The astonishing success of Facebook’s photos application led to a bout of soul-searching at the company. What was it, Zuckerberg and his colleagues asked themselves, that made photos so successful? Well, one thing was that you could so easily find new photos your friends uploaded. Each person’s profile included a “dashboard page” that showed which photo albums had most recently been updated. It seemed that users wanted to know what was new. Another recent innovation had been to order the list of friends on each user’s homepage according to which profiles had been changed most recently. They called that “timesorting,” and it won raves from users. Each time someone changed their profile picture, it quickly led to an average of twenty-five new page views.
What people did on Facebook was look at other people’s information. They were eager to learn what was new, what had changed, what had happened that they didn’t already know. Studying your friends’ profiles was an obsessive activity, but not a very efficient one. Click through and try to figure out whether anything had changed since the last time you visited. Was he still single? Did this photo mean she’d been to the Caribbean? How come he went to that party and didn’t tell me? Click click click. The information was good – you wanted to know it – but it was tedious to find.
So the company’s young leaders came up with the idea to build a page that showed not just the latest photos your friends had added, but all the things that had recently changed on the profiles of your friends. “We started asking, ‘How do we get people the information they most care about?’” says Moskovitz. “We wanted to build a screen that showed everything. So we came up with the idea for the News Feed.”
The new tool they arrived at will help users find the information that most mattered to them at any given moment. That might include everything from which party a friend planned to go on Friday to updates about the political situation in Tajikistan someone might have posted as a Web link. The point was to make sure you saw what you cared about, whatever it might be. The order in which information would be presented would depend on what you had shown – by your behavior – you liked to look at. Zuckerberg explained it to colleagues: “A squirrel dying in front of your home may be more relevant to your interests right now than people dying in Africa.”
All this brainstorming took place in the early fall of 2005. Shortly afterward, Adam D’Angelo talked to a new hire, Chris Cox, about building the News Feed. “I saw a glimmer in his eyes,” says Cox. “I could tell that for him it wasn’t about wanting to make money. He said, ‘Look, this is such a broken problem – I can’t find out what‘s going on with my friends!’ The internet could help you answer a million questions, but not the most important one, the one you wake up with every day – ‘How are the people doing that I care about?’”
They set to work on the News Feed. “For the next eight months, it was our labor of love,” says Cox, a tall, laconic, and brainy Stanford grad who had studied computer science, psychology, and linguistics. The idea was audaciously ambitious: to write a set of software algorithms to dissect the information being produced by Facebook’s users, select the actions and profile changes that would be most interesting to their friends, and then present them to those friends in reverse chronological order. Each person’s home page would thus be completely different, depending on who their friends were. “It was the biggest technology challenge the company had ever faced,” says Sean Parker.
The average user of Facebook at that time had about 100 friends. The software would have to watch every action generated by every one of those people. Then, each time you went onto the service, it would rank the activity of all your friends based on the likelihood that you would see it as interesting. That calculation would be based on, among other things, your previous behavior. Perhaps you noted that you were feeling glum or that you were going to the movies, or you uploaded a photo, indicated you liked the new Wilco album, or posted a link to a segment of the Daily Show. Facebook’s software would detect this new information and decide whether to send it to your friends, based on what it calculated is likely to interest them. It would infer this based on its observations of your friends’ previous behavior. If they liked hip-hop they might not get the Wilco info. If they never saw video they might not see the Daily Show link. It would apply such logic to every sort of information and activity on the site. It would repeat this process every fifteen minutes or so. Now multiply all these by 6 million – the number of active users Facebook had at the project’s outset. This was massive engineering and product design challenge.
The News Feed would be a radical change. “It’s not a new feature, it’s a major product evolution,” said Zuckerberg at the time. It would remake Facebook. It was necessary as a foundation for future innovations he was already thinking about. He evangelized it with conviction to the company’s engineers and product designers, not always successfully. “Many of us were ‘No no no, we hate this!’” says product manager Naomi Gleit.
Source: The Facebook Effect by David Kirkpatrick
Liked It? Let everybody know
Pick Your Book