After studying this section you should be able to do the following:
- Recognize that Facebook’s power is allowing it to encroach on and envelop other Internet businesses.
- Understand the concept of the “dark Web” and why some feel this may one day give Facebook a source of advantage vis-à-vis Google.
- Understand the basics of Facebook’s infrastructure, and the costs required to power the effort.
The prior era’s Internet golden boy, Netscape founder Marc Andreessen, has said that Facebook is “an amazing achievement one of the most significant milestones in the technology industry” (Vogelstein, 2007). While still in his twenties, Andreessen founded Netscape, eventually selling it to AOL for over $4 billion. His second firm, Opsware, was sold to HP for $1.6 billion. He joined Facebook’s Board of Directors within months of making this comment. Why is Facebook considered such a big deal?
First there’s the growth: between December 2008 and 2009, Facebook was adding between six hundred thousand and a million users a day. It was as if every twenty-four hours, a group as big or bigger than the entire city of Boston filed into Facebook’s servers to set up new accounts. Roughly half of Facebook users visit the site every single day, (Gage, 2009) with the majority spending fifty-five minutes or more getting their daily Facebook fix1. And it seems that Mom really is on Facebook (Dad, too); users thirty-five years and older account for more than half of Facebook’s daily visitors and its fastest growing population (Hagel & Brown, 2008; Gage, 2009).
Then there’s what these users are doing on the site: Facebook isn’t just a collection of personal home pages and a place to declare your allegiance to your friends. The integrated set of Facebook services encroaches on a wide swath of established Internet businesses. Facebook has become the first-choice messaging and chat service for this generation. E-mail is for your professors, but Facebook is for friends. In photos, Google, Yahoo! and MySpace all spent millions to acquire photo sharing tools (Picasa, Flickr, and Photobucket, respectively). But Facebook is now the biggest photo-sharing site on the Web, taking in some three billion photos each month1. And watch out, YouTube. Facebookers share eight million videos each month. YouTube will get you famous, but Facebook is a place most go to share clips you only want friends to see (Vogelstein, 2009).
Facebook is a kingmaker, opinion catalyst, and traffic driver. While in the prior decade news stories would carry a notice saying, “Copyright, do not distribute without permission,” major news outlets today, including the New York Times, display Facebook icons alongside every copyrighted story, encouraging users to “share” the content on their profile pages via Facebook’s “Like” button, scattering it all over the Web. Like digital photos, video, and instant messaging, link sharing is Facebook’s sharp elbow to the competition. Suddenly, Facebook gets space on a page alongside Digg.com and Del.icio.us, even though those guys showed up first.
Facebook Office? Facebook rolled out the document collaboration and sharing service Docs.com in partnership with Microsoft. Facebook is also hard at work on its own e-mail system (Blodget, 2010), music service (Kincaid, 2010), and payments mechanism (Maher, 2010). Look out, Gmail, Hotmail, Pandora, iTunes, PayPal, and Yahoo!—you may all be in Facebook’s path!
As for search, Facebook’s got designs on that, too. Google and Bing index some Facebook content, but since much of Facebook is private, accessible only among friends, this represents a massive blind spot for Google search. Sites that can’t be indexed by Google and other search engines are referred to as the dark Web. While Facebook’s partnership with Microsoft currently offers Web search results through Bing.com, Facebook has announced its intention to offer its own search engine with real-time access to up-to-the-minute results from status updates, links, and other information made available to you by your friends. If Facebook can tie together standard Internet search with its dark Web content, this just might be enough for some to break the Google habit.
And Facebook is political—in big, regime-threatening ways. The site is considered such a powerful tool in the activist’s toolbox that China, Iran, and Syria are among nations that have, at times, attempted to block Facebook access within their borders. Egyptians have used the site to protest for democracy. Saudi women have used it to lobby for driving privileges. ABC News cosponsored U.S. presidential debates with Facebook. And Facebook cofounder Chris Hughes was even recruited by the Obama campaign to create my.barackobama.com, a social media site considered vital in the 2008 U.S. presidential victory (Talbot, 2008; McGirt, 2009).
So What’s It Take to Run This Thing?
The Facebook cloud (the big group of connected servers that power the site) is scattered across multiple facilities, including server farms in San Francisco, Santa Clara, and northern Virginia (Zeichick, 2008). The innards that make up the bulk of the system aren’t that different from what you’d find on a high-end commodity workstation. Standard hard drives and eight core Intel processors—just a whole lot of them lashed together through networking and software.
Much of what powers the site is open source software (OSS). A good portion of the code is in PHP (a scripting language particularly well-suited for Web site development), while the databases are in MySQL (a popular open source database). Facebook also developed Cassandra, a non-SQL database project for large-scale systems that the firm has since turned over to the open source Apache Software Foundation. The object cache that holds Facebook’s frequently accessed objects is in chip-based RAM instead of on slower hard drives and is managed via an open source product called Memcache.
Other code components are written in a variety of languages, including C++, Java, Python, and Ruby, with access between these components managed by a code layer the firm calls Thrift (developed at Facebook, which was also turned over to the Apache Software Foundation). Facebook also developed its own media serving solution, called Haystack. Haystack coughs up photos 50 percent faster than more expensive, proprietary solutions, and since it’s done in-house, it saves Facebook costs that other online outlets spend on third-party content delivery networks (CDN) like Akamai. Facebook receives some fifty million requests per second (Gaudin, 2009), yet 95 percent of data queries can be served from a huge, distributed server cache that lives in over fifteen terabytes of RAM (objects like video and photos are stored on hard drives) (Zeichick, 2008).
Hot stuff (literally), but it’s not enough. The firm raised several hundred million dollars more in the months following the fall 2007 Microsoft deal, focused largely on expanding the firm’s server network to keep up with the crush of growth. The one hundred million dollars raised in May 2008 was “used entirely for servers” (Ante, 2008). Facebook will be buying them by the thousands for years to come. And it’ll pay a pretty penny to keep things humming. Estimates suggest the firm spends $1 million a month on electricity, another half million a month on telecommunications bandwidth, and at least fifteen million dollars a year in office and data center rental payments (Arrington, 2009).
- Facebook’s position as the digital center of its members’ online social lives has allowed the firm to envelop related businesses such as photo and video sharing, messaging, bookmarking, and link sharing. Facebook has opportunities to expand into other areas as well.
- Much of the site’s content is in the dark Web, unable to be indexed by Google or other search engines. Some suggest this may create an opportunity for Facebook to challenge Google in search.
- Facebook can be a vital tool for organizers—presenting itself as both opportunity and threat to those in power, and an empowering medium for those seeking to bring about change.
- Facebook’s growth requires a continued and massive infrastructure investment. The site is powered largely on commodity hardware, open source software, and proprietary code tailored to the specific needs of the service.
Questions and Exercises
- What is Facebook? How do people use the site? What do they “do” on Facebook?
- What markets has Facebook entered? What factors have allowed the firm to gain share in these markets at the expense of established firms? In what ways does it enjoy advantages that a traditional new entrant in such markets would not?
- What is the “dark Web” and why is it potentially an asset to Facebook? Why is Google threatened by Facebook’s dark Web? What firms might consider an investment in the firm, if it provided access to this asset? Do you think the dark Web is enough to draw users to a Facebook search product over Google? Why or why not?
- As Facebook grows, what kinds of investments continue to be necessary? What are the trends in these costs over time? Do you think Facebook should wait in making these investments? Why or why not?
- Investments in servers and other capital expenses typically must be depreciated over time. What does this imply about how the firm’s profitability is calculated?
- How have media attitudes toward their copyrighted content changed over the past decade? Why is Facebook a potentially significant partner for firms like the New York Times? What does the Times stand to gain by encouraging “sharing” its content? What do newspapers and others sites really mean when they encourage sites to “share?” What actually is being passed back and forth? Do you think this ultimately helps or undermines the Times and other newspaper and magazine sites? Why?
1“Facebook Facts and Figures (History and Statistics),” Website Monitoring Blog, March 17, 2010.
Ante, S., “Facebook: Friends with Money,” BusinessWeek, May 9, 2008.
Arrington, A., “Facebook Completes Rollout of Haystack to Stem Losses from Massive Photo Uploads,” TechCrunch, April 6, 2009.
Blodget, H., “Facebook’s Plan To Build a Real Email System and Attack Gmail Is Brilliant,” Business Insider, February 5, 2010.
Gage, D., “Facebook Claims 250 Million Users,” InformationWeek, July 16, 2009.
Gaudin, S, “Facebook Rolls Out Storage System to Wrangle Massive Photo Stores,” Computerworld, April 1, 2009, http://www.computerworld.com/s/article/9130959/Facebook_rolls_out_storage_system_to_wrangle_massive_photo_stores.
Hagel J., and J. S. Brown, “Life on the Edge: Learning from Facebook,” BusinessWeek, April 2, 2008.
Kincaid, J., “What Is This Mysterious Facebook Music App?” TechCrunch, February 2, 2010.
Maher, R., “Facebook’s New Payment System Off to Great Start, Could Boost Revenue by $250 Million in 2010,” TBI Research, February 1, 2010.
McGirt, E., “How Chris Hughes Helped Launch Facebook and the Barack Obama Campaign,” Fast Company, March 17, 2009, http://www.fastcompany.com/magazine/134/boy-wonder.html.
Talbot, D., “How Obama Really Did It,” Technology Review, September/October 2008.
Vogelstein, F., “How Mark Zuckerberg Turned Facebook into the Web’s Hottest Platform,” Wired, September 6, 2007.
Vogelstein, F., “Mark Zuckerberg: The Wired Interview,” Wired, June 29, 2009.
Zeichick, A., “How Facebook Works,” Technology Review, July/August 2008.