FriendFeed's meteoric rise Video

FriendFeed's meteoric rise Video Transcript

[ Music ] ^M00:00:03

>> I'm Dan Farber from CNET NEWS, and I'm joined by Bret Taylor from FriendFeed. Bret thanks for joining me.

>> My pleasure.

>> FriendFeed is taking off like wildfire. It along with Twitter seem to be all the rage now in these social networking circles. So how, first how is FriendFeed different from Twitter?

>> That's a good question. Well, really [inaudible], FriendFeed lets you see the content that your friends are discovering around the web. So when you log into FriendFeed, you'll see your friend John, dug this article on Digg, or they bookmarked this funny video on delicious. And so it's really focused on content discovery, and what people do on FriendFeed is discuss that content. So right now on FriendFeed, there's tons of discussions about last nights primary results. And there's also some discussions about probably one of Fred Wilson's blog post, because I'm subscribed to lots of techie [assumed spelling] [inaudible].

>> So you're a social network as well as being a place to aggregate [assumed spelling] all those different feeds...

>> Sure.

>> From the people you follow?

>> That's correct, but it is very focused on content. And I would say most people use Twitter for communication purely. So it's sort of almost like group instant messaging in a lot of ways. And I think a lot of people use it that way. And as a consequence, the stuff on Twitter are lots of messages. And the stuff on FriendFeed is a lot of content, and just the discussion is really focused on that content.

>> Now, some people, and I would point to Steve Gilmore [phonetic]...

>> Yeah.

>> As an example, are requesting that FriendFeed clone the Twitter functionality?

>>Yeah.

>> Is that something you'd really want to do?

>> You know, I think that there's a lot we could do to improve the communication mechanism [assumed spelling] on FriendFeed, but we're certainly not interesting in cloning any other product. I think in general when companies try to clone other companies products, [inaudible] usually lacks the user base in a lot of the sort of, the soul of the product that got people to use in the first place, so we're really trying to create something unique. Really focused on solving the content discovery problems, sifting through the huge amounts of information being published on the Internet. And I think that [inaudible] is distinct from a communication control. There's enough overlap that a lot of, and Twitters had some problems recently scaling to their user base. As a consequence a lot of people would like us to go in that direction. And I think that we will improve our communication tools, but we're certainly not interested in cloning their product all together.

>> I'm a Friendfeed user and I find it useful in terms of aggregating the content, and discovering things. But I also find it really difficult in many ways to get through it...

>> Yeah.

>> It's just like an overflowing river. What are you doing to solve that problem?

>> So there's a couple of things. Just last week, we launched a feature called Rooms. Rooms are basically places where multiple people can publish and share information. So you can think it almost as a mailing list, but what it would be on Friendfeed. And so that really helps you partition some of your sharing, and shared information into logical groups. So for example there's a conference in New York, I think this week or next week called Graphing Social Patterns used. And Dean Mclear [phonetic] is organizing the conference. Started a room on Friendfeed where people are posting articles related to that conference, and news about that conference.

>> Yes, he invited me to join that room.

>> Yes.

>> And so anyway there's, and there's also a room about Obama, and a room about, you know, racing, and you know, a huge spectrum of things. That said, it doesn't really solve the problem with just among your friends, the aggregate of not [inaudible] information is overwhelming. Over the next couple of weeks, we're launching a couple of features that will let you summarize the best-shared items from the past day, week, or month. And we're also working on a number of other problems, that really you can think of it as computers helping you synthesize [assumed spelling] the information. A lot of them are pretty experimental, so I'm not sure which ones will pan out yet.

>> Give me an example of some of those more experimental ideals.

>> Well, I can actually probably more [inaudible]; I can give you an example of when I was talking about the finding the best information. What, we have a lot of signals now on Friendfeed what people you know find interesting. So there's lots of comments and discussion on FriendFeed, which is obviously indicates an interest in the item that they're discussing. We also have this feature called Like, which basically is just sort of; you just click this like link...

>> [Inaudible] popularity contest.

>> Yeah and, but, which is also very passive demonstrated interest. And so when you're subscribed to say a 100 people, it could be overwhelming. But that means we have a 100 peoples signals of what they find interesting. And so we've developed an algorithm that tries to find the best information. And there's a lot of what I think are kind of created things in there that make it useful. [Inaudible] one problem people have is they're subscribed to maybe ten of their family members, and then a couple tech industry bloggers. And the tech industry bloggers are disproportionately [assumed spelling] active. So you might miss out on your family's photos from vacation. Because, you know, Robert Scobell [phonetic] is publishing a lot of blog posts in that timeframe, and it pushes down your family's content. So we have parts of an algorithm that are focused on diversity of information, so that you can make sure you see from the hundred people you're subscribed to that no particular person drowns out the other people you're subscribed [inaudible].

>> Is there going to be a way for users to actually twit the algorithm themselves in terms of setting some parameters?

>> Not right now. But it maybe something we consider in the future.

>> Well, great.

>> Sure.

>> Thanks for speaking with me [inaudible].

>> My pleasure.

>> I've been speaking with Bret Taylor of Friendfeed. I'm Dan Farber for CNET NEWS.

>>Thanks for watching. ^M00:05:23 [ Music ]