I’m your friend, feed me! FriendFeed is a website that exploits the idea of the newsfeed — the feature that helped popularize Facebook — to give you a stream of information about what your friends are doing on about 30 social networking sites. It will tell you if one of your friends posted new pictures on flickr, favorited a video on YouTube, dug an article on Digg, or watched a movie on Netflix—and serve as a forum to discuss the content. Yet another social networking venture? Well, consider that the four founders happen to be some of the brains behind Google Maps and Gmail. Here co-founder Bret Taylor discusses why he left the Google Mother Ship, how FriendFeed is letting users shape the product, and why the best Internet filters are your friends.
Did we really need another social networking site? Or are we just descending deeper into the echo chamber?
It’s a really good question. We come at it from the approach of solving an information problem, as opposed to creating a social network and solving people’s social problems. At its core, the way we talk about FriendFeed is as a content discovery tool, not just a social network. FriendFeed lets you share things on the Web with people you know. What’s unique about FriendFeed is that we can automatically import all the things you shared on a huge number of websites around the Internet. For example, if you favorite a video on YouTube, people automatically pick that up. They can play it online and comment on how you like way too many funny cat videos on YouTube.
What does FriendFeed suggest about the way we live now?
Over the past five years or so, one of the most interesting trends is the proliferation of user-generated content. When you talk with people, particularly younger people, about what they read or watched today, it’s not the NBC Thursday night lineup; it’s 10 YouTube videos published by people who are not professional video producers. It’s not just The New York Times article about the election, it’s five blog posts analyzing that article. The problem is, despite the fact it’s very easy to produce this information, the mechanisms to find what to read or what to watch haven’t really scaled. Having a list of the 10 most popular videos today isn’t that useful when there are 10 million videos published that day, because you’re really not reaching the long tail of that content and not personalizing it to your tastes. Our thesis is, the people you know are the best filters for that information.
Are discussions more civil than what you find elsewhere on the Internet?
When you’re talking to someone who you actually know, you have a normal conversation. If you’ve even seen the real comments on YouTube, you know they tend to devolve to the lowest common denominator. When people are talking in the anonymous masses, they tend to attract the type of people who like to call up a radio show and hear their own voices.
You were one of the creators of Google Maps. Why did you leave Google?
When I started at Google, there were 300 or 400 full-time employees, and when I left it was 15,000, or something like that. It had changed a lot as a company and in the back of my head I always had this idea to go off and try to forge my own path just because I like the environment and the excitement of doing a start-up. My job was more managerial than down in the trenches. It’s a fantastic company and I probably learned everything I know about computers there, but it was just time for me to try to do it myself.
Is it true FriendFeed started as an accidental side project?
One of my good friends, Jim Norris, was a member of the team that launched Google maps. We had actually been at Stanford together during the dotcom boom and we’d always had the inclination to try to start our own company. Being at Google for a while gave us a little more financial freedom to start our own thing. We left in June of last year and become entrepreneurs in residence at Benchmark Capital. We were actually working on some storage technologies originally. We were growing increasingly uncomfortable with the business model options of the technology we were working on. On the side, I started writing some scripts to download all the activity my friends were doing on the website, leveraging the existing syndication formats. We ended up getting more interested in this idea that our original idea and decided to run with it. In the process, we had been sending some of our prototypes for feedback to two other Google engineers, Paul Buchheit and Sanjeev Singh, the two original members of the Gmail team, who had also left Google, and they were really interested in the idea as well. We ended up deciding to found the company together.
How much have you raised?
We’ve raised $5 million from two of the co-founders, Paul Buchheit and Sanjeev Singh, as well as Benchmark Capital.
Not from you or Jim?
No. Paul was actually employee number 20-something at Google and Sanjeev was very early as well. So they’re financially in a slightly different group than Jim and I.
With all that money, are you going to grow fast?
We plan on growing our engineering group — which is the whole company, I guess — so we’re hiring a handful of engineers. But we’re not necessarily like a lot of other startups who instantly want to grow the company to 100 people and hire a bunch of executives. With a product like FriendFeed, the challenge is it’s a new product category. I think we’ll have to do a number of iterations before we find something that’s useful for the technology community in Silicon Valley and also to a more mainstream audience.
You’re an example of perpetual beta.
We release features every single day. We are a company designed around iterative improvement. Natural word of mouth marketing only comes if you’re constantly improving your product, addressing people’s concerns, and getting real usage and feedback. That perpetual improvement of the product is the right way to develop Web software. The products that continue to innovate and evolve are the products that people will continue to use the most.
One blogger observed that you guys were so enmeshed in conversation with users that he could tell when you took a staff retreat.
We always joke that creating a product like FriendFeed is really bad for our productivity because every time we open our browser we get distracted for two hours finding interesting stuff to read. We use FriendFeed as a communication tool with our users. I subscribe to 200 people and end up interacting with a lot of people. A lot of the early adopters of the product are very passionate about the product and tend to have strong opinions about the path that we should take.
How and when will you start making money?
We’re not currently monetizing the site. Probably the first thing we’ll try is some form of advertising. For example, you can connect your Netflix queue to FriendFeed so that your friends and family can see you’ve rented a movie. American Gangster shows up in my FriendFeed, there’s a lot of opportunity for related ads. That American Gangster movie is tied to me, almost as an implicit endorsement of that movie, which is a really interesting context that we think has a lot of potential for advertisers.