Crowdsourcing: Why many heads are better than one
Since its first appearance in a Wired article back in 2006, the term ‘crowdsourcing’ has revolutionized the world of work. Crowdsourcing, loosely defined as an arrangement in which a company outsourced work previously done by employees to a network of people, in other words, a ‘crowd’, covers a variety of applications today ranging from attempts at tapping the collective wisdom of the crowd to make rapid scientific advances to mundane tasks like filling out consumer surveys. The idea of crowdsourcing has also led to new concepts in other fields, such as that of crowdfunding, where several people pitch in money for a project. (Read more about crowdfunding here.)
Different models of crowdsourcing trigger different kinds of reactions. While people generally recognize the benefits of bringing communities together to tackle complex problems, some forms of crowdsourcing, such as websites that help companies outsource their routine tasks to the public around the world, have come in for criticism because of low wages and evasion of employee benefits. Some critics even compare such services, like Amazon’s Mechanical Turk, to virtual sweatshops that can pay as low as $1.2 an hour.
Despite the controversies, crowdsourcing has its share supporters who focus on the bright side—improvement of efficiency in innovation and problem solving. Karim Lakhani, the Lumry Family Associate Professor of Business Administration at the Harvard Business School, is one of the leading experts on crowdsourcing in the world today. Having extensively studied the world of open source software, Lakhani is now looking at how organizations can utilize resources from outside using mechanisms like innovation contests. He is also the principle investigator of the Harvard-NASA Tournament Lab, a virtual research facility that hosts innovation competitions in search for solutions to real-world problems.
In this interview, Lakhani discusses different institutions within crowdsourcing and why some of them are going to change how companies and organizations innovate in the future.
Q. Is there a broad and narrow definition of crowdsourcing? What are the different models?
A. Crowdsourcing now encompasses, in the common press, four distinct institutions for innovating. We have communities where people work together to solve a joint problem; we have a set-up of a contest where people compete on external platforms; we have online labor markets—companies like oDesk and so on—that, in fact, allow people to bid upon jobs. And then there is another whole set of perspectives, which is complementors—you can create a platform where lots of other people can innovate on your platform. Apple’s app store is a good example. It is basically crowdsourcing innovation for Apple so that Apple doesn’t have to invest in any of the applications. People do it themselves and the marketplace decides. So that’s the four things.
More recently, there’re also been much interest in crowdfunding. What’s interesting about that is it’s taking the same principles and applying it to the financing of innovation, the financing of ideas, which is also very critical. It used to be that we needed a single source to finance our innovative ideas, and now what we do is we can distribute that task of financing to lots of people.
Q. Can you give me a few examples of how crowdsourcing can help create new ideas, which may be likely to happen in more traditional ways? Like within a company?
A. If you look at what’s happened in the world of software—over the last two decades, communities have become a very important component of the software system, the open source software world. People share knowledge and ideas, and create Linux, Apache, Perl, Hadoop… you name it. What’s interesting is that many of the very important companies like Apple, Amazon, Facebook, IBM and Lenovo, they all rely on the open source software infrastructure. The core technologies that have been developed through sustained collaborations by people from around the world basically enable our modern economies to function.
Another example is what we have done with the Harvard Medical School, where we’ve taken computational biology problems that are fairly tough, very significant, and we can solve them through running a contest on online platforms, like a company called TopCoder. What we are finding is that these approaches are faster and we can solve problems in a few weeks instead of months or years. They are cheaper, and also we can create objectively better solutions. In the world of genomics sequencing and sequence alignments, we’ve published a paper that basically shows that the solutions we got are orders of magnitude better than what exists in the state-of-the-art academic medical centers.
A third example, an interesting one, is in the world of toothpaste. There is a company that makes toothpaste and they were having issue with how fluoride was flowing through their various plants. The fluoride was getting hard to mix and also it was getting hard to flow to all the pipes. They were stuck for a long time and eventually they decided to post the problem on one of those online contests settings, saying, ‘These are the problems we are trying to solve and can somebody help us?’ They were all chemical engineers, so the solutions they came up with were chemical engineering based. But then a physicist looked at the problem and said, ‘Oh, the best way we can mix fluoride and store it in transition is to, in fact, create negative-positive charge in your systems and the fluoride will go completely into the mixer.’
Q. Are there bad examples as well? And how can companies manage crowdsourcing projects better?
A. There’re a few examples where they didn’t work out as well. The most recent one is the X Prize Foundation shut down a $10 million prize in genomic sequencing. It was a very big prize and was getting lots of attention. The reason was, in many ways, the prize was not enough. There was already ongoing investment in the field of genomic sequencing so the $10 million prize was not going to motivate anybody to do more work.
Similarly, when PepsiCo runs various crowdsourcing challenges for creating commercials, sometimes they get commercials that are not flattering for their brand or offensive to their people. That can create problems for the sponsor and you have to learn how to deal with that.
For companies that are consider using crowdsourcing, there needs to be clarity in the questions you are asking; there have to be appropriate ways to interact with the community and there is a proper balancing of risk and reward. So among the risks that people can take, their time, money, and mental capacity, there has to be a proper reward. People think that it is, using an American term, the magic pixie dust–you just add crowdsourcing and everything will work. And I said, ‘No, there’s actually a ton of effort required to make crowdsourcing successful for the firm itself.’ What you’ve put into it is what you’ll get out of it.
Q. In your point of view, which areas of crowdsourcing applications have the most potential?
A. Certainly the world of software. I think we can do more and more of it this way. And firms should do so both in terms of contests and communities. It’s clear that software can be done more effectively this way. And then in software there is also analytics, algorithms and so on, all the Big Data stuff can be done through crowdsourcing. I think a lot of scientific problem solving also. If you can take components and modules out there, you can have the world’s scientific communities’ help in solving your problems. An interesting thing nowadays is a big push on marketing. So, in a lot of platforms, there emerged and there were firms creating digital content, web content, TV commercials, and full-fledged campaigns through outsourcing as well. That’s a huge industry. So companies like Tango, Victors & Spoils, MOFILM… all those guys have created very robust platforms that allow people to come up with great answers.
Q. There are some issues or controversies with regard to the online crowdsourcing labor market. There are problems with companies paying less than minimum wage or hiring underage workers. Do you think there’s a legal issue here?
A. Nobody is forced to work on these projects. It’s voluntary self-selection that drives participation. There is no real push for somebody to be forced to work on these conditions. Typically the system relies on people saying, ‘I think that I have the skills and the ability, and I want this reward for this work’. And the reward may or may not come depending on how many people submit and how good their work is as well. People know what the game is. All these are voluntary, and people can participate on their own wish or they can choose not to participate. The second thing is that people participate for many reasons. Certainly the monetary reward matters but people also participate for the sheer fun and problem-solving elements. They actually enjoy the process and they can meet other people who are likeminded. Even if you might not win, you might be able to create a reputation that is valuable both in the communities you participate in but also in the labor market. In terms of underage workers, I haven’t heard much about that. I don’t know what the driver is. Typically firms will say, ‘You have to be over a certain age to participate. If you are not, you need to ask the permission of your parents.’ So, that’s how it works.
Q. So you are saying that there shouldn’t be extra influences from the lawmakers or authorities, and we should just let the market decide?
A. Crowdsourcing relies on the market system so if there are people who don’t want the work, and there isn’t a big demand, then the work won’t happen, right? So the market will have to adjust and the prices will have to be raised. The fact is that we actually have a global system, and firms around the world can use this global system to solve their problems.
To be quite honest, I’m not sure what additional thing you could do to this. The whole point is that we are trying to bring efficiency into the marketplace. We are trying to create, to meet supply with demand, and you want firms and individuals to both solve their problems, important problems. So, let’s use the most effective way to do that.
Q. For the creative work, how do you deal with intellectual property (IP)?
A. There are typically two issues related to IP. One is who owns IP for the work that’s done? Secondly, what about competitive secrets and so forth? On the first side, it’s literally a matter of contracts, whoever the sponsor is of the community or crowd sets the rules for the game. And again people know what the rules are, so people know the license for this kind of software is going to be X, Y or Z. In many contest settings, the agreement is that you submit your proposal and you get evaluated. If the sponsor likes the solution, they will pay you the agreed-upon price amount and they will own the IP rights. It’s fairly straightforward.
A lot of people get worried about ‘if I post a problem, my competitors will know what I’m working on.’ I think that’s a legitimate concern. What we are finding is that the benefits of speed and effectiveness in problem-solving far outweigh any kind of a downside of your problems being revealed. And also often times what we are discovering is that the questions being posed are so specific and often devoid of the context that you can effectively mask the question in plain sight. I think the concern is legitimate and firms should pay attention to it, but I don’t think it’s a deal breaker.
Q. So how would crowdsourcing reshape companies or organizations in the future?
A. The modern corporation, the modern form of organizing companies got established literally in the 1900s, where you have a limited liability corporation, and you have a division of labor set up within a firm structure. That form of organizing human beings—in a company collectively working to create products and services, has done amazing things for the world economy. Now we are in the cast of a new era. Instead of saying, ‘Let’s create four walls, let’s bring people inside the walls, and work together inside the walls’, there’s now this view that we can create this network organization—as much as it’s inside, it is also outside. You can’t start a company—any kind of technology company or science company—without thinking about what is going to be your open strategy. So this notion that a company can do it all within its own four walls is going away. I think that’s exciting. I think we see it most clearly in the online world, in the digital world. This is fast approaching to physical products too, where you can use 3D printing and so on to bring online approaches to the physical world.
Q. The modern China is usually perceived as less innovative and full of copycats. Do you think crowdsourcing will help change that picture?
A. Yes, I think ‘copying’ isn’t all that bad in the first place. A big part of innovation is learning from others. You learn from others, you modify and you improve on those things. Throughout history, a lot of countries have relied upon the technology of others to bootstrap and then to rise up. I mean yes, we have to do it legally, and we have to be within the bounds of intellectual property laws and so on.
In the past, you needed to be in a foreign country, you needed to interact with people and learn from them in order to be more innovative. I think what the web world has done is that it collapsed that distance. You can now, in the world of software, participate in global communities and learn from the best. That’s actually unprecedented. It has, in fact, spilled over to China. The Chinese will learn how to organize themselves in this way and figure out what the right ways are to work online and to create online businesses. It can’t just be the US that does all the innovation. We want the Chinese, and we want the Indians to be doing it as well.
(Homepage image courtesy: cambodia4kidsorg via Flickr Creative Commons)