Network effects 1

This article was initially published in Viktor Kyosev’s (COO of Greenhouse) weekly newsletter, where he shares his views on the topic of startups, growth and fast-growth markets:

In my previous two newsletters, I focused on defining marketplaces and assessing what’s next for such models. Consequently, I touched on the concept of network effects as that’s one of the most essential concepts in business and, in particular, when it comes to tech businesses and marketplaces.

Today, I would like to discuss further network effects, how to recognize them, and why such dynamics can give businesses moats i.e. competitive advantage.

In a nutshell, a network effect occurs when a product or service becomes more valuable to its users as more people are using it.

Let’s take a familiar case and look at Airbnb being the most common example in my previous few posts.

What gives Airbnb a unique network effect is an interplay between supply and demand. Guests become hosts, and hosts also become guests; thus, the flywheel spins on and on.

Yet, network effects are not always present from the beginning in tech businesses.

The chart below illustrates how the network effect did not get triggered until the company received a critical mass of transactions.

In the early days, Airbnb focused on building features for the demand side because supply will always go where the demand is. While traditional forms of marketing e.g. branding, SEM, paid social, are important and can help, trust and safety are paramount in a marketplace.

Here I would like to mention that network effects are not available in marketplace models only. A study by NFX discovered that network effects drive roughly 70% of value in tech.

To name a few examples, the Big 5 – Apple, Google, Microsoft, Facebook, and Amazon are in the top 10 of the world’s most valuable businesses. Airbnb is now worth more than Hilton. Uber is worth more than GM. All of them are network effect businesses.

The study assessed 336 companies between 1994 (approximately the time when the internet was already widely available) and 2017, and the main criteria were that each tech company should have reached a valuation of more than $1 billion.

It turns out that having a network effect is the single most predictable attribute of the highest value technology companies — other than perhaps “having a great CEO.”

To drive the point home, I looked at Facebook because while it resembles a marketplace model (users and advertisers), it’s actually quite different, yet it demonstrates a strong network effect.

It’s important to point out how the Facebook product had inherent virality, it grew from one user to another as an organic consequence of use.

  • DAU – daily active users | MAU – monthly active users

Early on, Facebook identified that connecting a new user to 10 friends within 14 days of sign up was critical to improving retention. As a result, they used email contact imports, suggested friends’ features, and embedded widgets to drive engagement.

With time they optimized timelines, introduced the like button, relationship statuses, which increased the number of usage and triggered a network effect, where the more people use Facebook the more valuable network it becomes. After all, would you use the product daily if you could not connect with so many important people in your life?

Here it makes sense to emphasize what is not a network effect. It is commonly believed that if your product has some sort of virality that automatically illustrates a network effect, but scale and growth are not necessary network effects. The main difference being that viral growth increases just the speed of adoption, but it does not make a product/service more valuable. Another way to look at the concepts is to consider how virality is all about speed and growth, while network effects represent value, engagement, and retention.

As explained in Metcalfe’s Law:

the value of a telecommunications network is proportional to the square of the number of connected users of the system (n2)