Platform businesses are disrupting industries left and right, from Airbnb's redefinition of the hospitality business to the titanic effect of Amazon on all manner of retail. How do traditional businesses get in on the action?
In this segment of a six-part SearchCIO video series, Choudary answers the question of who needs to be involved in getting a platform business off the ground. The first step is identifying the three components of successful platform businesses: a network of producers and consumers, the infrastructure to govern that network and killer data science.
Who needs to be involved when getting platform businesses off the ground?
Sangeet Paul Choudary: If you think of the three components of a platform business model, very often organizations mimic these three components.
- You have a part of the organization that brings the network together. Bringing the network together is not just a simple marketing problem. It's an incentive/design problem. You need to ensure that you have the right incentives so that both producers and consumers are encouraged to participate. It's also a community management problem, and at the end of the day, it's also a behavior/design problem. You're often creating entirely new behaviors. Before Airbnb, we never stayed at a stranger's house, and now we do. So there's a new behavior creation involved as well.
- There's an organization that's involved in creating the infrastructure, which has two parts. There's the technology part of it, which we tend to understand very well because that's what we see in a platform. But then there's also the governance part of it -- the rules of the platform that need to be set in place. And governance involves community governance but also data governance. How can the data be used? What are the do's and don'ts around the data? What's the terms and conditions involved? All of these things are part of the platform organization.
- The final part of it is data science, where again there are two elements involved. One is data acquisition. When a platform starts, it starts without data. Before the platform can actually create value, it needs to start getting data into the system. That might mean connecting with other existing sources of data. It might mean creating incentives for existing users to bring in the data. All of that is the prerogative of a part of the organization that focuses on data acquisition. The other part of it is data intelligence: How do we take all of this data from the ecosystem and use that to make the platform more intelligent over time, so we're not just waiting for the ecosystem to participate, but the platform can actually figure out patterns and enable participation based on those patterns?