Lately, I’ve been thinking about SNS and networked political action. As a result, I’ve put together this small report on the Facebook feeds fiasco. I’ve posted the report here to my blog, but it will likely be easier for you to just download it as a PDF. I’m not exactly sure what I’m going to do with this report, hence no APA-style cites, but I wanted to jot this info down before it left my mind. I think what happened on the Facebook between 9/5/2006 and 9/7/2006 was very important, and will figure into my dissertation. Anyway, enjoy, and please feel free to comment/critique/add information, as I will certainly be revising this document later.
Fred Stutzman, UNC-Chapel Hill
Case Study: Facebook Feeds and Networked Political Action
Around midnight-1AM Eastern, Facebook mini-feeds were rolled out. Assorted LiveJournal users noticed and mentioned the service.
As evidenced, initial reaction in the blogosphere was negative, with many users claiming invasion of privacy. A-list bloggers, on the other hand, generally gave the service a thumbs-up. Throughout the morning of the 5th, discontent amongst users grew, with a number of users forming groups against the Facebook feeds.
Sometime during the morning of the 5th, Ben Parr, a junior at Northwestern University, created a group entitled “Students Against Facebook News Feed (Official Petition to Facebook).” In interviews (1,2), Ben states that he created the group, told a few friends, and went to lunch. When Ben returned from lunch, the group had grown to over 13,000 users.
Two main factors influenced the rapid expansion of the Students Against Facebook News Feed (SAFNF) group. First, the Facebook feed product was introduced poorly, causing discontent within the Facebook from the very beginning. The product was an invasive change that affected privacy expectations in the site. Furthermore, users were not prepared for the change in advance, nor were they able to opt-out of the service.
The second factor was an artifactual element of the Facebook feeds. The general purpose of the Facebook feed product is to inform the cohort of actions taken by a Facebook users. Common actions include writing on a wall, editing profile information, or joining or leaving groups. When a user logs in to Facebook, they were immediately presented with the last actions of their friend network.
On the morning of the 5th, Ben Parr’s friends saw that he created a group in the Facebook. They joined his group to show solidarity with his feelings about Facebook feeds. When they did join the group, that action was recorded on their news feeds, and thereby broadcast out to their entire friend network. As the average collegiate Facebook user has hundreds of friends, this action of solidarity in joining SAFNF became a living, viral advertisement for the SAFNF group. Each time someone who had a friend that joined SAFNF logged in, they were presented with a message alerting them to their friend’s intentions. As sentiment went pervasively negative against feeds, more people joined SAFNF, thereby virally spreading the group.
In the Facebook, groups are arbitrary affiliation vectors. Groups can be whimsical, such as a group named after a movie quote, or serious, such as a group dedicated to volunteering. Groups are costless to join, and they simply require a click of the mouse to join. Many users partake in a large number of groups; groups are generally thought of as a low-involvement way to make identity statements. Although barriers to group entry are low, this does not mean that group membership lacks meaning or consequence. As groups scale sufficiently, they become means for a mass audience to express a viewpoint. In Ben Parr’s case, the viewpoint was anger over the Facebook feeds, and it turns out his group was instrumental in the movement against the product.
Throughout the afternoon of the 5th, many users joined SAFNF. At 9PM Eastern, the group had grown to 30,000 members. Just two hours later, at 11PM Eastern, the group had over 45,000 members. (3) At the same time, users reportedly started venting their frustration on the Facebook blog (http://blog.facebook.com). According to an unverified thread (4), users left over 600 comments about the service, “every single one negative.” The Facebook blog reportedly removed commenting, and has never allowed commenting on its blog since. The group started growing at a sustained pace of approximately 15-20,000 new users per hour into the morning of the 6th. Reports have the SAFNF count at 90,000 members at 1AM on September 6. (5) By 3:45AM Eastern, Mark Zuckerberg issued the blog post that told Facebook users to “Calm Down. Breathe. We hear you.” (6)
As users woke up on the morning of September 6th, they were inundated with messages showing that members of their friend network had joined SAFNF. Over the course of the 6th, SAFNF membership was reported at: 300,000 members at 11AM Eastern, 400,000 members at 3PM Eastern, 500,000 members at 8PM Eastern. (7) At the same time, many users were voicing their frustration on the SAFNF discussion board. By midnight on the 6th, over 2600 messages had been posted in the discussion forum. (8)
Up until this point, the main coverage of the Facebook feeds event happened in the blogosphere, with a number of high-profile blogs (Techcrunch, Scobelizer) weighing in on the fiasco. On the morning of the the 7th, the Wall Street Journal, Washington Post and Los Angeles Times all ran coverage of what was being dubbed the Facebook Feeds Fiasco. (9,10,11) Throughout the course of the 7th, the group swelled to 750,000 members, representing almost 8% of the entire Facebook population.
On the night of the 7th, Ben Parr made the following announcement via the SAFNF website.
- I was read a statement from facebook by a news organization. Perhaps by tomorrow, there will be a feature to remove yourself from facebook news feeds, based on categories. This group will reserve judgement until we see the privacy upgrades implimented.
I think we’ve won, people. I think we’ve won.
- Ben (12)
At this point, the group had essentially accomplished its mission. The Facebook had agreed to make substantial changes to the news feeds, which were later documented in a post to the Facebook blog. Early on the morning of September 8, Mark Zuckerberg wrote:
We really messed this one up. When we launched News Feed and Mini-Feed we were trying to provide you with a stream of information about your social world. Instead, we did a bad job of explaining what the new features were and an even worse job of giving you control of them. I’d like to try to correct those errors now. (13)
On the night of the 8th, the Facebook introduced privacy controls to the feed product, allowing users the ability to opt-out of the feeds. Zuckerberg neglected to mention the SAFNF group, but it is commonly believed that the SAFNF action played a significant role in forcing the Facebook to roll back its feeds product. On the 8th, dozens of newspapers ran stories documenting the feeds fiasco, and the community response.
Key Facts (All data approximate)
- 9/5/2006, feeds introduced
- 9/5/2006, SAFNF created.
o 9/5/2006 Midnight membership 90,000 (Eastern, approximate)
o 9/6/2006 Midnight membership 600,000 (Eastern, approximate)
o 9/7/2006 Midnight membership 750,000 (Eastern, approximate)
- To date, over 5200 unique discussion threads on the SAFNF
o 9/5/2006, First post at 1:52PM
o 9/5/2006, 500 threads generated.
o 9/6/2006, 2100 threads generated.
o 9/7/2006, 1260 threads generated.
o 9/8/2006, 450 threads generated.
- Blog posts about SAFNF (via Google Blogsearch)
o 9/5/2006, 26 posts.
o 9/6/2006, 137 posts.
o 9/7/2006, 103 posts.
o 9/8/2006, 95 posts.
The Facebook news feeds are a valuable case study in botched software rollouts. First and foremost, the rollout was a substantial new feature that effectively changed the privacy dynamic in the service. Second, the users were kept in the dark about this change until the last moment, when it was rolled out to all members. Third, users were not allowed to disable the rollout, nor were they able to do anything about it. The rollout was further complicated because it contained retroactive data, i.e. news feed items prior to the launch. Users were very uncomfortable with this rollout and used any means necessary to voice their opinions; they turned to blogs, messages, wall posts and ultimately, the SAFNF group.
2. Students Against Facebook News Feeds
There are currently approximately 150 groups on the Facebook that are “against” the Facebook news feeds. Their membership ranges from the tens on the low side to low thousands on the high side. How did this one group gain such immense popularity? These reasons are speculative, but one can assume the following. First, Ben Parr and his friends were very active Facebook users. They likely had large friend groups, comprised of active users. They were likely also influential Facebook users. As the influential Facebook users joined SAFNF, the network effect started very quickly. Soon, the group had spread within the entire network, so that users at virtually every institution had a friend or two that had joined SAFNF. The most important reason for the spread of SAFNF, interestingly, is the news feed product. Because users could see the groups their cohort was joining, they were instantly made aware of the SAFNF group, as well as being influenced to join the group. This made for very pervasive spread of the group’s goal.
3. Facebook response
The Facebook initially did not respond to the SAFNF group. Over the course of the 5th and 6th, many employees of the Facebook created and joined groups that supported the Facebook feeds product. Many of these groups were derogatory to users who did not support the feeds product. Mr. Zuckerberg’s initial blog posting (6) was also widely considered a blow-off to the users of the Facebook concerned about the product. His message served to inflame the users, with many responding negatively to Mr. Zuckerberg in the SAFNF comment thread.
4. General concerns
Users generally considered the introduction of Facebook feeds as a privacy invasion. They commonly resorted to calling the Facebook by its nickname, the “Stalkerbook”, in blog posts. Many users considered the Facebook to truly be the “Stalkerbook” with the introduction of feeds. Interestingly, once Facebook introduced privacy features, users welcomed the service, with many users starting groups dedicated to admitting that they enjoyed the feeds product.
5. What went wrong
The botched rollout of the feeds, combined with perceived arrogance on behalf of the Facebook employees contributed heavily to the user revolt of 9/5-9/7. Facebook neglected to take into account users privacy and identity expectations, and they came off looking like they were playing loose and fast with a very fundamental aspect of the user’s experience. The fact Facebook completely neglected to introduce privacy functions was likely the largest motivator for the revolt. How product managers of Facebook neglected such an important feature is likely a result of groupthink, in which Facebook employees lacked the perspective of the average user. Had the Facebook introduced privacy features from the beginning, it is clear that the revolt of 9/5-9/7 would not be anywhere near as significant as it was.
Networked Political Action
The feeds fiasco is an important example of networked political action. In the course of just a few days, hundreds of thousands of users bonded together to express their negative opinion about the feeds product. This action was pervasive and effective, and completely organic.
A combination of message validity, timing, and network dynamics was responsible for the effectiveness of the action. Because the users were at stake, this drove participation. Further, participation was driven by the simplicity of the action – to make a statement, users simply needed to join the group. At a small scale, this simple action of joining a group has limited value, but at scale the group grows very powerful. The creation of large groups on Facebook is now somewhat of a game in the service.
The particular circumstances of the feeds fiasco are responsible for some of the success of the campaign, but there are some key takeaways. First, it is interesting to see that users understand the value of signifying political intention with group membership. Theoretically, groups could grow quite valuable as a democratic messaging platform within social network services. It is also interesting to see how rapidly information diffiuses through SNS networks. We know that 8% of Facebook joined the group, but it is obvious that many more didn’t, but were still exposed to the message of the group. Therefore, groups are excellent collective message dissemination vehicles. At scale, groups could easily reach all members of the SNS.
In a SNS, there is no notion of a democracy. Users are subject to the rules, regulations and changes in the SNS. As a result, the users will resort to using in-network means to express their voice. As we have seen, users can effectively leverage their collective will to influence the “rulers” of the sites in which they inhabit. Obviously, this type of action has interesting implications going forward.
9. IN BRIEF / INTERNET; Facebook’s Tracking Feature Draws Protests, Los Angeles Times, 9/7/2006.
10. New Facebook Features Have Users in An Uproar. Jamin Warren and Vaughini Vara, Wall Street Journal, 9/7/2006.
11. In Online Social Club, Sharing Is the Point Until It Goes Too Far, Susan Kinzie and Yuki Noguchi, Washington Post, 9/7/2006.