What is Crowdsourcing?

Jeff Howe’s Definition

Jeff Howe defines crowdsourcing as “the act of taking a job traditionally performed by a designated agent (usually an employee) and outsourcing it to an undefined, generally large group of people in the form of an open call.”  Mr. Howe coined the term “Crowdsourcing”, so he probably has the right to define it, but I’m not giving him that right. :)

The Scope of Crowdsourcing

This definition is too business-centric to capture how dramatically crowdsourcing will change how we work. Implicit in this definition is that there is a business, who has a job that it wants done, and who is making “an open call”. The changes that are starting are much, much bigger than this. Crowdsourcing is actually distancing the business from the “employee” more and more, and this trend will continue.

Open Source Software has already given us a model for how this works in practice. No one is “behind” Linux or MySQL. The community owns and runs these projects, without a corporation there to organize and reward their efforts.

There is still room for business in this new world, but they are enablers instead of enforcers. Google has crowdsourced advertising creation, selling, and buying, to the point that there are now tens of thousands of people who make their living through AdWords and AdSense, but would never consider themselves Google employees. Crowdsourcing companies will be moved more and more into the background – companies would be wise to see themselves as facilitators who help people connect to each other in easier ways. EBay and iStockPhoto and crowdSpring are all platforms for interaction, instead of the source or destination of what is actually produced.

My Definition

If I had to define crowdsourcing, my definition would be something like:

Crowdsourcing is when new technology makes it possible for a function to move from hierarchical systems like companies to dispersed and self-directed communities.

For Love Or Money

Photo by thomashawk on Flickr

In America, most of us have at least some control over the kind of job that we have. We choose, to a lesser or greater extent, what we do for a living. Because of this choice, I would argue that for all of us, work is a mix of love and money – if we could find a job we loved, we would be willing to do it for less for more money, and we would probably be willing to do a job that we really hated, for a price.

Most hobbies earn us little or no money. We might play in a local band or read books or take pictures. We do these things because we enjoy doing them.

Other people get paid, however, to play in bands and review books and take pictures. One of the things that makes crowdsourcing so powerful is that it makes it possible to get paid to do things that you love.

In the past, if you wanted to make money as a photographer, you would need to create a portfolio, go out and find clients, or get hired at an agency. There was a lot of overhead that went along with that decision. Now, you can just upload some photos to a microstock website, and get paid. Without doing any work.

Because crowdsourcing lets more people do what they love, the overall prices in the marketplace will go down. When you bring love into the equation, you don’t need as much money (this maxim also applies to relationships :)).

When Crowdsourcing Works

Photo by sheepbackcabin on Flickr

Crowdsourcing is not a panacea, and cannot be applied in every situation. While I believe that some elements of crowdsourcing will affect almost every industry, the rate and extent of its influence will be far from uniform. So far, the industries where crowdsourced production have worked share a number of traits in common. We can expect that future successes will come in similar industries.

The common traits of “crowdsourcable” industries include:

• Clearly defined objectives
• Dispersed workers
• Quality-based output
• Hobbyist participation

Quality-based payment

Crowdsourcing works best in a Merit Economy. When workers can be paid for the quality of their output, the overall cost of production goes down, without sacrificing quality. Stock photography is a prime example of this. When a photographer is paid based solely on the number of times an image is downloaded, the overhead disappears, and consumer prices plummet.

Clearly Defined Objectives

The objective and the reward must be clearly defined. Because crowd workers can choose what they want to work on, they will choose to pass on projects that are ambiguous, or where they fear being taken advantage of. This forces businesses to be very explicit in their expectations, and the reward tied to those expectations. When Netflix announced the Netflix Prize, the parameters were very clear, as was the reward for meeting those parameters.

Location Independent Workers

Crowdsourcing has seen the most success when the activities are not location-based or synchronous. If I am looking for a stock photo or reading a Wikipedia article, it doesn’t matter to me whether the photo was taken by someone here or in India, nor does it matter whether the article was written yesterday or last week (in general). Crowdsourcing works best when users can work independently, but as part of a greater networked community.

Open Source software is a great example of this principle. Most software is broken into modules, that can be worked on independently, before being combined together into the finished product.

Hobbyist Participation

So far, the two primary reasons that crowdsourcing has been able to reduce costs are: first, that it eliminates barriers to entry, letting people participate without needing to be hired, managed, fired, etc. Second, and more importantly, crowdsourcing lets people make money participating in things that would do for free. By allowing hobbyists to become part of the workforce, crowdsourcing dramatically increases the supply of workers, and workers who are doing something for love, with money as an added bonus. Hobby photographers would take pictures no matter what – selling them on iStock is a nice perk. Most bloggers write because they enjoy writing – if their blog makes a few dollars through ads, that’s great.

These hobbyists are destroying industries. A million hobbyist photographers will take more and better pictures at a much lower price than 1,000 or 10,000 full-time professionals. A million Wikipedia authors will create a higher-quality, more comprehensive encyclopedia than 1,000 full-time researchers, for free. A million bilinguals could translate the world’s content faster, cheaper, and better than all of the current professional translators.

Did I miss some traits? Let me know in the comments. ↓

The Merit Economy

For most of us, our compensation isn’t directly correlated with our performance. As long as we perform above a certain level, we can expect positive yearly reviews and modest raises. Some days are productive and beneficial to the company, while others are not as productive, but we aren’t paid less for our unproductive days.

In some occupations, results are directly tied to compensation. Farmers aren’t paid for how many seeds they plant, but how much they harvest. Authors aren’t paid for how many words they write, but how many books they sell. I am calling labor exchanges like that, where compensation is based on success, “Merit Economies”.

Creating Merit Economies

While some occupations will never become Merit Economies (Do we really want firefighters to be paid based on the fires they put out?), the Internet has allowed a number of companies create Merit Economies for certain occupations. In every industry where this has happened, the results have been dramatic and predictable: more people participate as producers but make less money, prices fall for consumers, and the profits of legacy participants erode. I want to give a few examples of where this is starting, and then conclude with some occupations that I think are ripe for this kind of change.

Examples

First, stock photography. iStockPhoto upended their industry by allowing amateurs to upload and sell their photos. Instead of vetting every photo that wasn’t of the highest quality, iStockPhoto let photo buyers make their own decisions about what they wanted to purchase. Photographers were then paid only for the photos that were successfully sold. Stock photo prices were slashed, as thousands of new participants entered the stock photo profession, most of them as a hobby or part-time job. In the traditional model, the best and the worst professional photographers all make about the same amount of money. In the iStock Merit Economy, there is a much larger discrepancy – a lot of people make a little bit of money, while only a few people make a lot of money.

The Netflix Prize. On October 2, 2006 Netflix announced that they would pay $1,000,000 to anyone who could improve their movie-recommendation algorithm by 10% or more. Less than 3 years later, they had their algorithm, and happily paid the $1,000,000 to the “BellKor’s Pragmatic Chaos” team. For $1,000,000, Netflix could have hired a team of PhD’s to improve their algorithm. That might have worked, or it might not have worked, but either way Netflix would be out $1M. By tying compensation directly to results, Neflix motivated the right people, and ensured that they would receive value for their investment.

Journalism. The other example that I want to point to is journalism. Internet blogging has decapitated the journalism industry, as people “pay” (via viewing advertising) only to read the news that is interesting to them, instead of paying for an entire newspaper. Just like with stock photography or the Netflix scientists, the best journalists can make more money in a Merit Economy, but most journalists are making much less. The key is that the consumer is still receiving quality journalism, just from more sources, and at a lower price.

Upcoming Merit Economies

Translators. Translators are generally paid by the number of words that they translate, not on the quality of those words. In the current model, a translator’s work is reviewed by 2-3 reviewers, all of whom are paid for the entire document. By gathering multiple translations for the same sentence, and then voting on which translation is correct, translation purchasers could get high-quality translations while only paying the translators who produced the good translations. (Disclaimer: I am currently working on a system like this for Lingotek)

Educators. Instead of paying all teachers the same amount, we could reward effective teachers, and give them a broader audience. With the ability to broadcast things over the Internet, why aren’t most of our kids taught Algebra by the best math teachers in the country, or taught Shakespeare by the best English teachers? By increasing the audience, we can decrease the cost of education, while increasing the amount that we pay our best educators.

What Merit Economies did I miss, and what other occupations could become Merit Economies? Let me know in the comments.

Is Crowdsourcing a Curse Word?

Josh Berkus wrote an interesting article today about how much he hates the word “crowdsourcing”. Here’s the crux of his argument:

By simply using it, the speaker is setting up an elitist dichotomy, since the speaker is never part of the “crowd”. They are part of the “elite”, instead, supposedly. It’s “us” vs. “them” language, and in a very classist way: we are the professionals, they are the unskilled; we are the visionaries and they are the dullards; we are the owners and they are the unwashed masses; we are the educated and they are the stupid; we are the strong and they are the weak. It’s the language of inequality.

While I disagree with Josh’s belief that the word “crowdsourcing” has undergone this sort of pejoration, there is no doubt that some businesspeople see the crowd as the “unwashed masses”. Many promote crowdsourcing as just another way to cut costs. They believe that it is a way to get your customers (read: suckers) to do things for free.

Those who believe this are wrong in two fundamental ways.

Crowdsourcing Isn’t Free

First, crowdsourcing isn’t free. Finding and supporting a “crowd” requires time and patience and money. Providing the resources and people to support and nurture a community can be more expensive than traditional models.

Even more importantly, crowdsourcing requires a shift in the way that businesses operate. Organizations need to realize that when they invite the crowd to participate, the crowd brings expectations of openness and transparency. Businesses that try to take advantage of their community without becoming more open and transparent will not get far.

Viva la Revolución

But businesses have much, much more than that to be worried about. Crowdsourcing is the beginning of a revolution. Just like the Industrial Revolution changed the mode of production from a rural, agrarian small businesses to urban corporations, crowdsourcing technologies will turn our current system on its head.

Instead of a hierarchical, top-down model, companies will be molded and driven by their employees. The line between employee and customer will become more and more blurred, as consumers contribute more and more to the products, and the direction, of the company. One can imagine a day when companies don’t need executive teams or boards, when the community is the company. Today’s corporations should be worried.