hypios

Last fall, Google launched Project 10^100, a call for ideas that would change the world by helping as many people as possible.  Google will choose 5 winning ideas (to be announced shortly) and has pledged $10 million US towards their implementation.  While the design of the contest isn’t very innovative—Google is essentially passing around a suggestion box, and the rewards are traditional grant-style funding—the company seems to have learned from previous crowdsourcing disasters. Here’s what they’ve done right (and what every company tempted to let the people decide should keep in mind) and what they haven’t:

1. Appeal to principles.
The project asks how the world could be changed by helping people. The 10^100 website is well-designed, with an appealing video.  It seems like this project is important, not just something a marketing department was too lazy to do and left up to the public.

Normally, when asked its opinion, the crowd likes to see the mighty fall.  If we’re not members of the community affected by a vote, for example the caring users of a product that we vote about, we have very little incentive to make the right choice.  When Time magazine put its Most Influential Person of 2009 up to an Internet vote, for example, the crowd chose the founder of 4chan, the forum where memes are born and productivity goes to die.  (The crowd in question was made up mostly of forum posters, who were later accused of hacking the vote, not magazine readers.)  Not only did Time have to honor what it called a “stunning result,” it had to acknowledge the influence of the forum that a 2008 Time article, quoting Star Wars, referred to as “a wretched hive of scum and villainy.”

Google isn’t using the crowd’s ideas for a new ad campaign or its own website, but for charity.  Project 10^100 appeals to people’s altruism, not their desire to make themselves heard.  If you submit silly ideas, that is, you’re not hurting Google, you’re hurting the starving children of the world.

2. Review ideas…
Even though they received 154,000 submissions from 170 countries, a team of 3,000 Google employees reviewed them all, eliminating jokes or infeasible suggestions and combining similar ideas into 16 larger projects.  This could be seen as controversial—the project website allows you to see “suggestions that inspired [each] idea,” but never credits specific users with these ideas.  The FAQ explains that with 154,000 submissions to sort through, they may have simply “missed some people,” which is plausible but unfortunate.

An idea review, however patronizing it seems, could have prevented other crowdsourcing disasters.  In 2008, the New York Mets baseball team asked fans to vote on a new song for their 8th inning sing-along.  Five million voters apparently thought it would be fun to Rickroll the stadium, so Rick Astley’s “Never Gonna Give You Up” was played once, booed, and returned to the shelf.  Voters were willing to write-in an obvious joke on their ballots, but were less enthusiastic when it was carried out.  (Or, more likely, those who voted knew they wouldn’t be going to any Mets games and actually sitting through the song.)  The crowd’s ideas and votes do not always reflect what a community—in this case, Mets fans—really wants.

3. …then let the crowd vote.
People still like to feel involved, though, and that’s why Google’s decision to let the public vote is smart.  The company could have simply reviewed ideas and selected winners, saving itself the effort of logging hundreds of thousands of votes.  Instead, Google let people review the 16 chosen submissions themselves.  Not only does Google seem to care about making a difference, it seems genuinely interested in what people think.

4. Get expert advice.
The five winning projects will ultimately be selected by an advisory board of experts in the projects’ fields.  While the experts will take voters’ choices into account, Google’s FAQ makes it very clear that they reserve the right to “discount any votes that [they] deem invalid.”  While this means that Google’s project is not strictly crowdsourced, it’s also protected from the attempts at vote-hacking which plagued Time’s contest.

Of course, this depends on the advisory board, an area where some companies have not fared well.  Kraft’s naming its vegemite-and-cream-cheese spread ‘iSnack 2.0′ is often cited as a crowdsourcing (and marketing) disaster, though the name was recently ‘corrected‘ to ‘Cheesybite.’  Kristen Vang of thinqdesign, however, points out that while the crowd submitted the unfortunate name, someone in Kraft’s marketing department actually selected it as the winner.  An advisory board doesn’t guarantee good results.

5. Use common sense.
The five selection criteria for winning ideas are simple and intuitive: reach, depth, attainability, efficiency, and longevity. Even if you don’t necessarily agree with utilitarianism (the philosophical position that states that what makes more people happy should be pursued), these guidelines are straightforward and represent a strong choice by Google to try to define ‘the project that will do the most good.’

Is Project 10^100 a new and innovative way to crowdsource?  Not really.  It will be interesting to see which ideas are selected and how Google implements them, though. The plan is to partner with existing organizations, who are perhaps better equipped than Google to judge projects’ feasibility.  This grant-style approach is fairly traditional compared with the pay-for-solution incentives or rewards for incremental improvements offered by other so-called crowdsourcing efforts we’ve posted about.

Google’s accomplishment here is learning from other crowdsourcing disasters.  We look forward to the project’s conclusion and, as the 10^100 website says, “may those who help the most win.”

Photo from gruntzooki’s photostream on Flickr.

Share this post!