FriendFeed is a current Web 2.0 darling. The service performs the increasingly valuable job of presenting, in one place, all the online activity of the friends you want to follow. Twitter posts, blog entries, YouTube favorites, Last.fm listens, Flickr photos, you name it...FriendFeed lets you track it all (except Facebook updates). You can also talk about your friends' activities on FriendFeed itself, a clubbier environment than joining the fray on, say, a YouTube feedback page.
The service is not the only social aggregator, nor is it the first: Plaxo Pulse does a lot of the same stuff, as do smaller operations like Profilactic and Iminta. But FriendFeed is well-funded, and since launch the team has taken to rolling out new features quickly. So what's next? I visited the FriendFeed offices in Mountain View, Calif., to talk with co-founders Paul Buchheit and Bret Taylor.
FriendFeed is currently a "social-network aggregator." It picks up the stuff you do online and tells your friends about it, saving them the hassle of visiting all your online hangouts to see what you are up to. But as many people have noticed, this leads to social overload. It's too much information to process. Buchheit and Taylor were clear with me that they have more work to do on FriendFeed to make the core aggregation feature more useful. In particular, they want to add intelligence to the service so it highlights what you're interested in, not every last thing your friends are doing.
One of the use cases that they have not solved for, Buchheit told me, is the infrequent user. Say you visit FriendFeed after an absence of a few weeks. Some of the "old" information will still be relevant to you--week-old photos of your nephew's birthday party, for example. But other, newer news will be irrelevant due to its age, like a three-day-old blog posting about a stock you're following. FriendFeed needs to a better job of differentiating content so old-but-relevant news can surface ahead of data that might be newer, but is less valuable.
"Aggregation is a mechanism, not a product," Bucheitt told me. The product was built so that, "You don't subscribe to sources, you subscribe to people." The founders told me, again, that their vision is only partly realized in FriendFeed today.
The founders want FriendFeed functionality to surface in multiple places. One could argue that without a broad footprint on the sites that FriendFeed aggregates from, it will not likely get broad traction. The recent release of the FriendFeed API is a step in that direction. Also, an "add to FriendFeed" button will be released within a few days. Publishers or users of various social sites will be able to install this on their feeds, pushing the FriendFeed message to their followers.
FriendFeed's most popular request is for Facebook integration. While there is already a FriendFeed Facebook app, ironically FriendFeed cannot read Facebook data itself. Users want FriendFeed to follow Facebook photo updates, and the team is working to implement that, but Facebook doesn't make it easy, I was told.
Despite requests, the team is not working on a desktop app. But other developers are, as has been reported in TechCrunch. Buchheit and Taylor expect the first AIR-based FriendFeed desktop app to surface within days.
Another thing not being developed: a mobile version of FriendFeed. The founders note that the current site works acceptably well on an iPhone, and that other developers are working on a mobile app using the new APIs.
The team continues to work on the user interface, though, trying to balance readability with compactness. "Spacing is undervalued," Taylor told me, but he's still sensitive that using the site requires too much scrolling on the page.
FriendFeed recently launched the "Twitter back" feature that allows users to reply to Twitter messages from within FriendFeed. Similar functionality will be added to other services that FriendFeed handles. Next up will be 'Seesmic back." Other services will get "back" features as their APIs allow. "We are really open and eager to do these things," Taylor told me.
FriendFeed is currently not bringing in revenues. While the founders have not settled on a business model yet, Taylor did notice that the FriendFeed philosophy--it's a site for finding things to read more on and has "very little actual socializing"--is consistent with advertising that works. Since users can include feeds that can be monetized (Netflix movies queued up; Last.FM music tracks listened to), it's possible that the site could make affiliate revenues when users follow up on friends' content recommendations. Standard in-line advertising (like Pownce has) is another possibility.
Meanwhile, FriendFeed's burn rate is low. There are only six employees, and the company is basically stealing office space with a 50-cent-a-square-foot rental in a Mountain View building that until recently was scheduled for demolition. The housing crunch has given the building owner a new perspective, however, and FriendFeed will have to move soon, since the landlord recently scuttled his development efforts and plans to raise rents closer to market level.