The Power of Network Effects for Start-Ups

Since 1994, over 70% of the value created by technology companies is attributed to companies built around network effects (according to a report by Andreessen Horowitz, a16z).

Network effects occur when a product or service becomes increasingly valuable as its user base expands. Having been harnessed by companies like Google, AirBnB, and Facebook, every founder looking to build a company these days should understand them. Espcecially as they help carve out a competitive edge that practically monopolises billion-dollar markets.

What Exactly Is a Network Effect?

Under network effects, a product or service becomes increasingly valuable as its user base expands.

Let's consider the telephone. When only one person has a telephone, it's practically useless — there's no one to call, no one to receive messages from. The device's intrinsic value is practically zero since the whole purpose of the telephone, communication, can't be fulfilled. However, the moment a second telephone enters the network, the value of each telephone increases significantly. Now, there are two people who can communicate, making the initial investment in the telephone worthwhile.

As more and more people adopt telephones, the network's value continues to expand exponentially. Each new user who joins the network enhances the value of the network for all existing users because they're potential new people to call or receive calls from. The larger the network grows, the more valuable it becomes to its users. This creates a powerful feedback loop that accelerates the network's growth and consolidates its competitive advantage. This phenomenon highlights the growth and resilience of social networks and platforms that leverage network effects. As a second mover struggles to build a moat, your existing, value-creating user base gives you a significant competitive advantage.

A Bit of Science

Robert Metcalfe, the inventor of Ethernet and the co-founder of 3Com, contributed a foundational paper on the network effect. His work on the value of networks, commonly referred to as "Metcalfe's Law," posits that the value of a network is proportional to the square of the number of its users.

The paper is titled "Metcalfe's Law after 40 Years of Ethernet" (IEEE Computer, Dec. 2013). Metcalfe and his co-authors argue that the value of a network grows quadratically as a function of its size. While it's not a study in the empirical sense, it forms a theoretical underpinning of how the value per user in a network can be seen as a squared function.

For empirical evidence, a more recent study titled "Metcalfe's Law is Wrong" by Bob Briscoe, Andrew Odlyzko, and Benjamin Tilly (IEEE Spectrum, July 2006) found that network value doesn't increase quadratically with size,

   n²,

but rather in a linearithmic scale:

   n*log(n).

Even with this lower rate of increase, each additional user increases the value of the network more than the last. This supports the idea that value per user grows over time.

Implications of Network Effects: Unleashing a Virtuous Cycle

Rapid Growth and Scalability

Start-ups that capitalize on network effects can achieve rapid growth. This is due to the fact that each new user might bring additional users onto the platform. This cascading effect can fuel scalability, accelerating the start-up's growth trajectory and fostering a vibrant and engaged user community.

Competitive Advantage and Barriers to Entry

Network effects can erect formidable barriers to entry. This makes it challenging for competitors to gain a foothold or disrupt the market. The inherent structure of network effects creates a 'winner-takes-most' dynamic where the market leader, benefiting from a larger user base, can provide a significantly superior value proposition compared to rivals.

Increased Value for Users

As the user base expands, the value of the product or service increases for each user, leading to enhanced customer satisfaction and retention rates. The relationship between user base growth and increased value is not merely linear but exhibits the characteristics of a quadratic function. This phenomenon underscores the strategic importance of customer acquisition and retention in businesses that leverage network effects.

Are There Downsides to the Network Effect?

While advantageous, there is a significant downside of network effects. That is, their susceptibility to collapse if the number of users decreases. 

As more users join, the network gains value, but losing users can trigger a negative cycle, diminishing overall benefits and discouraging new users. This downward spiral can render the network obsolete or unviable, making user retention and adaptability crucial for start-ups relying on network effects.

The Way Forward

In light of the transformative potential of network effects, it is crucial for entrepreneurs and investors to identify and understand these dynamics in their early stages. 

Network effects can offer start-ups an unfair advantage, catalyzing rapid growth, fostering competitive moats, and increasing user value. Recognizing these effects in your venture or the companies you invest in can be a game-changer, potentially unlocking unprecedented levels of growth and value creation.

Despite the current wave of hype surrounding technologies like generative AI, its future success isn't clear. I’m personally most excited about companies that combine these new technologies with clearly calculable network effects.

In AI companies, that effect is typically reached through data collection. In addition to improving your performance, more training data will make your product more valuable. This means more users will become involved, more training data will be generated, and so on.

When I talk to venture capitalists about AI companies, some of them look at them as if they were all alike. Though, I don’t believe this is the case. By combining this knowledge with an understanding of AI, we can determine which ventures will scale and which will not.

I hope this article was helpful for you. As a founder or investor, pick business models that scale, irrespective of the technology you are passionate about.