eqtventures

Facebook, Data and Growth: Henrik Landgren

Henrik Landgren

It’s often in life that it’s the times where you come across painful stumbling blocks or challenges that cause you to question what your priorities really are. But it isn't just people who can have these moments of pain and reflection; businesses can have them too. It was such a moment that led Facebook to focus on analytics, according to Alex Schultz, Facebook's VP of growth marketing, analytics and i18n.

Alex and I spoke at a recent round table dinner that EQT Ventures hosted, on the topic of Analytics: from startup to Global Growth. Alex said that Facebook had gone through these moments of real challenge a few times. Each time the site stopped growing for a short period. But it was the first time, in October 2007, that forced a change in thinking.

He said that the social network still doesn't know exactly what caused that this halt in growth but it led to the creation of the Growth team. Many of Facebook’s internal growth charts go back only to June 2007 when the company set up Hadoop and they could begin to collect and structure data on its growth to help understand what was really going on.

It was interesting to compare Alex's experiences with my own, at Spotify. There was no data team there either, at first, and thanks to Hadoop, we could also store all events and start analyzing to understand what our users actually did. We started by working with the product team. We began asking ourselves what to build next and, in doing so, had to examine what the users were doing. Soon we had reached deeper questions, such as "what is the difference in behaviour between people who stay for a long time, versus people who churn out?" We realised we needed data to answer those questions and allow us to make product decisions based on user behaviour and not gut feelings.

Facebook’s mission is to make the world more open and connected and give people the power to share. However, the company identified three main barriers keeping people offline; access to infrastructure, affordability and lack of awareness of the value of being online. In an effort to address these barriers, Facebook created Internet.org, an initiative to connect 4 billion people who don’t have access to connectivity.

Facebook has a variety of programs under Internet.org, and through these efforts has connected 40 million people, based on best estimates. As with many of the company's ideas, it did not come from nowhere; Alex said that it was informed by earlier projects, such as Facebook Zero, a partnership with mobile internet providers that allowed for free access to Facebook.

Alex said that until 2012, data was collected and used by the tech team, but product analytics as it exists today, where it is used across the business was not in existence. I was interested in how that change was brought about because I've found that in encouraging people to adopt analytics, the hardest thing to do is to change the mindset. People become enormously reliant on their gut feelings. That shouldn't stop but people have to train their gut with analytics.

At Facebook, Alex said, adoption of analytics was helped by the model put in place by Naomi Gleit, one of Facebook's first employees and now VP of Product Management for Social Good. She introduced a three-step framework: understand, identify and execute. Analytics, said Alex, helps the understanding at the beginning and then, once the product team has decided which idea to pursue, it can help to validate them.

That isn't to say that there is no place for gut feeling. Alex said that early in 2008 Facebook was trying to get to 100m users. No social network had ever done that and many people thought it was impossible. The company was caught in "analysis-paralysis", said Alex, when it came to determining the importance of friends to growth and usage of the service.

I brought up the famous magic growth number of 10 friends that Facebook has talked about. Alex said that number came from the founder Mark Zuckerberg himself to set a benchmark - 10 friends in 14 days - for successful growth. It was informed by our analysis, but the exact number was entirely arbitrary, said Alex. They could have picked any point along a smooth curve of user engagement but once they had something to focus on, they were able to move forward.

Alex had plenty of insights into the power of analytics within a business and it was a fascinating evening. The most powerful insight the crowd took away with them might be the fact that even a company as big and successful as Facebook has to go through the occasional near-death experience. We also had these defining near-death moments at Spotify, and it is fascinating to realize how similar startups really are when they evolve themselves and grow from start-up to global.