everyone is looking at different things. For this reason, it is not an effective organizational tool. Centered public networks are not particularly effective either because they are open to anyone; the team becomes too large and arbitrary. This leaves us with centered-private network.

In traditional network applications, the interaction takes place between user and information. Let’s call this user-to-information model. (See Entering the Interaction Age by Andrew J. Milne for more elaborate explanation of it.). For instance, a business may have a contact database installed on their server, and all the employees might access it, but this is not an online network that we are discussing here because the users do not interact with one another. Other examples of user-to-information model are medical record database, library database, accounting system, booking/reservation system, and inventory system. In comparison, centered-private network, such as groupware or collaborative software, is an application where the interaction takes place between users; user-to-user model. A good way to tell if an application is user-to-user or user-to-information is whether it’s possible to share one login among all the users. In user-to-information model, the password is protecting the information only, so there is no real need for everyone to have his/her own login. If the application is truly user-to-user (collaborative) like in Basecamp, everyone using the same login would not work at all. In this sense, private-centered network is a relatively new phenomenon, and has gained in popularity just recently.

Successful implementation and integration of a collaborative application can be quite challenging because collaboration is not something we can force on people. In the user-to-information model, a business owner could require an employee to keep the inventory database up-to-date, and there are ways to account for his success or failure. With a collaborative application, this is not always possible. You cannot, for instance, force your employees to share knowledge, opinions, and ideas because they could always pretend like they don’t know anything, or have no opinions or ideas. Each member must feel inspired or motivated to collaborate. Those who are indifferent to the overall success of the organization would not bother. The design of the system could do only so much in providing inspiration or motivation for its users. Because of this, even a perfectly designed groupware could fail to motivate people to collaborate. In most situations, the motivation to collaborate must already exist before a collaborative application is installed within the organization. Because people’s motivation can be fragile or limited in degree, when designing collaborative software, usability becomes a bigger concern. If the application is hard to use, the users may not feel motivated enough to overcome the difficulty in using it.

Because motivation can be fragile, and because true collaboration cannot be forced on people, the best way to implement a collaborative application is to mimic Jane Jacob‘s approach to urban planning. That is, rather than forcing your grand vision of how things should work, you observe how people are already doing things, and find ways support the positive aspects by providing better tools. This usually means you need to start small and simple, and grow the application organically based on your observations. If you installed a large complex application to entirely change the way people are doing things now, you would probably face stiff resistance.

It is interesting to note that many of Jane Jacob’s ideas for urban planning can be used for developing collaborative software. This is because understanding human behavior is the most critical part of organizing a group of people. Whether it’s on line or off line does not make much difference. In real life, we organize in many different ways: friends, families, work groups, volunteer groups, political groups, social movements, support groups, religious groups, etc.. Depending on the purpose of each group, the way to organize and manage is different. Simply using one scheme to organize them all would not work. In some groups, there are hierarchies while others have none. In some, participation is required, while others are voluntary. Some have shared goals, while others do not. Some groups have diverse members, while others are organized around a very specific demographic. And so on... Just as in reality, online networks have to be designed differently depending on their organizational characteristics. Many Web developers get frustrated when their users do not participate in the online networks they created. This often happens when the developers did not understand the behavioral differences between different organizational types. For every project, they blindly use the one and the only method they know. In some cases they succeed, so they blame the users when they don’t. I believe that one of the first steps in developing an online network should be to understand the behavioral differences, and all other decisions should be based on that understanding. The differences are often very subtle and may not be obvious to casual observers, but they are factors that can make or break the entire project in the end.