How to Better Manage FriendFeed for Relevance

FriendFeed is an amazing new service that is both a content aggregator as well as a social network. At its core you add feeds to the various places you post social media online (your blog, your RSS reader, your Flickr or Zooomr account, your Netflix queue, etc.) and then people can follow all of your media in one place.

The more interesting aspect to FriendFeed though are the likes (endorsements) and comments (conversations) that take place around the social media that you publish — valuable user created metadata. FriendFeed has quickly eclipsed blogs as one of the most active place for the best conversations on the web.

But the real power of FriendFeed is still coming. Steve Rubel blogged today about the potential for FriendFeed to empower search. Steve hits the nail on the head here:

“Social contextual search addresses Google’s Achilles Heel – superfluous content. Right now when users scour the web they can’t easily separate content they trust – i.e. what’s been created by their friends – from everything else. It all gets piled into pages of indiscernible blue links that all compete for attention. However, if you can just search just what your friends think and prioritize it over everything else, you have a very powerful recommendation engine.

As an early Friendfeed enthusiast I find myself increasingly turning to its terrific search engine when I need product and service information. You can give this a try yourself here. However, it works best when you have added a bunch of people whose opinions you trust. Advertisers will soon be tripping over themselves to make sure their ads show up at the precise moment when such searches are executed”

And who better to finally capitalize on social search than a couple of old Pros from Google (i.e. the Founders of FriendFeed).

Yahoo made it’s first run at social search when it bought Flickr unbeknown to most. Most people thought Yahoo bought Flickr as a photo sharing content property. Really Yahoo bought Flickr first and foremost as search. It was the search team not the photos team (which originally passed on Flickr) that bought Flickr. The earliest application for social search can be seen in Flickr’s patent for interestingness. A patent that I personally feel is a patent example of patent abuse. By the way if anyone has an update on the status of this patent application I’d love to get more information on it. Yahoo largely has not capitalized on social search however.

So the question becomes how ought FriendFeed improve itself to enhance search and discovery for relevance. One of the problems at FriendFeed is noise. Noise = content by the most active FriendFeeders (myself included) that dominate the site. For some this is not a problem but for others it is.

Think about discovery and search this way. On the one hand you have a blogger that you kind of like and follow but don’t know personally who publishes 100 bits of media a day and you subscribe because every so often one of those 100 bits is a gem — maybe you put up with 19 twits about Obama to get that one gem about photography for instance. On the other hand you have that girl who you really, really, really like who nobody follows and who puts up 3 bits of media content a day. Now you are vastly more interested in that special girl that you like’s social media than the windbag blogger — especially since one of her three bits every day is a self portrait on Flickr.

Now on FriendFeed today, that blogger is seen all over your stream. Because he/she is popular and has lots of fans and followers his/her content is bumped a lot. That special girl posts and nobody likes or comments and so her media quickly gets buried in the sea of the worldwide talk show (as Scoble calls it).

Because we can’t be on and monitor FriendFeed 24/7 (unless your name is Louis Gray or Robert Scoble) we miss important conversations. Friendfeed has started to address this problem by presenting “the best of” FriendFeed each day, week, month. The problem is that this feed still pulls in the windbag bloggers and not necessarily that very hot and very fine girl that you have that big crush on.

So what’s the solution? Allowing users to rank their contacts. So I might rank the girl that I like (in my case my wife mrsth) as a 100 score out of 100 possible. I might rank my brother who I kind of like as well as a 90. I might rank Robert Scoble who is a popular blogger and also a good personal friend as a 75. And I might rank a blog like Engadget, or Gizmodo, or TechCrunch, that I don’t always need to follow but enjoy reading from time to time as a 10.

Now what FriendFeed could do is offer me the “best of” FriendFeed for *me* personally. Everybody’s best of would be unique, reflecting not only the comments and likes that media gets, but also their subjective ratings. So if my wife posts a photo on Flickr and it gets 2 likes, this would be shown to me ahead of a post on Engadget with 30 comments and 10 likes.

Give people the ability to score their contacts and you better empower both their search and discovery of content based on their trusted social contacts. This could be the ever ellusive king of all search, social search, that people have been talking about for the past 5 years but that has never quite yet arrived.

You can find me on FriendFeed here.

You can follow a conversation on this post on FriendFeed here.