Crowdsourcing 101: Episode 3 - So You Think You Can Crowdsource?
I wonder how this genetic gold rush effects the search for infectious agents? After all, we now know that whatever genetic predisposition there is towards ulcers it's the Helicobactor pilori that accounts for the vast bulk of the cases. Also, it's becoming clear that bacterial infections in the walls of blood vessels plays a roll in arteriosclerosis and infarcs. Schizophrenia seems to have an infectious component as well. With everyone tumbling into the genetic gold mine how much science is going to be done searching out infectious diseases?
- From a comment posted on Wired.com by arpad
I began the last article with a rant about the quality of comments posted to YouTube (as an example of why it's understandable to fear the crowds). As evidenced above, the crowd can respond wisely. And comments on sites such as Wired.com can produce more astute feedback, than say YouTube. (Note, I'm biased here, I can't love Wired more than I already do - were it a guy, I would blush and wave at him, as I do firemen.)
The attention span component certainly plays a large role in the quality of comments: YouTube viewers are looking for a quick fix (I watch it too, so I'm not judging), while Wired readers often aim to delve deeper into the pot and hence stick around long enough to make often well-articulated, well-thought-out comments. This can sometimes be true as well for the level of participants in crowdsourcing projects, especially when their motivations are primarily based in passion.
And, in a subtle way, we all tap into the wisdom of crowds every time we search for information on the web - initially by using a search engine (Google "which organizes websites based on how they link to each other" or Wikipedia founders' Wikiasari) and then when we, among other things, assess the poster's/site's reliablity.
Wondering how to insert a Windows Media player into your blog? (I was) – well the information is out there, posted by someone who knows more about the subject than you (OK, me). And, just like the companies who employ crowdsourcing, we must filter the information (in our case, choices) to find the best answer (e.g. this site explains it all clearly and is clean and organized: I think I'll trust it, rather than one that looks like Geocities circa 1998).
Only a select few of the comments on Wired.com, in fact, make it into the actual magazine. Filtering is critical and not everyone's contribution is focused, relevant or equal.
properly, [Crowdsourcing] can generate new ideas, shorten research and
development time, cut development costs, and create a direct, emotional
connection with customers." In fact, when appropriately integrated, Crowdsourcing "can be a great way to access new ideas, find solutions to problems or quickly build out that impossible task".
You can use Crowdsourcing to encourage feedback, get others to vet and
weigh ideas, have problems addressed and solved early in the process,
help market and promote your product, and even reduce risk by giving customers what they want (e.g. Threadless).
Of course, it's not right for every company. But if you want to make Crowdsourcing work, here are some tips.
Pick Good Crowds, Ask Good Questions
It is imperative that you are clear on whose input you are soliciting – is it everyone? Likely not. You may only want the feedback of a select few. "For any crowdsourcing activity, the first step is to pick the right crowd! Equally important, you must ask the right question" and "it's a good idea to focus the discussion around one area and clearly define what you're trying to achieve and what the community is all about."
Many cite InnoCentive as a perfect example of filtering the crowd since it "limit[s] audience participation by natural selection. People who join InnoCentive Inc.'s "open innovation marketplace", for instance, tend to be scientists, engineers, inventors and business experts because they're called upon to respond to highly complex challenges posted by organizations, or "seekers."
There is something to be said about how not knowing everything about a company allows people to come up with more creative ideas and solutions and be more inclined to (dare I say it... yes, I dare) "think outside the box".
But some claim decision making is near impossible without that understanding and that it is imperative that contributors be familiar with the specific company/industry/and or task at hand. "[A] microchunk isn't really just a simple task - it comes with a history. Much thought and time and action has been put into whatever it is to get it to the current state. An understanding of that history is necessary if you are going to move the task forward, even if the work itself (the microchunk) only takes a few minutes."
"Smart companies want to assemble the crowds with the most sophisticated knowledge about their business problems to maximize the impact of the small percentage of idea generators within them."
Watch out for the lowest-common denominator, or an information cascade, especially with regards to voting. Winners of voting can merely be the ideas that most people agreed on, not necessarily the best one. And voting itself can easily be influenced by others if results are made visible during the process.
Voting, however, is a nebulous arena since it is a function that the crowds do best (e.g. "American Idol has produced highly successful artists"). "When a company like John Fluevog Boots & Shoes asks its fans to submit and vote on new shoe designs - that is a model based on the wisdom of crowds. The wisdom of the mass is more likely to identify a winner than a select few."
Making Your Site User-Friendly and Ready-to-Wear
As mentioned in Part 2 of this series, one of the reasons these types of crowdsourcing projects fail is the platform. Programming and usability of the site should be a priority and should be solid before you get the crowds involved. You don't want them to arrive, find nothing there or discover that the site is difficult to navigate and leave. We've all left a site because it was impossible to navigate, or simply boring. Poor site planning or construction will be infinitely more destructive when you need the crowd to be completely engaged, stick around and come back often.
"The most important piece of advice I can give you about this is to make your site/product/software useful even if you only have one user. People keep trying to create sites where users can share all this stuff, but unless you build a critical mass, the site isn't very useful."
Anyone contributing their time needs to feel that they are getting something in return. Contributors will only stick around and continue to participate if this is the case.
The topic/project needs to inspire passion in its contributors: Michael Sikorsky of Cambrian House notes that "[CH's] members care less about money than they do about meaning. Their labor has to hold meaning for them."
Speaking about crowdsourcing project, Assignment Zero, Jeff Howe observed: "What the interviews make clear is that contributors volunteered to tackle subjects about which they were passionate and knowledgeable. In this they held a considerable advantage over professionals, who often must complete interviews with little time (or inclination) for advance research."
Further, in order to keep the crowd around, the arrangement has to be transparent and inspire trust "For the community to be truly engaged, it is extremely important for the company to be very transparent." You can keep the crowds posted and in the loop in many ways - one example would be Cambrian House's weekly updates and emails (Jasmine Antonick, VP of Communications, at a panel discussion).
Antonick also notes, "people will not work on something if they don't feel that they are gaining."
And, be absolutely certain to reward. If you want to inspire
loyalty it is imperative that you compensate people who work for you
(even those in the "webosphere") in one form or another. Mzinga's
Aaron Strout emphasizes that it is about partnering, rather than exploiting.
And your fairness will reap rewards in kind, because "[b]y providing rewards or incentives consistent with the value of the ideas being submitted, you can get greater participation from qualified users and a higher level of confidence in the quality of the ideas being submitted."
Accept the Loss of Control
Being "out there" is not always easy (e.g. I can hear the snide comments coming from my friends when I wrote that sentence). Similarly, when you ask for input, you may not always like what you hear.
"Crowdsourcing isn't for everyone, so make sure you have the fortitude required to make the effort pay off. Humility, a thick skin, and a receptive management culture are key prerequisites. Be prepared to see and hear some things you might not want to. The people who participate may really like your business or your product, but the way they articulate it may be very different than what you'd do yourself."
Leadership (without censoring)
Crowdsourcing is not all that democratic – it can't be. "Groups need leaders because they need direction." "...having an expert in place as a product manager can provide guard rails to keep things on track. The product manager can bring a single, unified vision and - this is critical - can communicate back to the community why a particular idea is not being used."
Collaboration in crowdsourcing is possible as long as there is essentially a leader or guiding force. Then people can vote and do whatever is required of them within that structure. Collaborations which involve almost complete consensus is very difficult to achieve, and when it is, that consensus averages out into a sort of blandness.
"Every authentic example of collective intelligence that I am aware of also shows how that collective was guided or inspired by well-meaning individuals. These people focused the collective and in some cases also corrected for some of the common hive mind failure modes." (Jaron Lanier)
Even when you open your world up to include the crowd, as Charles Leadbeater, author of the upcoming book "We-Think" writes, "It rarely works as a free-for-all. It requires some core norms and rules of behavior, but not many. It does require leadership but of a particular, open, conversational kind. It thrives on decentralized cooperation and people taking responsibility for working together. So it needs a leadership that makes the conditions for that possible."
These organizations need organization. Although "a self-policing community (possibly, with some moderation) can help weed out low-quality input and spam", it cannot all be turned over to the masses. A caveat: monitor and moderate gently: "Once you slow communities down they realize they are being censured or they're being interrupted, and their natural momentum begins to either slow or dissipate." (quote from an interview with Barry Libert)
Examples of Successful Crowdsourcing
- Lego Mindstorms NXT, Tremor, Threadless, Trendwatching, GoldCorp and Netflix’s contests
- Cambrian House
- Citizen journalism (such as Nowpublic, Gannett, and Muckraker's use of crowds)
- And others
The series concludes with "Concluding Thoughts" (Hey, I used up all my clever titles in the past two posts...). (Wanna prove crowdsourcing right? Suggest a better title by contacting me or submitting one below.)