Wednesday, January 2, 2008

The overload of online social networks: part III

It's gaining momentum, since my last two posts about this. I've just read a bit about current efforts to cope with this ever increasing overload of social-ish information production and consumption.

Over the last weeks I've been thinking a bit on a (possible) successful recipe to fully meet this goal. But first let us first examine the nature of social networks:


  • Professional: typically links persons based on professional connections, lists CVs, support job hiring, etc. Such examples are facebook or linkedin;

  • Friends: these one-size-fits-all social networks are centred on connecting friends in two perspectives: friends I know and care about (e.g., facebook), and friendships (worships?) I want to make (e.g., myspace);

  • Utility: these are a bit anti-social networks, where everyone shelfishly contribute to the whole. del.icio.us and digg come as prime examples of leveraging collaborative efforts (tagging and voting) to meet each individual's expectations (e.g., discovering new resources or having personalized news);

  • Topic-centric: these networks are built around a common interest. Bakespace (recipe sharing) and last.fm (music profiling) are well-known examples, amongst hundreds (thousands?) of others (dogster anyone?);



This categorization leverages a particular aspect of utmost importance (already discussed by everyone else): there is a tendency to register into different social networks based on their offerings - a kind of vertical market segmentation - which is usually broken by network effects of some behemoths such as facebook or myspace. Nonetheless, it is an understandable way of differentiation between competitors. However, one surely knows that these categories are far from being disconnected. Especially due to widgetization, cross-pollination has become a normal practice.

When thinking about aggregating social networks, one realizes that each person might have different identities/personas. A first question arises? Do I want to share my personas with everyone else? (Actually, I don't mind at all, as you can see from my homepage, with my dumb/cool personal aggregator.) But does someone wants to share their myspace alter-ego with everyone?

This question leads to another questions: what's the purpose of a social network aggregator? Is is to build *THE* social network of all social networks (kinda like a meta-network) that everyone can access, or is it just a humble tool that helps each individual managing their personas and frame each network accordingly (all in a single page)? First thought: privacy matters.

Hence, here is my recipe for a successful social network aggregator:


  • Identity: I don't want to have to register 1000 times, write the same stuff everywhere, put up the same picture of myself. One time is enough. OpenID helps a bit, as well as solutions built on top of it (such as sxip's solutions). It's Identity 2.0;

  • Privacy: Having a granular way to select my identity aspects that I want/need to share is critical. As well as to whom I show them. Do I want to share a photograph just with my closest friends (from what social network, btw?), or with everyone? FOAF helps describing social networks in terms of connections, which could be leverage to cope with these privacy aspects;

  • Read access: An aggregator has to access my information, mix it appropriately, and leverage it (new visualization methods anyone?). Only this way I won't have headaches by just thinking about seeing *all* the updates on *all* social networks;

  • Write access: Do I want to post something on twitter as well as on facebook's wall? Do I want to update my status accross different social networks? Do I want to publish a picture on a subset of my profiles? An aggregator has to cope with these things. They are typically done by hand, albeit a set of existing APIs leverage these tasks. Google's OpenSocial comes to rescue, fortunately;

  • Sustainable business model: I won't provide an answer to this item (actually I don't have one), but someone has to figure it out. And no, google ads doesn't count...



Finally, and coming back to a previous questions, what would be the visible side of an aggregator? A simple login page open to the public and everything else is closed to each individual (walled garden - privacy matters, don't forget), or picking up each individual's granular control of privacy and expose the Network to everyone? An open garden might benefit some (hyper-targetted advertising?), but disturb others...

And what about after transposing the login page, what would I be faced with? Some kind of billboard/mashup containing every information I want to access, as well as updating facilities to cope with the overload issue? Or should an aggregator go a bit further and become a first class citizen in the social network battlefield? That'd be intersting, from an end user's point of view: one time login, infinite procrastination :)

One last thought: where to start. Identity consolidation. After that, connection networks and semantics might be built upon it.