In addition to these 5 general design principles, I’d like to offer two other pieces of advice. First, the immediate reaction that you might encounter when you propose a mass collaboration project is “Nobody would participate.” Of course that might be true. In fact, lack of participation is the biggest risk that mass collaboration projects face. However, this objection usually arises from thinking about the situation in the wrong way. Many people start with themselves and work out: “I’m busy; I wouldn’t do that. And, I don’t know anyone that would do that. So, nobody would do that.” Instead of starting with yourself and working out, however, you should start with the entire population of people connected to the Internet and work in. If only 1-in-a-million of these people participate, then your project could be a success. But, if only 1-in-a-billion people participate, then your project will probably be a failure. Since our intuition is not good at distinguishing between one-in-a-million and one-in-a-billion, we have to acknowledge that it is very hard to know if projects will generate sufficient participation.
To make this a bit more concrete, let’s return to Galaxy Zoo, the human computation astronomy project discussed earlier in this chapter. Imagine Kevin Schawinski and Chris Linton, two astronomers sitting in a pub in Oxford thinking about Galaxy Zoo. They would never have guessed—and never could have guessed—that Aida Berges—a stay-at-home mother of 2 who lives in Puerto Rico—would end up classifying hundreds of galaxies a week (Masters 2009). Or, consider the case of David Baker, the biochemist working in Seattle developing Foldit. He could never have anticipated that someone from McKinney, Texas named Scott “Boots” Zaccanelli, who worked by day as a buyer for a valve factory, would spend his evenings folding proteins, eventually rising to a number 6 ranking on Foldit, and that Zaccaenlli would, through the game, submit a design for a more stable variant of fibronectin that Baker and his group found so promising they decided to synthesize it in their lab (Hand 2010). Of course, Aida Berges and Scott Zaccanelli are atypical, but that is the power of the Internet: with billions of people, it is typical to find the atypical.
Second, given this difficulty with predicting participation, I’d like remind you that creating a mass collaboration project can be risky. You could invest a lot of effort building a system that nobody will want to use. For example, Edward Castronova—a leading researcher in the field of economics of virtual worlds, armed with a grant of $250,000 from the MacArthur Foundation, and supported by a team of developers—spent nearly two years trying to build a virtual world within which he could conduct economic experiments. In the end, the whole effort was a failure; nobody wanted to use Castonova’s virtual world (Baker 2008).
Given the uncertainty about participation, which is hard to eliminate, I suggest that you try to use lean start-up techniques (Blank 2013): build simple prototypes using off-the-shelf software and see if you can demonstrate viability before investing in lots of custom software development. In other words, when you start pilot testing, your project will not—and should not—look as polished as Galaxy Zoo or eBird. These projects, as they are now, are the results of years of effort by large teams. If your project is going to fail—and that is a real possibility—then you want to fail fast.