Conclusion: The Rules of Crowdsourcing
The author mentions spectacular failure of the first firm that explicitly called itself "a crowdsourcing company" and that sold software solutions that would enable companies to have plug-and-play solutions. This is a common problem, which often leads critics to denounce a faddish technology as a failure. But it's not the technology, just the way it was used. There is no one-size-fits-all crowdsourcing software that can be installed to begin getting benefits.
He admits that his perspective is one of a journalist rather than a practitioner, but having covered crowdsourcing successes and failures, he has noticed a number of "rules" that the successful efforts seem to follow (and the failures seem to break) and will share them now:
Pick the Right Model
Crowdsourcing isn't one thing, but an umbrella term for various practices whose common element is that they involve leveraging a large group of people in some way. Just as there is not one best way for everyone to do investing or marketing, so is there no best way to do crowdsourcing.
The two basic questions are "what are you trying to do?" and "whose help do you need?" Consider the basic tasks for crowds:
- Contribute intelligence to solving a problem
- Contribute opinion to making a decision
- Contribute labor to complete a task
- Contribute money to finance a purchase
There may be other things you want from your crowdsourcing campaign - but you need tyo know what you want and why you want it.
Pick the Right Crowd
Once you know what you want and why you want it, consider whom you should ask. There are billions of people online, so even if you want people who are "one in a million" personality or very rare skills, there are still thousands you can reach.
Deciding what kind of person you want to help is important - throw open the door to the masses, and "the masses" will show up. Sometimes, that's what you want - other times you want a more specific group of people. Getting people who match the characteristics of your market is critical, or who have knowledge or skills germane to the task, is important to getting the kind of help you need.
Offer the Right Incentives
You know what you want to do and whose help you need. Now ask: why would they do it? Too often, there is the assumption that people "like" to show up and work for you for free, and that if you build a virtual community, people will flock to it. Failure to consider WIIFM from the crowd's perspective is the main cause of empty playgrounds.
People want something tangible - cash and prizes still work for many - or they want something psychological, such as the pride they feel and esteem they get from others when their contribution is recognized.
People will readily donate time and money to a nonprofit organization, but they feel taken advantage of when you make a fortune off of their free labor - so be cautious. If they sense deception, they will leave - and tell others to avoid you.
Keep the Pink Slips in the Drawer
One misconception about crowdsourcing is that the crowd will replace your employees - fire the marketing department and let the crowd promote your products.
The example is given of a company that did this, and the results were not good. The company held a contest to design advertisements for its product. It received a flood of submissions, the "vast majority" of which were "offensive, in violation of copyright laws, or simply bad." The company had to hire an outside firm (having let go of its staff) to sort through all the submissions, ultimately found nothing usable, and abandoned the contest and hired an agency (at greater expense than internal staff).
You will need staff to support and encourage the crowd, as well as staff to organize and filter. One online magazine mentioned that it was even more expensive to produce an issue entirely of user-submitted content than to pay writers. It isn't done to save money, but to make readers feel more connected to the publication.
Support and Direction
Another common misconception is that a crowd works in isolation and maintains its energy and direction without any further support. Generally, the larger the organization the less likely it will be to organize itself without direction and guidance.
He mentions his own experience with volunteer journalism - a lot of people were enthusiastic, but because nobody told them what to do, they did nothing and eventually lost their sense of purpose and wandered off. But at the same time, if you give them too much direction and subject them to too much control, they lose interest all the same.
In software products, there are "benevolent dictators" who step in as needed to provide guidance in open-source software projects. They write the plan that indicates what the software will do, provide just enough direction to keep everyone working in the right direction, and otherwise find ways to guide and support without becoming directive.
Crowds are most efficient when a task can be broken into the smallest possible components - modules that can be independently produced before being assembled into a whole.
While people can be enthusiastic about participating, doing a large task is demotivating, as is having to be closely coordinated with others. So a simple task that can be done quickly is more appropriate for crowdsourcing. A person will gladly spend a few minutes translating a few sentences of text into another language, and they might do this several times - but ask them to translate an entire article and they will not be so enthusiastic.
Remember Sturgeon's Law
Sturgeon's law maintains that "90 percent of everything is bad" - and in speaking with those who have sponsored crowdsourcing, suggesting that only 10% of the contributions will be of acceptable quality is a gross overstatement.
Technology enables people to do things easily - but that does not mean that they are able to do them well. Owning a digital camera does not give a person the ability to shoot professional quality photographs any more than owning a typewriter gives them the ability to write a good story.
So if quality matters, you may have to ask far more people to participate than are necessary to do the task, and spend considerable time filtering through poor contributions to find something adequate.
Remember the Law of Large Numbers
If even 99% of everything is awful, the remaining 1% can be quite substantial if you have a large number of people. Also keep in mind that the crowd can often help sifting through submissions by voting/rating them - they can be the first line of filtering (though they should never be the last).
You cannot command the crowd, you can only guide it - and it may choose to reject your guidance. So remember the principle: government governs best when it governs least. The more you try to control, the more your leadership will suffer from mutiny and defection.
Serving the Crowd
Most who show an interest in leveraging crowdsourcing are concerned with what that crowd can do for them, and fail to consider what they can do for the crowd. And then they are left wondering why nobody wants to participate in their project.
You don't have to pay or bribe them, or at least not very much. Remember that people online crave fame and esteem - being recognized and thanked, praised and admired, are powerful motivators.