top of page
  • Writer's picturewright.economics

Seven questions to evaluate network effect moats

By now, every investor and entrepreneur on the planet is aware that network effects, when present, can be a great source of defensibility and unfair competitive advantage. However, many investors and entrepreneurs don’t go beyond superficial statements such as “startup X has compelling long-term defensibility via marketplace network effects” (actual quote). That’s a problem because there is huge variance in terms of how weak or how strong network effects are, and that ultimately determines whether or not they create a defensible moat.

We have put together a list of questions that we use when evaluating investment opportunities with network effects. These questions are designed to probe as deeply as possible for the strength of network effects and associated defensible moat. In what follows we share those questions, each of them followed by a brief explanation of the logic behind it.

For the sake of completeness, it is useful to remember that a product or service exhibits network effects when the value to a user increases in the number of other users that buy the same product or use the same service. Network effects lead to defensibility by creating a virtuous cycle: the more users join, the more attractive joining becomes to other users, leading to more users to join, and so on. This covers both “same-side network effects”, where all users are essentially the same from the provider’s perspective (e.g. users on social networks) and “cross-side network effects”, where there are two or more distinct customer groups that interact/transact with one another (e.g. buyers and suppliers on marketplaces; users and app developers on iOS or Android). Most of our questions below apply to both of these types of network effects.

Here goes the list:

1. To what extent do users really care about how many other users also buy the same product or service?

This is of course the most basic question one should ask in order to establish whether meaningful network effects exist in the first place. It can be helpful in ruling out cases where network effects only play a marginal role, if any. For instance, some have tried to argue that WeWork creates network effects because “the more tenants that rent space in a WeWork location, the greater the opportunities become for networking and meeting like-minded people”. Even if that were true, it would be at best a very modest consideration relative to the nature of the space, the location and the lease terms. And answering this question clearly can help avoid the still common confusion between network effects and virality. GoPro has virality (when people share videos filmed on GoPro, it increases awareness of GoPro) but that doesn’t give it network effects which would require a person’s value from using a GoPro increases with the number of others using a GoPro. Finally, this question is also helpful in articulating the nature and “depth” of network effects. For instance, do they come from the possibility of discovering new matches (e.g. finding a driver on Grab or Uber), from making transactions/interactions easier (e.g. buying a product from an eBay supplier and having it shipped), from enabling communications between users (e.g. WeChat), from providing information to other users (e.g. ratings and reviews of hosts on Airbnb), or some other source of network benefit?

2. How quickly does the extra value created for users diminish as additional users are added?

Needless to say, network effects are stronger when their marginal value decreases more slowly. The 1,000,000th user on Facebook still provided a lot of additional value to other users, as opposed to the 1,000th English language tutor on an online tutoring marketplace like Preply.

3. For marketplaces, do buyers view suppliers as distinct/differentiated, or do they view them as interchangeable providers of the same product or service?

All other things equal, network effects are stronger when suppliers are differentiated in the eyes of buyers (e.g. Airbnb or Cameo) than when they are largely commoditized (e.g. hopps or Uber). The reason is that when additional suppliers contribute variety to a marketplace, they provide more value to users since there is a greater chance that users can discover their ideal match. This is why a key feature of many successful marketplaces (e.g. eBay, Etsy, TaskRabbit, Upwork, Youtube) is that they enable a long-tail of suppliers to thrive. And this provides defensibility since otherwise it is easy for competing marketplaces to start and draw from the same pool of undifferentiated suppliers.

4. Is the network effect global (e.g. Airbnb, Upwork) or local (e.g. Task Rabbit, Uber)?

All other things equal, truly global network effects (users care about users in all locations) provide significantly more defensibility than local ones (users only care about users in the same geographic location) for obvious reasons. It is a lot harder to become a dominant platform at the national or international level in an environment with local network effects, which typically allow multiple competitors to start in and dominate different locations.

The four questions above are aimed at evaluating the strength of network effects. The next three questions are aimed at figuring out to what extent those network effects actually translate into defensibility.

5. How difficult is it for buyers/suppliers/users to “multihome” (i.e. be active on multiple competing platforms)?

Obviously, network effects do not provide much defensibility when users can easily multihome (e.g. GoJek and Grab in South East Asia, Lyft and Uber in the US). Factors that make multihoming less likely include the monetary and non-monetary costs of joining (e.g. buying and learning how to use an iPhone, inputting information and pictures on Instagram), the propensity to transact/interact repeatedly with existing connections (e.g. Facebook, WeChat), the importance of reputation/ratings (e.g. Airbnb, Amazon, Etsy), the complexity of managing schedules/transactions on multiple platforms due to capacity constraints (e.g. Airbnb, OpenTable), and so on.

6. How easy is it for users to coordinate their adoption decisions?

In some cases (e.g. Clubhouse, Evite, Houseparty), users who interact can easily coordinate on where to connect: the organizer of a call or event sends an invite to all participants (e.g. via email or twitter). This is bad news for the defensibility of network effects since those users can just as easily coordinate on a competing product/service that has slightly better features.

7. Does the matching of buyers to suppliers (or between users) have to be synchronous or can it be asynchronous?

All other things equal, if the scheduling of the suppliers must be synchronized to match buyers’ demand, so the marketplace has to create sufficient liquidity at any given time for it to be useful, this will make it more defensible. The fact that ride-sharing platforms like Lyft and Uber need sufficient liquidity to keep wait times low makes it harder for a new entrant to take them on compared to if rides were only those scheduled well in advance.

A final word of caution is that even when network effects are strong and defensible (based on the list of criteria/questions above), that may in some cases not be enough to guarantee a highly profitable business. For example, a marketplace creating strong network effects by matching buyers and suppliers may have trouble extracting a lot of revenue if the two sides are inclined to take their transactions off the platform after discovering each other on the platform. We discuss this issue (known as dis-intermediation or leakage) in depth in a series of substack posts starting with this post.


Commenting has been turned off.
bottom of page