01 Jun 2013

Collective Wisdom

Crowdsourcing for innovation, efficiency, and profit


Faculty Q&A

Crowdsourcing—it's a relatively new word for the centuries-old approach of opening a challenge to outside communities. The difference today, according to Karim R. Lakhani, is that thanks to the Internet, crowdsourcing has moved into new and unexpected industries and organizations, establishing itself as a mainstream strategy for innovating and gaining a competitive edge. Lakhani, the Lumry Family Associate Professor of Business Administration, is coeditor of Perspectives on Free and Open Source Software and has published cases on organizations that practice crowdsourcing, including Data.gov, TopCoder, OpenIDEO, Google, Threadless, and Wikipedia. He serves as principal investigator of the NASA Tournament Lab at Harvard, which helps the space agency in applying open innovation approaches to big data and computational challenges, and works with Harvard Catalyst, the University-wide initiative led by Harvard Medical School.

What are some examples of different kinds of crowdsourcing?

One well-known format is community-based. The idea is that there are some problems that require collective input and cumulative knowledge-building. Open-source communities are a very powerful way to get that done. We used to think that software had to be developed by lots of engineers working under a strict managerial structure. Open-source software communities have shown that we can have a highly distributed model involving contributors from all over the world. Some people said this would never work. But open source often creates products that are better than those engineered within companies. A few of the biggest beneficiaries of open-source software have been companies like IBM, Google, Facebook, Twitter, and even Apple.

Talk about the role of contests.

This kind of crowdsourcing has a strong historical precedent. In the 15th century, Filippo Brunelleschi won a competition to determine who would design the dome for the Florence Cathedral. Even the design of the HBS campus began as an architecture competition. Contests work well when we don't know what the right approach may be to a problem. Today they're used in a variety of situations, from creating logos to software code.

"Spot labor" markets are also a form of crowdsourcing, correct?

Yes, you can post any sort of task—whether it's editing a document, providing website copy, or verifying directory entries—and people with the abilities and time available will bid on it. Companies like Elance and oDesk make it possible to find a global labor supply to do a job accurately and cost-effectively.

What about the fear of losing the ability to monetize intellectual property (IP)?

Those are valid concerns. It really depends on the crowdsourcing approach you use. With a contest, the IP is yours for the price of the prize. In a community-based setting, you give up IP in exchange for complementary assets that create value for customers. Then there's the fear that my competitors will find out what I'm working on. Open companies have adopted a nuanced attitude that embraces transparency as a way to attract crowds and gain energy. A company like SAP has realized that relying on communities is a way to build support for features that customers will value, instead of blindly pushing out solutions and hoping for buy-in. It's a matter of firms realizing that they can benefit more from executing faster and better than they can behind a veil of secrecy.

But how do you manage a crowd?

It sounds oxymoronic, doesn't it? It's obviously a very different control process when you have people who you rely on for solutions, yet they don't report to you. Plus, they can enter or leave the process at any time. So far, it's a challenge that we've seen managers approach intuitively. It's definitely one of the questions I'm looking at in my research.

Your work seems to run across so many fields. You're collaborating with NASA, for example, and researchers at Harvard Medical School.

My research is focused on understanding how communities and contests can be better designed and managed to solve real-world challenges. My colleagues and I at the lab work closely with our partners to design field experiments that advance academic knowledge and create practical managerially relevant insights. At NASA, we're leveraging the knowledge of people from around the world to solve big data challenges; for example, algorithms that optimize power generation on the International Space Station or that detect the location of craters on the moon.

At Harvard Catalyst we are pushing the frontiers and thinking around the design of academic research programs and initiatives. For example, we launched an open and global "ideas challenge" to generate research topics around type 1 diabetes, with $30,000 in awards. Winning ideas came from diverse actors, including an earth scientist, a college senior, and a retired dentist, among others. Opening up participation to nontraditional actors resulted in the sort of truly novel perspectives that can lead to breakthroughs in an established field of research. Often the complaint in academia is that our research is trapped in journals that no one reads. It's exciting to produce work that matters while also moving the needle on health care.


Post a Comment