Are we really friends? The trouble with buddy lists in social applications.

The trouble with buddy lists is that we end up collecting friends like baseball cards. Because I’m on your buddy list, it could be because we’re really closely connected, it could be because we met once at a conference two years ago, or it could be because we both tacitly just want to make ourselves seem slightly more important or popular by having a large number of friends next to our name.

Letting users browse each other’s buddy lists is a great way to discover other people on the system you might want to add to your list (one valid use case). But in reality, public buddy lists are pretty terrible at what should be their primary purpose – as a way of giving any understanding of who someone really is or their true set of social relationships/connectedness.

Just by virtue of being public, buddy lists get distorted by social pressures for the following reasons:

1. Pressure to accept buddies. It feels impolite to turn down a buddy request even if you are not truly closely acquainted with somebody.

2. Popularity contest. There can be a bias to maintain a big list to impress others.

3. No granularity. Most social networks don’t have much way of indicating what sort of relationship you have with somebody. Facebook for instance is a little better at this, we met randomly, we are part of the same organization, we are dating etc. but for the most part the connections are binary and not granular. We’re either friends or not – but not, are we a close or a weak tie?

4. It’s socially awkward to take people off. People rarely take people off their lists as the more public your list is, the more awkward that might be.

And don’t forget Sampling bias. For many social networks, the breadth of your buddy list may say more about how much time you spend with other computer nerds, rather than saying anything useful about the quality and breadth of your actual connections to people in the “real” world (the 1st life).

By comparison, this is why subscribing to people by rss is great and why I’m in no rush to make my OPML public. Because it means I can read who I want to read as often as I want to read them, without my actual relationships with these people being distorted by social pressures. And if my interests or connections change over time I can quietly unsubscribe from people without it being a big deal or offending anyone.

So, how could social networking applications solve this buddylist problem? Here’s one idea, maybe not applicable for all networks but one that could work well in a say… an Enterprise 2.0 platform

Keep 2 lists.

Sure, keep an explicit buddy list that you control, but make it private.

And then let for the purpose of being able to broadcast an accurate portrayal of a user’s actual “aboutness”, let the system keep track of a user’s public buddy list implicitly.

Imagine a last.fm type system that reports back to you on the actual number of your interactions with people within the system and the extent to which those interactions are reciprocated.

The system could then easily spit back for the benefit of each individual user statistics on the “aboutness” of everyone in the organization (or at least on their current professional connectedness) like little last.fm-style bar charts of who you work with most closely that you could display on your profile page. Weighted by relevance and always up to date (much unlike any official org chart I’ve ever seen on a corporate intranet).

In fact, throw out the old org chart while you’re at it. What you’ve got now is a practical working map for for modeling the new enterprise.

Also implied by this idea is that if you’re going to appear connected to someone, you’ll actually have work at maintaining that relationship.

Are you ready for this new level of social transparency? You can count me in.