Digging deeper into how platforms are conceived, used and then managed, it is worth asking do they always imply a two-sided market? You might also ask does it matter? The answer is yes. We need to understand what defines a platform. Nick and I have emphasised the transaction engine. But is the emphasis misplaced.
Those markets are generally understood as a refinement of network effects. A network effect happens, for example, when a company like Facebook grows its member base. Each member will benefit if the platform has more members, i.e. more friends. The increased utility of a social network lies in knowing more people in it – to a point.
In the case of many software markets, recruiting new users is delegated to existing users (e.g. DropBox used double incentive marketing to reward its users for bringing in new users – however it had to reward both the existing user and the new user, which is why the technique is called double incentive).
This needs less incentivizing when the addition of users makes everyone’s experience better (e.g. as it does on Skype). Participation needs to be incentivized when the network effect is low (as in DropBox).
Clearly not all platforms offer network effects. Yet the network effect can have a profound effect on how business models are designed – it is cheaper to market them.
Some platforms benefit from an indirect network effect:
AirBnB users benefit when there are more apartments; Uber users benefit when there are more taxis. These types of services have grown as match-pairing technology improved and it might be the ability to match needs rather than a network effect that makes them successful.
This is substantially different, however, from Skype where each user needs another user, not another Skype or service provider.
GrabCad members do need other members but not in a truly imperative sense. Having access to other members lies in the category of “very nice to have” rather than “must-have”. The counter parties truly add value without causing dependency.
Here are Parker & Van Alstyne’s examples of two-sided markets:
” credit cards, composed of cardholders and merchants; HMOs (patients and doctors); operating systems (end-users and developers), travel reservation services (travelers and airlines); yellow pages (advertisers and consumers); video game consoles (gamers and game developers); and communication networks, such as the Internet.”
The network effect on AirBnB and Uber looks weak in comparison. Indeed even these examples are indirect rather than direct, with the exception of credit cards. There is a point where any market can claim a network effect if the definition is diluted. Cars are platforms for parents and children? The catwalk is a platform for attraction and repulsion?
In many instances, though, the new platform is enabling productive activity. In the case of GrabCad, the objective is to share design drawings in order to reduce the development time of new products. In open source platforms a similar productive activity is taking place.
In these cases too the platform has a network effect but is not bound by a two-sided market. It is bound instead by a moral framework. The only way to achieve the moral objectives of the community is to recruit ore members and evangelize it use.
In cases like App Stores, which are the epitome of the modern business platform, the network effect is actually negligible. There is no compulsion for anybody to buy an iPhone because of the apps on it. On the other hand the availability of thousands of apps makes it possible to achieve two ends:
The first is an advocacy community for the iPhone and Apps Stores
The second is extreme personalisation for every user.
Platforms have many characteristics that make it difficult to nail down one definition. In unusual cases they have network effects. In some cases they are advocacy communities that propel a platform symbiotically with its advocates. In other cases they are bound by a shared moral objective. Each of these gives a clue to how to engineer participation.
Here are some other examples.
The last column in the table below refers to the degree to which an actual exchange is a focal point of the platform. Exchange can be a high or low focus or it can be direct or indirect.
“Multiparty” is a new development in platform environments, beyond two-sided markets. For example Apple’s struggle with Health Kit is going to be making it work for multiple-sided markets including hospitals, family doctors, labs, devices, monitoring services, analytics and users. In that sense the market itself becomes ecosystem-like and poses new degrees of complexity on marketing and coordination.
|PLATFORM||Network effect||Two-sided market||Multi-sided market||Exchange|
I’ll come back with some more thoughts in the next few days. Here is a definition provided by Gawer, taken from Parker & Van Alstyne:
“Industry platforms are products, services or technologies that are developed by one or several firms, and which serve as foundations upon which other firms can build complementary products, services or technologies.”
I don’t find this enlightening but I do think it illustrates the difficulty of finding a good definition.