Where’s the Real Social Networking?
Among the slew of services from Google in the past year was Orkut, a social networking service. For many, Orkut was the first experience with the concept, although Friendster, Ryze, LinkedIn, and others came on the scene earlier. The idea behind these sites is that as social beings, we interact with a given set of people regularly and trust them over strangers. Each of these persons has their own circles of trust and familiarity, and in this way, everyone is connected to everyone else in a vast human network. This network has no definite structure. Nobody keeps a written list of who their friends are and who they trust, leave alone sharing that list. An individual’s view of the human network is limited to their social interactions.
But what if you could articulate your relationships? What if you could share this list with others and define a structure for the global human network? What could you potentially gain from such a well defined structure? This is the potential that social networking services hope to realize. Friendster made an early start as a dating site where you make new friends based on familiarity with your existing friends. LinkedIn set out to build the business equivalent: when you have several candidates for the job, the most reliable recruit may be the one your colleagues have previously worked with.
Orkut now wants to be the super-cool all-in-one network that works for everybody. Orkut not only maintains a list of your friends, it also provides several communication tools: a messaging system similar to email, but restricted to Orkut users; community forums, where several users can start new discussions on a common topic and reply to each other; photo hosting, for those who feel like letting others have a peek at their world; a scrapbook, which is like a guestbook where you can leave comments for a person that others can also view; and testimonials, where you sum up your impression of a person, and that person can then choose to display it to anyone viewing their profile. Orkut also allows you to declare yourself a fan of any of your friends, and rate them on a 1-3 scale of “trusty,” “cool,” and “sexy.” Users get to see who their fans are, and an aggregate of their ratings.
In short, Orkut goes out of the way to make itself an interesting site with plenty of activity for users. Users frequently enthuse about how cool the network is. And yet, somehow it doesn’t make sense. Orkut carries the aura of a dotcom business plan that promises great special effects but embarrassingly glosses over the fine details. Can you actually make an explicit list of all your friends? Can you articulate your level of trust in a person? Can you convince all your real life friends to join the site, so that your social network online matches your real social network? What if you build a fake network on Orkut? Declaring someone as a friend is really simple: clicking on the Add button on their profile. The other person also has to approve the request, but almost everyone accepts because they fear having to explain themselves if they refuse. Nothing stops you from declaring just about anyone as a friend, and indeed, some people took the liberty to do just that during the early days, forcing the site admins to restrain them—an event since branded as being sent to the Orkut Jail. Unfortunately, the admins also came down hard on some legitimate users who really had a lot of friends, and who now had a lot of friends to complain to.
People network around an activity. When you need a web designer, you remember the person you worked with earlier, or the person that someone you trust recommended. Social networking services encourage the reverse approach: first you declare each other as a friend, then you figure out what to do with each other. This doesn’t make sense. Given that technology has never managed a good social interface, it is far more likely that people will form relationships around the use of a technology, rather than a technology defining how human relationships are supposed to work.
The real online social networking, then, will come from collaborative repositories, where several people work together to build a common pool of data, code, or whatever else. Social networks built around an activity are sustainable, as long as you have an engaging interest in that activity. Once you accept this, you will realize that the Internet has always been about social networks—they just were never clearly defined. Perhaps the clearest definition comes from weblogs.
My favourite example is LiveJournal.com, which provides a free weblogging service. You can sign up for your own journal (same thing as a weblog) and make a list of all the other LiveJournal users you like reading. Their latest journal entries are then collected into a single Friends page, so you don’t have to visit each journal individually to check for updates. LiveJournal also lets you set viewing rights on your journal entries. Everyone can see it (public), only you can see it (private), or only your friends can see it (protected). Notice the catch here: your “friends” on LiveJournal are people whose journals you like to read. They are not necessarily your friends. Yet, this apparent inconsistency has an interesting effect: because there is a trade-off to declaring someone as a friend, LiveJournal users are encouraged to be careful with it. Contrast this with the social networking sites where you have people competing to have the largest friends list possible. Your friends list on LiveJournal is certainly not representative of your social network, but insofar as maintaining a journal is an activity you regularly engage in, LiveJournal provides a reasonably accurate representation of your network resulting from that activity.
In effect, LiveJournal emphasizes activity over network, but in the process builds you a more accurate network.
LiveJournal also has fake accounts, but since there is no critical need for authenticity of accounts here, they are tolerated unless the accounts are used to abuse the faked identities. People have used LiveJournal to create journals for, among other things, personal use, political discourse, display of art, web comics, and entertainment using fake profiles, such as the journals of Raavana, North Korean dictator Kim Jong Il, and NASA’s Mars exploration vehicles.
What Happens When You are Bored?
Modern society is a consumer of passive entertainment. We would rather sit in front of a television than engage in an activity, because TV requires no work. You do not have to participate to be entertained. LiveJournal is like that. When you want to be entertained, you go reload your Friends page, look at what’s new—and even better than television—you get to jump in and participate in the discussion when something stimulates you. Orkut is an interesting place until your enthusiasm runs out, and then there’s no place to go for passive entertainment. No page you can reload to get more.
We like our television programs to end because we can then change the channel or switch off. Even passive entertainment needs boundaries where you can stop consuming. On LiveJournal this boundary is when there are no more new updates from friends. On Orkut, where the primary activity is visiting other people’s profiles, there is no end point. This lack of clear end point can discourage one from starting at all.
Orkut probably recognizes this since they are now busy with an online magazine, complete with an editor and staff writers. This is unlikely to work though, because Orkut seems to be trying really hard to be interesting, to provide everything their users want, instead of letting their users do it for them.
All-in-One or Mix and Match?
The all-in-one service provider scenario brings us to another interesting dilemma, one that has repeatedly played out on the Web. Should you use an integrated provider, one that provides a consistent interface and carries your social network across all the services, or should you use specialized providers and rebuild from scratch everywhere? With weblogs, for example, Blogspot and LiveJournal provide the weblogging software, the hosting, and the community. Software like Movable Type requires your own hosting and has no community (owing to the lack of central identity verification). Observation of prior systems shows us that the integrated services take off faster, but limit users in the long term. Mix-and-match configurations take longer to evolve, but provide greater flexibility. Social networking sites could similarly find themselves torn up by more focused services, especially since they already do such an unreliable job of their primary purpose, defining social relationships. The strongest attack is coming from weblogging systems:
We have looked at LiveJournal, which provides a social network definition system that is more meaningful than the social networking sites, even if restricted to only one activity. Flickr does the same with photographs.
Trackback is a system for distributed commentary across weblogs (non-LiveJournal). In the absence of central authentication, the only way to verify that the person who posted the comment is who they claim to be, is by having that comment hosted where only that person has access: their own weblog. Trackback then forms the link between original post and reply post.
Going a step further, the TypeKey authentication system provides authentication for posting comments on weblogs anywhere on the Web, so your community is no longer contained within a single service provider (like LiveJournal or Orkut or any other).
Technorati.com is a service that tracks links between weblogs and ranks by popularity. Blogdex is a similar service that looks at the links and determines what the weblogging world is currently gaga about.
FoaF, (Friend of a Friend) is a protocol for listing your friends in an XML file that can be posted on your website. FoaF Explorer is a site that lets you use this FoaF data to explore relationships, much like you browse on Orkut or Ryze. FoaF suffers from the same problems as the social networking services with articulating what would be otherwise vague relationships, but it is worth watching as a non-centralized alternative for the same functionality.
Some older social networking systems that evolved for special purposes:
One of the oldest is PGP’s (Pretty Good Privacy) Web of Trust model for verifying that a person’s electronic encryption key (for sending encrypted email) really belongs to that person. PGP allows users to sign each others’ keys as a mark of verification. Signing only implies that you have verified—it does not imply you trust the person. When sending encrypted mail, you can then look at the attached verification keys, and check those keys in turn for something leading back to you. The shorter the chain, or the more the number of verifying keys, the greater the assurance that the key really belongs to the person you think it belongs to. This trust model has also been adopted by ratings websites like epinions.com and mouthshut.com.
Others like Wikipedia.org, the free encyclopedia, eschew complicated trust models and put their faith in community goodwill. Wikipedia allows anyone to edit any page on the site, even without authentication. A complete editing history of each page is saved, and it is expected that any errors or acts of vandalism will be corrected by others visiting the page. The model works remarkably well, so much so that publishers of traditional encyclopedias have been forced to sit up and take notice.
Wikipedia’s free-for-all model does not mean they have no community. They do have a forum where changes and proposals for new pages are debated, and an active (but unarticulated) social network of their own. Their example reiterates what has been said earlier:
The Internet has always been about social networks. Networks formed around activities are sustainable. Networks that exist solely for the coolness factor will have no future once the next cool thing arrives. Networks that try to fight this by providing too activities will find themselves struggling against specialized service providers. Don’t let the hype bowl you over.
These pages are worth reading for their insights into the reasoning for and the process of networking, the possibilities with a well defined network, and the problems with current social networking services.
Update, December 2006
It appears Orkut’s users have found a sustainable activity on the site: scrapping. It’s so popular that the site now allows responding to scraps without visiting the other person’s own scrap page, offers Google Talk integration, and even has a third party scrap messenger, Scrapboy. Why is scrapping so popular? I suppose because it makes conversation public, much the same way as blogging, while also maintaining a shroud of privacy: you only see one half of the conversation, unless you’re willing to take the time to visit each person’s scrapbook and find the other half.
This however does not explain why Orkut is so popular especially with Indians (and Brazilians — the two apparently make up 70% of the network). Perhaps a different explanation is in order. Given the nature of internet penetration and available bandwidth in India, Orkut may represent the highest return on a web presence for the lowest effort invested. Does that make sense?
1 This example is from Friendster, but I didn’t want to shift discussion from Orkut to Friendster, so have not identified the network.