Does the $1 million Netflix contest foreshadow a winner-take-all economy?
For more than three years, teams of heavily invested and entirely uncompensated computer scientists poured their creative efforts into a goal that only a true techie could appreciate -- improving the algorithms that Netflix uses to make movie suggestions to its customers. It was called the Netflix contest, and to the winners went the spoils: the $1 million Netflix Prize.
After the Netflix contest awards were handed out on September 21, even the members of the losing teams professed that the learning experience of trying to solve the intractable problem of improving the movie suggestion algorithms by 10 percent was more far more valuable than the prize money, let alone their time spent.
Perhaps, but to me the Netflix Inc. (NFLX) crowdsource lovefest is actually emblematic of a new and troubling economic trend. Increasingly, companies are seeking to turn what previously would have been paid work into winner-take-all contests. If this becomes the norm, such contests could create a more Hobbesian work life for many people in trades easily attacked by the contest cancer.
The economic theory is simple. Participants gain valuable experience and have fun. But the economics of contests are so appealing to the challenge-issuing companies that I am concerned this trend will further depress wages for rocket scientists and graphic designers alike.
The ongoing decline in real wages spurred on by global competition is already a serious problem. Reed Hastings, the company CEO, crowed in a New York Times article that the average cost of labor supplied by the scientist-contestants was less than $1 per hour. Popular Web technology blog ReadWriteWeb called the $1 million price tag "a steal." This is more a peril than a sign of promise and a trend to be watched very closely for possible exploitation.
Am I a killjoy? Certainly, everyone loves a contest. And worse than being a killjoy, I have been a willing contributor to this trend. Last year, I used a service called CrowdSpring to obtain a logo for a new website I wanted to build. I offered $200, which was close to standard at that time. I got back roughly 40 logos, including some entries that had multiple designs. Even in a best-case scenario, these logos must have taken a minimum of 30 minutes to design. A more realistic estimate of the time required would be a full hour. In other words, talented graphics designers were working for $5 to $10 per hour to design my logo. A monotonous data entry job pays more than that. Anyone seeking to use this type of contest as a form of subsistence might as well apply for food stamps.
A growing array of sites and companies are offering similar compete-for-work models. GeniusRocket runs a similar site targeting more complex marketing and advertising needs, right up to one-minute spots designed to run on television. InnoCentives bills itself as a marketplace for business research, where groups can compete to solve business problems for prizes. And more companies seeking to set up competitions aimed at delivering real work are in the process of launching.
While this constitutes an incredibly tiny fraction of the actual economy (probably measured in tens of millions of dollars), the trend underscores a number of unpleasant things about work today. First, that much of the workforce is replaceable and interchangeable. Second, that the lure of a prize can appeal to the irrational side of human psychology. This may sound paternalistic, but personally I find it unpleasant to encourage people to do things that I know are not worth their time. Perhaps they have their reasons -- they are just getting started in their careers, etc -- but that's also what internships are for, and at least with those you get some actual connections.
Many participants in the Netflix competition told The New York Times that they relished working on such a large, rich set of data and were able to develop new programming skills and techniques that would stand them in good stead in other business and research efforts. In this, the contest echoes a troubling trend in academia, a professional sector that has increasingly become a two-class society, with low paid post-docs wandering from lab to lab to survive while major universities continue to reduce the number of tenured, and even untenured, permanent positions. The views of those computer scientists who claimed the contest was a hoot may be revised in the not-so-distant future when such bake-offs truly become a standard version of outsourcing.
It's bad enough to invest time and energy into a huge RFP that one's company has no guarantee of winning. How about handing over the entire work product with no guarantee of any compensation? In this light, I think Netflix CEO Reid Hastings would do the world a service if he would modify his model somewhat and make it more like the World Series of Poker-type competitions. In those contests, a much larger percentage of contestants actually walk out with some money, and the winner does take home a lot, but not everything.
Netflix immediately followed up the announcement of its prize-winner with word of another contest that has somewhat similar rules. When asked if he would participate in the next one, winning team member Bob Bell told the Times that he liked the idea, but was too tired. Welcome to the winner-take-all economy, my friend.