10 Key Lessons Facebook Learned from Friendster (and what I learned to apply to Bebo)
10 Key lessons Facebook learned from Friendster’s mistakes
I learned them too and applied them to Bebo (sold to AOL for $850M in 2008)
by Jim Scheinman, one of the first employees at Friendster & employee #1 at Bebo (running Marketing, Business Dev, Corp Dev & Sales)
Follow Jim on Twitter: @jimscheinman
1. Focus. It’s critical for a startup to have a clear focus, especially early on! Focus comes in many flavors. One key area of focus is product-focus. A start-up must choose what problem they are solving and stick to it until they’ve solved that problem, or it becomes clear that it’s time to pivot. It was very difficult in the early days of Friendster to choose one direction and stay the course. Many smart investors, executives and team members all had great ideas on where to take this new thing called social networking. As we learned at Friendster, being the pioneer can be an advantage when it comes to getting an early lead on the competition, but can also be a major challenge when it comes to focus [more on product focus below in the product section]. There’s also a customer or demographic focus that needs to considered. Friendster’s product was launched to everyone over 13 yrs old and struggled to focus the product on a specific demographic. As a striking contrast, Facebook’s focus on the college audience (and in particular, Harvard) helped them hone in their product to a particular audience and make it a great experience for Harvard students and later all college students and eventually everyone. At Bebo, we decided to focus on the high school and college market in six key English Speaking countries (US, Canada, UK, Ireland and New Zealand). We knew this demographic would be interesting to one another and would also be much easier to monetize once we had scale (see Revenue below). As facebook and Bebo proved, it was better to build the product for a single demographic at first, and expand later after we delighted our early customers rather than to go broad and try to narrowly tailor later.
2. Product (ship new product fast & often). A start-up must balance the need and desire to constantly innovate to stay interesting, fresh and ahead of the competition with the critical need to stay focused! While it's critical to introduce new great features in order to keep members engaged with the product, in the early days, it’s more important to keep your limited resources focused on the key product features and not to get carried away with features that might confuse the product direction or burden site performance (more on this in Technology lessons below). Friendster had several challenges here--first, as the pioneer in this new social network space, Friendster was the test-bed for everything. It's hard to be the pioneer! Others who came later, like facebook, were able to focus on those features that they knew were critical and avoid the ones that didn’t work. Friendster didn’t have the luxury early on to watch others make the mistakes. Moreover, there were several product features that Friendster focused on early on that in hindsight were not 'worth it.' The #1 requested feature from early members was an integrated IM experience, which Friendster made a priority. However, it was simply not worth the massive engineering effort required at that time. Like facebook showed later, it would have been wiser to focus on easier to execute features that drove virality at first and a feature like IM later after massive scale was achieved. Additionally, since Friendster had such huge site performance issues, it was virtually impossible to launch new features: the team was too busy trying to keep the site working and couldn't get resources for new features. Moreover, Friendster was concerned that any new successful feature would end up driving more traffic that would further deteriorate site performance. A terrible dilemma for any start-up! Like facebook, at Bebo, performance was of paramount importance and we started with the key features we knew (or hoped) would lead to critical mass and avoided any major new features until we knew the site was stable and performance would not be impaired. Moreover, when we came up with innovative features ideas, we constantly launched, iterated and tested new smaller and relatively easier to execute features monthly (and in some cases weekly) and kept those that worked and killed those that didn’t. Speed to execution is a critical success factor for a start-up.
3. Performance. The site must be fast! Friendster made some early ‘bet the company’ issues on the product side that proved to be very costly. For example, some in the company held fast to the idea that we needed to show how each member was connected to one another by six degrees. You can imagine how taxing on the Friendster database this was to figure out and display every time someone used the site, especially with 2003 technology. As Myspace, facebook & Bebo later proved, members didn’t really care about how they were connected to people six degrees away. One or two degrees are sufficient for most people. In another example of poor decisions leading to slow site performance, Friendster’s homepage showed so much content that there often required dozens of database calls to just load the homepage. This significantly slowed the rendering of that most important homepage, and in some cases it took several minutes to load. These painful and valuable lessons were easy to apply at Bebo—we never launched features that would sacrifice site performance. And Zuckerberg learned the same valuable lessons—as he says in the movie, The Social Network (paraphrasing): “we’ve got to keep the site working fast, we can’t end up like Friendster.”
4. Technology (& Good Luck). Don’t experiment with the newest technologies when you have a fast growing startup! Friendster was growing at an unprecedented rate and as such needed to push the envelope on the available technology to keep pace. Unfortunately, the team probably went too far out on the limb with the bleeding-edge and experimental technologies. In contrast, at their launch, facebook couldn’t afford to scale more than one college on each server, so they opted to work with the model of purchasing servers for each new launch. While this might not have been the most efficient method, it enabled facebook to scale and keep their site working fast while Friendster was constantly mired in scaling and speed issues. Every great company has had a tremendous amount of good fortune or luck along the way. Facebook was lucky early on that their strategy of scaling from university to university was something that the current technology was able to manage. Early on, facebook was able to simply buy new servers for each university avoiding many of the early scaling issues that Friendster faced which went globally from the start. At Bebo, we built everything with proven technologies (even Oracle database solutions) that we knew we could scale without sacrificing site speed. Bebo certainly had its moments of incredible good fortune, like the time we hit a wall with Oracle early on and the site crashed. We were able to locate the one expert based in London who could fix our problem but at first told us he was too busy. We were devastated, but he quickly reversed course and fixed our issue and later told us that it was because his daughter was using Bebo and loved it and was upset that her dad wasn’t helping us get the site working again. What an unbelievable turn of events!
5. Viral. If the goal is to grow fast without spending a fortune, one must focus on viral features! At Friendster, the concept of viral marketing was nascent and today we take for granted many of the insights and features that the early team at Friendster created (either by design or by accident!) in regards to virality. Friendster didn’t put a priority on viral marketing that is needed to succeed, we didn’t know any better at the time. Virality is a science and today many entrepreneurs are well versed in such terms as the ‘viral coefficient.’ Start-ups must focus on the details of viral marketing to make it work—it’s not easy, but when it works, it’s magic. Facebook focused on their site’s virality from the start, and always had large teams of folks focused on the user growth and viral paths (teams of over 40 people even from early on). There’s been a lot written about facebook’s focus on viral marketing recently in Quora and on blogs (see Andrew Chen’s “Growth Hacker”). It’s one of the key reasons for facebook’s success. Zuck was probably the first CEO to create a specific viral team—a brilliant idea—that kept a huge focus on viral growth at the company (to this day). It can’t be someone’s part-time job if you want to be successful at it. At Bebo, everyone on our small team focused on virality. There was constant A/B testing to improve the viral coefficient. I applied all my knowledge of viral growth that I learned from Friendster to Bebo. Bebo Founders, Michael & Xochi Birch had great experience in viral growth from their previous start-ups and we constantly experimented with viral features. Often some of Bebo’s most loved product features (like ‘share the love’) were our best viral hooks as well. That’s nirvana for a start-up -- when you can match your viral features with features that your customers love.
6. Customers. Know Thy Customer! Early on, Friendster was inundated with members from the Philippines. There’s nothing inherently wrong with that, except that Friendster’s strategy was to grow in the US first and later worldwide, primarily because we’d have an easier time monetizing US members via advertising. Moreover, when new members from the US would join the site, Filipinos would message to them in their native language (Tagolog), which was really confusing to the US members. In a viral product, like a social network, you need to be diligent on where the ‘virus’ spreads. In Friendster’s case, it spread early in South San Francisco, which has a very large Filipino population who then in turn spread the site to their family and friends in the Philippines. It turns out the Filipinos (like Brazilians—remember Google’s Orkut?) are extremely social and in Friendster they discovered their ‘dream product’—a free website to stay connected with friends and family in their country and abroad and to meet new people. Great for the millions of Filipinos who flooded Friendster, not so great for Friendster who was already struggling with trying to keep the site working for the millions of US customers growing increasingly irritated with the terribly slow site performance. In stark contrast, Facebook’s demographics were Harvard students, then later Ivy League students, then all colleges, and eventually everyone. In the early days, if you were a Harvard student, when you joined facebook, the people on the site looked like you and thus looked interesting. They also had some great college-focused features that made your life as a college student better—like checking out that cute guy or gal you saw in class that day. And, thus, facebook spread like wildfire to one key demographic early on setting the tone for the product. At Bebo, we made sure to focus on what we wanted to be our core demographic—high school and college students from the US, Canada, UK, Ireland, Australia and New Zealand. We came up with several strategies to ensure this success…one of which was simply shutting off the viral features outside of those countries.
7. Hiring. You must hire A+ players! There was a lot of pressure early on at Friendster to hire quickly, especially on the engineering team. Friendster tried to hire too many people too quickly. It’s simply not possible to hire all A+ team members as quickly as was being required at Friendster. Zuck surrounded himself with some talented and well-connected advisors early on who helped him in recruiting and hiring great talent. At Bebo, we hired slowly and only amazing talent. In fact, a team of only 6 of us built the site to reach 1 Billion page views. With today’s technology, you don’t necessarily need a lot of people to scale a successful site, just the right people.
8. Leadership. A start-up needs to have steady leadership, especially at the top! Great leadership is required for companies’ success at any level; however, start-ups are especially susceptible to problems with early changes at the top. It’s often the product vision of the founding CEO that is absolutely critical to lead the way and enable the company to stay the course. Friendster had 3 CEOs in 2 years. Mark Zuckerberg is still the CEO of facebook. Zuck’s clear vision on focusing on the product and performance over revenue in the early days was critical to facebook’s eventual success. It seems self-evident today, that without Zuck’s laser-focus on his vision of connecting the world in thru a social graph, facebook wouldn’t be where it is today. Moreover, replacing Zuck with a ‘professional’ CEO early on would have likely led facebook down a different path, one in which revenue would have likely become a key focus too early and one that would have likely led to the demise of facebook. Michael Birch was the founding CEO of Bebo and remained so until the company was sold to AOL. His clear vision for Bebo, his leadership and product-focus was also critical for Bebo’s success.
9. Politics. There’s no place for politics in a fast-moving startup! Friendster’s product and technology teams simply didn’t get along. There are a lot of reasons for this, which I won’t go into detail here, but suffice it to say that it’s critical for the key teams on fast growing startups to work well together. Every start-up has its fair share of politics, that’s human nature. But, it must be at a level that doesn’t cripple or paralyze the company. Facebook in the early days went thru several heads of product before Zuck settled on the right person who could balance their own product desires with Zuck’s vision for the company. At Bebo, we decided not to employ any formal product managers, but rather empower each engineer to be a product manager. Admittedly, this was unorthodox, but it worked. We never had any conflicts between the product team and the engineers! And, we shipped new features every two weeks. Constant innovation and testing was a key to our success.
10. Revenue. Focus on the Product First, Revenue later! This is a tough one for many start-ups and frankly not always the right call. But, for companies that are growing fast like social networks, it’s vitally important to reach an inflection point in growth and engagement before starting to focus on monetization. In 2003, when Friendster was launched, this was a novel concept. Friendster raised a large round of financing ($13M in 2003) and there was pressure to start early on focusing on monetization strategies and revenue. Since this was one of my key responsibilities at Friendster, I’ll take some of the blame here ;-). We created a new marketplace of social advertising where brand advocates and social influencers could be directly targeted. These new revenue strategies are now implemented successfully in massive scale on facebook and other social media sites. I’m proud of creating this billion dollar new advertising marketplace, however, in the early days of Friendster, we should’ve been focusing 110% on our customers and growing the user base and engagement. It would have been wiser at Friendster to hold off on our brand advocate revenue strategy for another year or so. Any energies and web real estate focused on monetization early on, meant energies and real estate not devoted to making the product and site great. Zuckerberg at facebook was very clear from the beginning that they were laser-focused on user growth and not monetization, as they had very little advertising and spent very little effort on optimizing their revenue. Once they hit hundreds of millions of customers, they started finally to focus on their monetization strategies and their recent revenue numbers are a testament to their patience. At Bebo, we opted to have no ads on the site for some time in the early days and added only a few ad placements later and avoided the annoying lowbrow banner ads (like those annoying punch the monkey banners—remember those?). Once we achieved a certain scale at Bebo, we then were able to do some very creative integrated big brand integrated advertising deals that were highly profitable.
