Yup. Facebook friends trend towards bloat, which is problematic. Twitter followers/followees can be somewhat more manageable because there's a different set of assumptions about what following means vs. "friending".
The difficulty is that this is a very deep problem, and it is "solved" only partially by a very complex, very subtle, very tricky set of societal and personal customs and behaviors. The biggest problem from a social networking service standpoint is that computers very much prefer clear, explicit data, whereas human relationships exist in varying degrees of fuzziness. Most people don't sit down and write out a complete list of everyone who they consider a friend, nor do they create a tiered set of levels of friendships or a ranked list, nor would that necessarily be helpful as such levels are often in flux as friendships change, evolve, grow, and dissolve.
This is a problem everywhere with human behavior, it's one of the reasons why google was successful, for example. Back in the day every search engine was very practical and straightforward. Searching a set of terms brought you a list of web pages that had those terms on them, ranked by whatever convenient concrete metric seemed reasonable. To improve the quality of search results people would have to manually provide cues to search engines, via meta tags for example. Google decided instead to use a more sophisticated search strategy, extracting the implicit meta data about web pages via page rank and such-like. In short, google made web search conform to human behavior, rather than trying to get human behavior to conform to the requirements of search.
I suspect the same breakthrough is waiting to happen in social networking and will have just as big an impact.
Facebook could start by realising that I'm not interested in Farmville updates from people I have never interacted with beyond accepting their initial friend request, and am very interested in updates from the people whose events I attend and messages I reply to. They have a wealth of data on how I intentionally interact with other Facebook users.
But most of the manual features for managing friends are already in place - you can update particular lists of people through groups, events and private messages. It's really not that tricky to decide who you want to invite to that awesome house party or on the cheap trip to Vegas.
Other features, such as the "how do you know them?" and focus on work/school/city networks seem to have actually been removed due to lack of use despite offering very intuitive and automatic categorisation of friends. Facebook's trajectory towards openness probably reflects a lack of desire towards using it in a structured way. A hypothetical startup whose main strength is removing the spam in the newsfeed and adding in the features they chose to remove probably isn't going to keep Zuck awake at night. I preferred Facebook before Zynga and gratuitous "liking", but I'm not going to persuade my closest friends to move to a clone to get rid of the bloat.
FWIW I think Dave's example of friend bloat is probably more a reflection of his social circle than underlying changes in Facebook itself. He'd have had a lot more friends in 2006 if he'd been attending university, and considerably fewer now if he was working 9-5 for an unglamorous small business and nobody read his blog. The number of friends I had grew far faster in early 2006 than in any subsequent period, and people I shared fleeting conversation whilst waiting for lectures with were arguably far more tangentially connected to me than, say, my sister.
"Facebook could start by realising that I'm not interested in Farmville updates "
OTOH, should Facebook be "smart" about this, or should they just provide a button to hide all Farmville updates? In fact, they try to do the first, and also do the second, so I don't know why people still complain about this.
I think they should do both, in a better and more obvious way. Certainly, some level of filtering of the feed has alwaye existed, but 1997 Yahoo tried at search and 2004 Hotmail had rudimentary spam filters and reporting tools. Google admittedly didn't have to fight against strong network effects to grab market share there, but it was the ability to incrementally improve that made their product offering more compelling rather than anything staggeringly original. And Yahoo and Microsoft had a better starting position to make improvements had they realised the true value of the underlying data on their user behaviour...
To be fair to Facebook's team, the feed does seem to have improved recently and not just because I've manually blocked Zynga apps. But with their data and talent there really shouldn't be any suggestion that a competitor starting from the ground up could better facilitate the social interaction they started off specialising in.
That's an interesting analogy but flawed because getting traction for a better search engine does not depend on network effects nearly as much as it does for a better social network.
If none of my friends use Google, it's still as valuable to me. If none of my friends use Facebook,...
Sure, network effect is a big deal for sites like facebook, but it's not everything. Network effect didn't save myspace. If something much better than facebook came out people would start switching to it, and quite rapidly the network effect bonus would vanish. My point is that there is every reason to believe that the possibility there could be a social networking site significantly better than facebook is very high. Precisely because the problems in play are incredibly hard problems and nobody has solved them very well yet.
That's interesting. I like the reference to Google. In a way though, Google could rely more so on technical innovation to drive itself forward.
That's obviously not the case with Facebook, as seen from the controversy over their privacy settings. It's gonna be interesting watching technology drive culture at the crazy rate it is today. What a time to be alive!
The difficulty is that this is a very deep problem, and it is "solved" only partially by a very complex, very subtle, very tricky set of societal and personal customs and behaviors. The biggest problem from a social networking service standpoint is that computers very much prefer clear, explicit data, whereas human relationships exist in varying degrees of fuzziness. Most people don't sit down and write out a complete list of everyone who they consider a friend, nor do they create a tiered set of levels of friendships or a ranked list, nor would that necessarily be helpful as such levels are often in flux as friendships change, evolve, grow, and dissolve.
This is a problem everywhere with human behavior, it's one of the reasons why google was successful, for example. Back in the day every search engine was very practical and straightforward. Searching a set of terms brought you a list of web pages that had those terms on them, ranked by whatever convenient concrete metric seemed reasonable. To improve the quality of search results people would have to manually provide cues to search engines, via meta tags for example. Google decided instead to use a more sophisticated search strategy, extracting the implicit meta data about web pages via page rank and such-like. In short, google made web search conform to human behavior, rather than trying to get human behavior to conform to the requirements of search.
I suspect the same breakthrough is waiting to happen in social networking and will have just as big an impact.