Will the value of a network grow exponentially with time?
Metcalfe’s law: the value of a telecom network is proportional to the square of the number of connected users of the system.
Startups often feel that it justifies exponential growth projections, very high expected revenues in years 3-5, assuming that user base always grows.
This law also means that a small network has very low value. Users have little incentive to join as a result.
So while entrepreneurs hope to beat Facebook providing new or better quality and functionality of service, they forget that the value of their service is nothing compared to that of Facebook – its network has huge value.
Here is the catch: the value of the service would be great when they have 1,000,000 users, but when it is 100 – the value is zero, so it is not always clear why it will ever get to even 10,000 users.
Even Titans FailApple has over 160 million iTunes users and unlimited resources to market a network, still Ping had limited usage so far.
Google is used by even more people and also has unlimited resources, still Orkut has succeeded only in Brazil, and Buzz has not too many users.
In addition to being concerning facts in themselves for a “social” startup, it makes sense to ask: Why have they failed in social networks?
Many Users Do Not Equal A Network
Metcalfe’s law defines a potential value of the network, rather than actual. For big networks the
square of the users number becomes close to the maximum number of connections.
Metcalfe assumed a telecom network that has only one value – to connect, and all the connections are short term.
For a social network, where connections are long lasting, it is wrong to apply Metcalfe’s law. It is more precise to take the number of existing connections as the value of the network.
Photo: Eugene Stenly
It explains Apple’s and Google’s failures – their services values were dependent on network value, but as the networks values initially were very low (many users, but few connections), services didn’t work out as expected.
In addition to well established networks (Facebook, Twitter, Linkedin, etc.) there are so many new social networks and services, that the probability of success in this space is very low – we just have no time for them all.
Why Has Ping Had Limited Success?
Users expected Ping to be a useful experience and more than a million immediately subscribed.
But Ping, having no original connections between iTunes users, had little value and many users cooled off and stopped using it, also being unable to establish connections with many friends and artists, as they were not on the network yet.
If, for example, Apple had tried to make iTunes experience better for connected users, relying on existing networks (Facebook, Twitter), Ping could have been more successful.
Also Apple could have used “share with friend” history to build an initial network.
Three Roads to Success
Is success possible at all? It seems for some time it may be:
- “Thin edge” approach (Chris Dixon seems to be the first to use this term here)
- Network of service providers
- Find network, create a service
Thin edge/Single user approach
- Create users base. Provide a service (S) with utility independent of the users network – this service will be valuable for its first user, so the user base will grow. A good example – Prezi.com.
- Create network. When the user base is big enough, enhance the service, so that it becomes more useful for connected users (when and how to do it is the main question). Use existing networks first to implement social features.
- Leverage and capitalise network. When network has many active connections between users, implement services with utility dependent on the network.
Network of Service Providers
- Find a service that can be “crowd-sourced.”
- Build a system to facilitate transactions between providers and customers.
- Provide incentives and advantages for providers to sign up and to use this service, rather than sell independently.
Issue 1: there are many such networks already, such as elance.com, etc.
Issue 2: It is not a network yet, as connections between members are not used – it is only a service.
Find a Network, Create a Service
- Find network (maybe loose one) that already exists, but which value is underutilized and necessary services aren’t provided to this network.
- capitalise it by making connections stronger and by implementing services valuable to this network.
Photo: Eugene Stenly
Github.com is a good example:
- It provided services to networks of software developers, that existed before, but were loosely connected.
- It allowed new networks to appear.