Should an Algorithm Determine Content?

From Beta Daily Finance:

Making online media pay: Demand Media vs. The Texas Tribune

Bruce Watson

This has been a tough year for the media business: as august newspapers and magazines like the Seattle Post-Intelligencer and Gourmet have closed their doors for good, other outlets are scrambling to develop strategies that will guarantee long-term economic survival while allowing them to continue producing quality content.

This month, Rupert Murdoch entered the fray, going toe to toe with Google (GOOG) over advertising revenues and access to News Corp.'s (NWS) online newspapers. As he tries to work out a revenue-sharing scheme with Microsoft's (MSFT) Bing, it will be interesting to see how the Internet's low-cost distribution structure ultimately affects the production of news and other media.

Traditional newspapers and magazines use three revenue sources -- subscriptions, newsstand sales, and advertising -- to fund highly expensive news-gathering, writing, and editing operations. While some publications like Murdoch's Wall Street Journal have managed to make paid content work, most Internet sites provide content for free, which puts funding solely into the hands of advertisers. The question is whether Internet ads can generate enough revenue to fund media production.

Make It Cheaper and Make Lots of It

Demand Media, a company based in Santa Monica, California, offers one solution to this problem: lower the cost of producing content until it becomes profitable. It works like this: Demand uses an algorithm to scour the Internet, focusing on ads, keyword searches, and other publishing platforms, in order to determine the topics that people want to read about. A second algorithm then generates story ideas, predicts how much ad revenue they will generate, and determines how much they are worth. Freelance writers and videographers write or film the pieces for either a one-time payment, generally in the $10 to $15 range, or a cut of the ad revenue.

According to Wired, the Demand system currently produces somewhere in the neighborhood of 4,000 articles and videos per day; by next summer, it hopes to reach 1 million per month. This massive array of articles and videos -- and the resulting advertising -- has allowed Demand to achieve the holy grail of online content providers: it is profitable. In 2009, its owners anticipate revenues of $200 million.

As someone who has made a living from freelance writing, I tend to view Demand the same way that a cow views McDonald's. For a while, I looked into working for the Demand Studios: seduced by the possibility of getting on one of their premium sites, I considered the economics of churning out articles on relationships and personal finance for eHow. I soon realized that, at $10 to $15 per piece, I would need to work something like twelve hours per day to finance my irritating addiction to food, shelter and clothing.

Given that kind of schedule, I would have to forego things like organizing, proofreading, and editing my work. While I might get a lot of bylines, none of them would be pieces that I'd want prospective employers to see; more important, the breakneck pace wouldn't really give me much time to look for something better. In short, I could easily see a future in which my work for Demand would transform me into a sort of Internet serf, cranking out work on the Demand treadmill in the hopes of one day getting an assignment to something more prominent.

To be fair, Demand isn't quite the low quality content machine that its critics sometimes make it out to be. While eHow is jam-packed with slapped-together articles on topics like "How to select Mickey Mouse bath toys," "How to chew gum," and "How to tie your shoes," Demand also hosts premium sites -- like Trails.com and Golflink.com -- that feature well-written, well-edited articles. And Demand's Cracked.com transcends its namesake magazine, offering an endless stream of thought provoking lists about pop culture.

The End of the Cheap Road?

For writers, the real terror isn't Demand itself, but rather sites like GetAFreelancer.com that suggest where the cheap content road may ultimately lead. On Get a Freelancer, reverse-auctions pit writers from around the world against each other in a fight for the lowest possible price. Right now, for example, there is an open auction in which a company is trying to get a writer to produce forty 200-300 word articles about marriage. The current low bid, $93, means that the auction winner will make roughly $2.32 per article, or between $0.007 and $0.01 per word.

Apart from the steel-cage-match motif of Get a Freelancer, there is the question of value. At these prices, the site can't hope to produce articles that are fresh, new, or creative. While many are search-engine-optimized and highly clickable, they aren't the sorts of things that readers are likely to spend a lot of time on. Adding little to the sum of human knowledge, they are the writing equivalent of bubble gum: bright and shiny, but possessing no nutritional value.

Pay for Highly Focused Quality

On the other hand, Demand Media is highly profitable, which is more than one can say for The Texas Tribune. If Demand represents the scattershot, market-driven, profit-based end of the online content spectrum, the Tribune represents the other extreme. Very narrowly focused -- it only covers Texas politics at the state level -- the site has a staff of sixteen well-paid writers, editors and Web developers producing the kind of incisive, thoughtful, well-researched journalism that is generally associated with print newspapers.

The trouble is, this sort of well-written, vitally important content doesn't pay all that well. The Tribune, for example, is a nonprofit, largely funded by donations and sponsorships. In Texas, where state loyalty and identity borders on obsession, it seems like the Tribune model might work; in less affluent, less cohesive states like West Virginia or Michigan, it might be less effective. In general, the nonprofit model of news seems as inadequate an answer as the profit-driven model of content.

Comparing Demand and The Texas Tribune is admittedly problematic. After all, Demand is profit-driven, while the Tribune is nonprofit; Demand employs thousands of freelancers, while the Tribune has a small, cherry-picked staff; and Demand is designed to answer every question under the sun, while the Tribune covers the comparatively microscopic topic of Texas state politics. Perhaps most importantly, the Tribune is a news source, while Demand focuses solely on general-purpose content that doesn't have an expiration date.

While relevant, these differences obscure a more basic similarity: Demand and the Tribune are both focused on providing intriguing, eye-catching content that will inspire clicks. And, while their respective models are miles apart, they offer two possible solutions to the problem of profit and quality. The lingering question is if there is a way of navigating these two goals to produce information that is valuable to readers, scalable, and profitable enough to survive.