September 17, 2008

Bret Taylor on FriendFeed's Road to Monetization, Early Surprises

This evening, at a panel on lifestreaming put on by the MIT/Stanford Venture Lab, FriendFeed co-founder Bret Taylor spoke about the popular aggregation and lifestreaming service's early months, explained what he and the team are trying to do through developing the site, and what we can expect from FriendFeed, as the company builds plans to monetize and further expand its growing user base. The panel, moderated by All Things Digital's Kara Swisher, also saw participation by angel investor Jeff Clavier of SoftTech VC, Leah Culver of Pownce, and Loic LeMeur of Seesmic.

I attended the session and took notes via laptop, so all quotes are "best effort."

Bret's presentation stated that FriendFeed, which currently supports 43 different Web services, and is now tracking greater than 100 million individual entries, is designed primarily to enable content discovery and social media consumption through a shared experience with friends and peers.

Growing Crazy Fast After a Slow Start

In front of an audience at the Stanford School of Business, Bret, in a quick presentation utilizing Apple's Keynote, recounted the team's reaction to the site's initial traffic spike following launch coverage in the New York Times and TechCrunch which evaporated in mere days. He said it took four months to return to the initial activity level, in between which the team went through varying stages of excitement, strategizing, realism and depression, while they openly questioned what they might have been doing wrong - having a history of successful product launches at Google. However, not too long later, traffic began to balloon in the beginning of 2008, reaching a hockey stick spike from March to June, during which the team's excitement turned to sheer panic, as they looked to scale their product and maintain speed and reliability amidst unprecedented demand.

Initial Development Missteps or Oversights

When FriendFeed added the ability to "like" items and make comments to a friend's feed, it opened up the opportunity for significant discussion among peers, and helped catapult the site from a simple aggregator to a destination site for many. But Bret admitted the team didn't anticipate the success these additions would have, and they didn't put as much development work into fleshing out those features.

As he said yesterday evening, "The discussion parts of our site have been almost the sole driver of our growth. It's been interesting to watch, and in retrospect, it was obvious. It was initially one of the underdeveloped parts of our site."

Also, while FriendFeed has been lauded for their highly-capable "hide" features, letting you block individual posts, posters, or services, there have been many requests for better ways to filter through the noise and find information that's most suited to your own likes and dislikes. But so far, the team is still playing catch-up. Bret added, "For the one year or so we have existed, we put less into relevancy and more into filtering tools. We are working on relevancy now. It's reflected in the different ways that people use a feed reader, as some see it as a new e-mail box and others ask to show the things that are interesting right now."

Handling Competition From All Sides

Swisher asked, as many do, if there are too many sites in the lifestreaming space, as there were too many calendaring sites in the Web 1.0 timeframe. She speculated that by the time the Web 2.0 shakeout occurs, that maybe three will survive and one, like Google with horizontal search, will end up with the lion's share.

Bret said that the lifestreaming space is a new category, and that it was "healthy" for many people to be working on "the content discovery problem". With the advent of syndication formats like RSS and social networking, he said subtle differences would be very important, and that on the Web, there is a history of natural fragmentation that enables niche sites to succeed. However, he did warn against sites adding so many features that they miss their core position. He said, "Every application grows until it has e-mail. Every Web site grows until it has all the features of a social network."

On Staying Independent

Every few weeks, somebody seemingly speculates that FriendFeed would make a great acquisition target for somebody, and the name that almost always comes up is Google, where the team's co-founders were last employed. But, as he and Paul Buchheit have consistently said, Bret again repeated the plan is not to sell the company, even if the road to business success isn't 100 percent clear.

"We're not interested in selling. We wanted to forge our own culture, to create a sustainable company," Bret said. "We have different perspectives on how to build a company of scale, and we want to build a company that scales."

Finding an Eventual, Sustainable, Business Model

Bret said the $5 million in seed funding FriendFeed raised this Spring was done anticipating the economic downturn, enabling the company to have a long runway before seeing the cash disappear. The team's hope, he said, was to find an advertising-based solution that delivers revenue without damaging the user experience. As he said, in response to questions from Kara Swisher, "there is a huge spectrum in the effectiveness of advertising," from ads that have high click-through rates, like those from Google AdWords, to the less-targeted AdSense, which delivers low conversion rates and "lots of accidental clicks." For FriendFeed, he speculated the site was "somewhere in the middle, but hopefully on the good end," where if links were mixed into the service that were sponsored, they would be done in line more with users' experience than image ads adorning the sides of the browser.

While Swisher coyly teased some of the panelists about their being "pre-revenue", Bret said one of the keys to launching a successful business model in the Web 2.0 atmosphere would be to not do so too early, and when they do, to do so in a way that is both quantifiable and analytical. "It makes no sense to try and monetize when you have only 2,000 users," Bret said. "It's too early and the early adopter audience does not reflect the behavior of mainstream users." He cited the early successes of Overture and Google AdWords as forging the quantifiable advertising market, but admitted they weren't yet sure how ads on FriendFeed would work. "We want to experiment enough to not run out of money before having to raise more, or we will have a sustainable business," he said.

What's Coming Next

Bret clearly hinted at a move to improve relevancy, and help users find signal in the noise. The team added "top" posts of the day not too long ago to help use the wisdom of crowds, and that feature could be improved. It's clear ads are coming, but maybe not immediately, as the company continues to try and scale in terms of users before worrying much about revenue. He also gave praise to the search engine that returns results just from your friends, but said it could be improved, as missing a result in FriendFeed would be much more impactful than a missing search result in Google. But he also spoke of "focus" and doing "one thing well", so it could be that those of us asking for messaging features aren't going to get our wish.

One thing you can expect FriendFeed not to do is to immediately give in to the demands from the early adopter tech geek set, who can at times be very demanding. While Seesmic CEO Loic LeMeur said the "tech geeks and geek press would have you make products for the geeks," Kara Swisher helpfully added that group was pretty small to begin with. "It's 14 slightly-overweight white guys," she offered.