Insights On Network Effects Using A 2 x 2 Matrix

Context

A 2 by 2 matrix is a great problem-solving tool (Read this book to know more!).

It uses 2 carefully-chosen parameters in order to segment a situation into 4 parts. Often this breaking down yields surprising results. Example: BCG’s Growth-Share matrix.

Let us try this tool for analyzing Network Effects.

What could be the 2 right parameters for our 2 by 2?

I would choose the following: Switching Costs and Critical Mass.

Read this article by Andreesen Horowitz to know about Critical Mass.

Below is a quote from the article:

Critical mass is present in a platform business if a sufficient number of users adopt an innovation in a system so that adoption of those innovations becomes self-reinforcing. The classic example people cite to illustrate this point was the competition between the Beta and VHS videocassette formats in the 1970s. Beta was considered to be the superior format in terms of quality but VHS reached critical mass first.

Now, the criteria for choosing the 2 parameters is: each parameter should be a good clustering variable. It should be able to divide up the whole situation into 2 parts such that each part is different from the other.

A network effect business that has high switching costs will be a very different animal than that with low switching costs. So, this parameter is an interesting one.

A network effect business that has high critical mass requirements will be a very different animal than that with low critical mass requirements. So, this parameter is also an interesting one.

There may be better parameters still. But I will go ahead with these 2 for now.

Equilibrium outcomes, using a 2 by 2 for Network Effects

Let me now explain each cell.

High Switching Costs + Low Critical Mass: In such a market Low Critical Mass implies that a player can start seeing network effects at low scale too. This could be for instance when the marketplace is hyper-local/local. A cab-aggregator is an example (because it can just offer its services in NCR region of India and still benefit from local network effects). Now, if this condition gets combined with High Switching Costs (e.g. by charging upfront membership fee), then we have a 1st mover advantage market with multiple player in different niches. Since Critical Mass is low, multiple players will be able to sustainably target multiple market niches (e.g. different cities) even with limited funds. At present I have no example in this category.

High Switching Costs + High Critical Mass: In such a market High Critical Mass implies that a player will start seeing network effects only at high scale. This will mean that only players with deep pockets will be able to build the network. Now, if this condition gets combined with High Switching Costs (e.g. by charging upfront membership fee), then we have the holy grail of network effects: a winner-takes-all market. MS Office has been an example of this until recently.

Low Switching Costs + Low Critical Mass: In such a market Low Critical Mass implies that a player can start seeing network effects at low scale too. Multiple small players will be able to enter the different niches of the market with limited funds. Now, if this condition gets combined with Low Switching Costs, then each player will strive to identify and defend its own niche (perhaps through superior customer service and value-add). We will have a competitive market with multiple profitable niches. Examples: Online classifieds business (will get split into verticals as what is happening with Craigslist, Cab-aggregation (will get over-powered by localized players, Hotel-aggregation business (AirBnb is already getting hit by niche portals).

Low Switching Costs + High Critical Mass: In such a market High Critical Mass implies that a player will start seeing network effects only at high scale. This will mean that only players with deep pockets will be able to build the network. Now, if this condition gets combined with Low Switching Costs, then each player will strive to snatch away other’s network! This will lead to very intense and unprofitable competition! Examples: eCommerce business (will keep on competing each other through discounts, until one of them has no more cash to burn).

Key Insight

Not all Network Effects create a winner-take-all market. Only those with high Switching Costs and high Critical Mass do!

1