[Editor's note: This is an OpEd piece by David Gal, an assistant professor of management at Northwestern University]
Last year, Facebook chief executive Mark Zuckerberg suggested that a PC operating system was the inspiration for Facebook’s new “Platform.” With Platform, anyone could write applications for Facebook. Facebook’s in-house applications would get no special treatment, he declared.
The analogy to an operating system is appealing. For many years Microsoft’s Windows operating system has benefited from the large number of applications written by outside developers. People buy Windows, not necessarily because it is the best operating system, but because it has the most applications. Like Microsoft, Facebook does not have a monopoly on great ideas nor unlimited bandwidth, and a platform ostensibly allows Facebook to leverage the talents of the entire developer community to its benefit.
However, Facebook’s most important strategic asset is not its community of developers but its network of users. Does Platform build or leverage this strategic asset? I believe Platform was intended to leverage the network, but, in fact, it squanders it, by fragmenting the network across a large number of applications in each application category. Enter a search on Facebook for any category of application, from dating to chess, and you are faced with a large number of choices.
But aren’t more applications, as the Microsoft analogy suggests, a good thing? Abstracting lessons through analogy can yield unexpected insights, but analogies can be stretched too far. A large number of applications is a benefit to Microsoft Windows, because being able to run the largest number of applications is an important source of value for users of an operating system. However, the most important source of value for users of a networked application is the number of relevant users using the same application. When multiple applications compete for users in a network, then fragmentation results, and an important source of value is frittered away.
Won’t Darwinian selection ensure that the best application in each category will prevail? Maybe, but probably not. Darwinian selection does not work so well in a networked world. Multiple competing applications may coexist in a category, leading to diminished overall adoption of the category. Moreover, those applications that come to dominate a category will not necessarily be the best or the best-maintained; rather, they will likely be the first to attract a large number of users in a particular category. This is because in a networked application, having a large number of users may outweigh any functional benefits that a competing application can provide.
To conclude, Facebook’s focus should be on building, maintaining, and leveraging its core strategic asset, its network, and Platform is not well-suited to these goals.
First, platform does little to build the network, because while a large number of applications is an important reason for people to buy Windows, a large number of relevant users is the most important reason for people to join Facebook. In fact, rather than attracting users to Facebook, most Facebook applications acquire users through Facebook (and then get to split revenues with Facebook).
Second, useful applications are important to maintain the network, but the most useful applications are those that have the largest number of relevant users, that are well-maintained, and that do not require users to share their personal data with third-party developers. Platform applications are unlikely to meet these criteria.
Lastly, leveraging the network implies using that asset to pursue applications from which Facebook can derive revenue, such as dating, jobs, fantasy sports, event-ticketing, auctions, etc. By relegating these applications to Platform and focusing on advertising revenue through impressions, the user-base is fractured, and the value of the network is squandered.
VB's working with marketing expert Scott Brinker to understand the new digital marketing organization. Help us out by answering a few questions
, and we'll help you out with the data.