What is wrong with the social graph?

The web turns into a social space. Every new web app tries to grab a piece of the social media pie. To be viral is a new disease that spreads all over the web. From my point of view, this is a very unhealthy evolution. I blame the base of any social network, the social graph, which does not represent the reality well enough.

How many friends do you have? Not on Facebook, Twitter or FriendFeed, but in the real world? Have you ever met them face to face? Probably yes, right? What type of content do you usually share with them? Do you show them pictures from your vacation in Hawaii? Perhaps you talked about your frustration from buying high-priced live concert tickets on the secondary market. Maybe you recommend your favorite restaurant to some of your friends from time to time.

OK, now log in at www.facebook.com. Or use MySpace, Twitter, FriendFeed, whatever. One of your so called “friends” shares a new picture of her little puppy. Simply cute overload, as some of you would say, but what the hell does it have in common with me? One more example that I “really” enjoy seeing is status updates. I love “Watching very funny video at http://tinyurl.com/snipr.by. Join me now!” You read the status update first on Twitter. Then it automatically show at Facebook and then, last but least, FriendFeed follows. God bless data portability in this form!

These two above mentioned scenarios illustrate a huge gap between handling the real world and treating the virtual space. To harmonize our offline and online presence, we need to change our attitude towards one of these two worlds. Most likely it won’t concern the real one 🙂

Social media is a new pill that we take each morning and evening. True, many people use it continuously 10+ hours a day. In the current state of social network the pill does not make our lives any better. What I can see is mostly a lot of noise, redundancy and irrelevancy. How to get rid of this? Adding some kind of semantic information to social networks could be a solution, at least in part. For instance I like the Link – Blogroll admin section of WordPress because it uses XFN specification for describing the social graph. The system asks me what the relationship is like when I add a new link.

It is definitely the way to go as more details about a certain relationship is given the higher content relevancy can be reached within the social network. Furthermore this is the direction we should take if social networks are supposed to represent the reality.

Another problem is caused by both people’s uncontrolled creativity and an easy access to digital technologies that allows us to share almost anything. Simply there is too much content on the web and especially inside social networks. Users need tools to cope with this information overload. Take Facebook and its redesigned homepage with the News Feed. I can pick my friends and add them to friend lists. By choosing a given friend list I can filter the lifestream. Moreover I can hide inappropriate friends or applications with a single click. To a certain extent these features provide a good help to me but not the perfect one.

So far the most efficient filtering system I have found at www.friedfeed.com. Each item in the FriendFeed stream is accompanied by a Hide link similar to the one at Facebook. But there is a difference. Except the basic hide functionality for hiding the unwanted or irrelevant content, the user can set up advanced Hide options.

I am really a big fan of this Hide feature. Let’s say you want to display only blog posts (or tweets) from TechCrunch.com that have at least one comment (or so called “Likes”) from any FriendFeed user. No problem here. Moreover you can set up multiple filtering criteria. This way you can clean up the displayed life stream so you see only the relevant content. One more thing I like at www.friendfeed.com is the stream item clustering. That means they put similar or equal activity items together (tweet and Facebook status update showing the identical message for example) into a single (though expandable) activity line. Ultimate redundancy killer!

The entire world of social networking is still very immature. We should be slowly heading away from Snowball effect games, Six degrees of separation experiments and “popular” quizzes to something more useful. Current social networks are more or less empty containers where selling ads are the last instance the providers rely on. It needs to be fixed.

  • Great post Jan and very true.

    In the context of CoWorking (…) I came across this post: http://tinyurl.com/d8bmsf coining the concept of 'diagonal relationships'

    While we allow for a wildgrow of horizontal (therefore not indepth) 'relationships' we may start neglecting our real (vertical) relationships.

    We somehow need to find a balance between the two -> what we may start calling diagonal relationships.

    Time will tell if we are able to adapt and lets hope we'll not feel amputated when our 'friends' turn out not to be….

  • Thanks Dave,

    the diagonal relationships sound very reasonable and corresponds to my call for a wider implementation of the rich data model (XFN for example) of social relationships.

    Then we could see a new breed of social applications that could add more (or at least some) value to our lives.

  • Funny…for me the usage of friendfeed at all is entirely redundant…and so I don't. 😛 I'm kinda joking (except that I truly don't bother with it), but the reality is that social media aggregation itself needs to be finessed quite a bit in order to not become a total timesuck that people just avoid entirely.

    There was an interesting discussion as part of a roundtable conversation regarding privacy with someone from Facebook at SXSWi this year. Basically upon discussing granularity options ad nauseam, the clear message from sites with a more prolific and diverse user base was that although users claim to want granular privacy options, they rarely use them because they tend to be prohibitively complex for people who aren't powerusers of a given social network.

    From a usability perspective, providing users this level of control up front can be intimidating to the point where people leave the site instead of completing the task at hand. Perhaps a good balance is to provide it to those who want it hidden behind a click + ajax or something like that (but inevitably participation at that point would dwindle to the point of questionable resource investment benefit).

    In all, this is quite an interesting post that brings up a lot of key points about information that relevant to the social graph, and yet generally absent from it. I just wonder if there's a more implicit way of collecting said information so that the learning curve doesn't prevent critical mass from ever occurring with services who attempt to tap into it.

  • Great comment Lena, I totally agree.

    One of the key points you mention is a gap between a poweruser and a mass user. The web app, be it a social network, aggregation service or something else, should be designed for both user groups. Usually I would expect a basic feature set for masses and a power-feature set for advanced users.

    Generally speaking, if a given web app offers only over-featured version then it will hardly reach the critical mass.