Executive Thought Leadership

The Network as the Platform

An Interview with Information Economics Expert Marshall Van Alstyne

July 2007

Dr. Marshall Van Alstyne is a prominent researcher in the area of "information economics." Van Alstyne, an associate professor at Boston University and a visiting professor at the Massachusetts Institute of Technology (MIT), recently met with Cisco Director Roger Farnsworth to discuss the concept of the network as the platform. Excerpts from their conversation follow, edited for length and clarity.

FARNSWORTH: Can you briefly define the concept of a platform as it pertains to networking and information technology?

VAN ALSTYNE: A "platform" represents the components used in common, across a product family, that also exhibit network effects.

FARNSWORTH: What are the attributes of a successful platform?

VAN ALSTYNE: You need a certain amount of openness, extensibility, and modularity. Another requirement for success is an excellent peer review or other quality-control system. This becomes particularly important as multiple parties begin contributing to the enhancement of the platform. Quality control represents the role of integration, where someone selects best-of-breed parts and tests them to make sure that they play well together.

FARNSWORTH: Who are the beneficiaries of a platform evolution?

VAN ALSTYNE: There are three beneficiaries – consumers, third party developers, and the platform sponsor – but the benefits of evolution happen in phases.

Initially, there is competition to establish a platform. We see that today, for example, in the competition between DVD and Blu-ray Disc (BD) technologies [for the storage of high-definition video and other optical media]. There are few benefits for the losers and you have to move quickly to try to get a standard adopted. Once that happens, developers can build products that work with the standard platform and then consumers can begin realizing its benefits. Over time, competition shifts to development on top of a given platform. Consumers and the platform sponsor both benefit.

FARNSWORTH: How is a platform transition different from normal technological advancement?

VAN ALSTYNE: With normal technological advancement, you tend to see incremental improvement over time—short-term enhancements. But for a platform transition, you see the traditional S-curve growth. As folks converge around a platform, the networking effect takes off and moves the platform forward.

But a lot of mistakes get made when people see this exponential growth continuing indefinitely. At some point, it tapers off, and you have to figure out what's going to be the next layer on top of the existing platform. At platform maturity, there is even risk of a different technology offering a new and difficult-to-contest kind of competition. We see this now in the threat that traditional phone service faces from VoIP.

FARNSWORTH: How is innovation affected or best managed when platforms take hold?

VAN ALSTYNE: Imagine learning to run in place. For a platform, you have to continually look at how to evolve and add new components to the system. The mechanism for doing that is, again, openness, so that others can build to the platform. As competition broadens for particular components, you fold those high value but low margin elements into the platform itself. This then becomes the next common layer for innovation to be built on.

FARNSWORTH: Is the network's emergence as a platform becoming more evident?

VAN ALSTYNE: I believe that it is. What's interesting is that as the network platform becomes more important, it also becomes a bigger part of the technology we take for granted. It's because those systems work—because companies are making sure their investments are solid and extensible—that you can take the technology for granted, and then you can add the next layer to it. The irony is that the very success of the network platform makes it so ubiquitous that it also renders it less noticeable.

FARNSWORTH: Are the concepts of Web 2.0 service-oriented architecture and software as a service just beginning?

VAN ALSTYNE: Let's look at Web 2.0 in a couple of different ways. One of them is software as a service; another is as user-generated content. In terms of software as a service, I think we are going to see the continuous opening up of platform elements such that different people add different features through the APIs. And you're going to need ways to integrate them, so you need consistent service to provide that integration. And you can charge for that value.

But the other element of Web 2.0 comprises user-generated content services such as YouTube and Flickr. This content requires a backbone and appropriate systems in place to make it accessible. As the cost of transactions drops, it becomes possible for content, services, and ideas to be generated across the marketplace and be provided across the network backbone.

FARNSWORTH: How does a technology make the leap to platform status?

VAN ALSTYNE: That's a tricky question. I want to highlight the difference between what would be a traditional technology, with a fairly linear set of supply-chain relationships, and a platform, with a more triangular set of relationships between the platform sponsor, the co-developers that contribute to the platform, and the users.

Making the technology-to-platform transition requires connecting these parties with each other and not just through you; rather, you create an ecosystem among them. So you can think of a platform as having a triangular set of relationships that you're managing simultaneously as opposed to a linear supply chain that you're managing sequentially.

FARNSWORTH: Is a more open technology more likely succeed as a platform?

VAN ALSTYNE: We've done some theoretical modeling that suggests you cannot get elements of growth if you don't open the platform at least at some level. The tricky part is getting the right balance. If you open it completely, you limit what you can offer and charge for uniquely. You absolutely need openness, but you also need something that is your core nugget value, your novelty for which you should be duly compensated. So, yes, it must be open but you also need to know there are limits.

FARNSWORTH: That being said, what types of licensing and partnerships are most appropriate to platform environments?

VAN ALSTYNE: The answer really depends on your ability to identify the key partners and players in your environment. Traditional licenses and partnership arrangements should be used for the situations where you can identify the best players. But as you develop a more open ecosystem, you want to move in the direction of more open licensing.

A lot of the time you want players with the best ideas to be able to self-identify. In an economic sense, this poses an interesting problem. An innovator may have a really clever idea, yet might not want to give it up when coming to you to negotiate the license. So you need an open licensing regime where innovators can pursue their ideas on top of your platform without having to go through too much negotiation.

Lawyers often make a classic mistake in that maximum protection does not always equal maximum value. You need to find a happy medium where you open up platform licensing enough to get innovative contributions back.

FARNSWORTH: Is the title of "platform" one that can be taken, or must it be bestowed?

VAN ALSTYNE: They can label it a platform but, you know, you can label it a duck and it may quack like a duck but it may still smell like something else.

I think that a lot of technologies would aspire to the level of platform. But in many cases, they won't actually get there, particularly if they can't engender participation by third-party co-developers in addition to bringing the users on board.

FARNSWORTH: How can the network's emergence as the platform help the transition to an interactions economy?

VAN ALSTYNE: Again, if we look at the platform as the basis of this triangular relationship between the platform sponsor, co-developers, and the users, you're really facilitating interactions among multiple parties, as well as interactions between each of them and the platform. You need a platform, an open licensing arrangement, and an interactions economy to get that. For an interactions economy, you need to help others interact with each other, not just interact with you.

FARNSWORTH: What overall economic effect is the network's emergence as a platform likely to have?

VAN ALSTYNE: I believe the economic effects ought to be huge. If you can get incremental 3rd party innovation on top of the platform, not just your own, you create enormous wealth among those in the ecosystem triad. New wealth is created, not just transferred from someplace else.

FARNSWORTH: As a follow on, are the economic effects of platform evolution likely to be more micro or macro in scale? Is a platform more likely to have a global or local impact economically?

VAN ALSTYNE: I think you will observe, in some sense, micro, macro, and global effects. Micro effects will be simple interactions and access to new features and attributes that users wouldn't otherwise get without third-party co-developers providing them. The macro effects are that you're getting more players involved in the system.

Then, people you haven't even identified as potential partners can join the ecosystem. At that point, the platform becomes much more of a global entrepreneurial ecosystem in which multiple parties can participate independent of the usual corporate or geographic boundaries.

FARNSWORTH: What about managing intellectual property?

VAN ALSTYNE: This is not just a technology question. It's also an information and ownership rights question. And you need to balance the ownership rights and the technology and the wealth that's created across the product space.

Products are becoming increasingly not just technology, but technology plus information. Physical goods are delivered with digital goods; what is the nature of the content? What are the product attributes? What consumer groups are interested? How do you manage tastes and preferences? What are the privacy and security issues. All of this is information. And if you put the property rights in the right places, you motivate the people to give you the benefits of their investments, their insights, and their complementary products and designs. If you hold the intellectual property rights too close-to-the-vest, people will keep the benefits of their complementary information for themselves.

So, in closing, I'd encourage the sponsors of a platform to consider both how they manage the technology and how they manage the information on top of that technology.