Winner Takes All: Case Studies in How Online Marketplaces Are Creating Modern Monopolies

The following is an excerpt from “Winner Takes All: Case Studies in How Online Marketplaces Are Creating Modern Monopolies” by Shirish Nadkarni.

According to Wikipedia, “A network effect (also called network externality or demand-side economies of scale) is the effect described in economics and business that an additional user of a good or service has on the value of that product to others. When a network effect is present, the value of a product or service increases according to the number of others using it.” We have seen many examples of how powerful network effects can be in creating unstoppable, winner-take-all juggernauts. As we discussed earlier in the book, eBay was an early entrant in the auctions market in the US. Many later entrants like Yahoo and Amazon failed to take on eBay because of the strong network effects it had established. The buyers on eBay came because it had a critical mass of sellers and sellers came because it had the largest audience of buyers available. Unlike eBay which charged for listings, Yahoo and Amazon made the listings free of charge. Even then, the sellers stuck with eBay’s platform as they saw more sales transpire through eBay. Similarly, companies like Airbnb and Instacart have established dominant positions in their market because of the strong network effects they have.

While network effects can be very powerful when the network has achieved a critical mass of users and suppliers, it is very hard to establish early network effects. Generally, marketplaces have gotten going by first acquiring a critical mass of suppliers e.g. as DoorDash did by signing up restaurants with a promise of creating new demand for these restaurants. Another strategy that is sometime possible is to sell products to suppliers that provide value on their own. A great example of this strategy is OpenTable which sold a customer and reservation management system to restaurants without promising that it would drive customer demand to these restaurants. Its only when OpenTable had signed up enough customers that it created its own destination site that drove reservations for its restaurant customer base.

Not all network effects are as strong or global as we have seen with eBay and Airbnb. Some network effects can be fairly weak or localized. Let’s explore some of the nuances about the strength of network effects.

Multi-tenanting

Multi-tenanting is practice of buyers and/or sellers to utilize multiple networks at the same time. If it is easy to multi-tenant and to utilize multiple networks at the same time, it becomes much harder to establish strong network effects. Food delivery apps are a good example of a category where multi-tenanting is common. It is relatively easy and advantageous for restaurants to participate on multiple platforms as they have done to date. Similarly, it is easy for users to try out different food delivery apps especially as many of them were doing a “$10 off your first order” promotion to get users. As a result, the food delivery market has not been a winner take all market with multiple major players like DoorDash, Grubhub and UberEats.

Multi-tenanting is also a problem in the ride sharing market. Many drivers log on simultaneously to Uber and Lyft and take whoever provides a ride first. As a result, both companies have spent

an enormous amount of money trying to acquire drivers exclusively to their platform. Again, like food sharing apps, the ride sharing market has not been a winner take all market.

Local Network Effects

In many situations, the network effects are not national or global in scope as we have seen with eBay and Airbnb. The network effects have to be established market by market. OfferUp is a good example of this phenomenon. It originally launched in Seattle where it proved that the model worked and was able to raise funding that allowed it to expand to other markets. The company also kept a relatively low profile in the press so as not to attract competitors in markets where it has not yet entered. Over time, OfferUp was able to attract large funding rounds that enabled it to rapidly expand to all the key markets in the US before other players had the opportunity to enter these markets. However, if multiple players had entered different local markets it would have been much harder for OfferUp to create nationwide network effects.

The same phenomena applies internationally where you often see copycat competitors establish themselves before you have a chance to enter that market. In many situations, the US company generally ends up acquiring a foreign competitor to enter the market if it determines that network effects in the local market are too hard to overcome. Given the larger market size in the US and the easier access to venture capital in the US, it becomes relatively easy for the US company to make the acquisition of a smaller foreign player.

Network Effects Throttling

Not all network effects show an ever-increasing strength as the number of players increase. They can get throttled or even reverse depending on the product. For example, with ride sharing there are initially strong network effects as drivers and users join the platform. However, once the market has been saturated with drivers so that the wait time is 5 minutes or less it doesn’t really help if more drivers join the platform. As a result, in the US, Uber hasn’t been able to kill off Lyft as it has also been successful in attracting enough drivers to its platform to establish a viable service in the US.

Similarly, Facebook is a great example of a network with very strong network effects that have dissipated or even reversed to some extent. Most teenagers and young adults no longer use Facebook and instead use Snapchat as their social networking platform. The reason is that with Facebook, the teenagers have been forced to accept their parents, uncles/aunts and grandparents as friends. Teenagers are not very comfortable sharing their private moments with their relatives whereas they are much more likely to do so on Snapchat which is mainly populated with young teens.

Commoditized vs Differentiated Supply

It is a lot easier to establish strong network effects if you have a differentiated supply as opposed to a commoditized supply. Ride sharing is an example of commoditized supply where as a consumer you don’t really care who your driver is – you are trusting the app to make the right selection for you. As a result, once the network reaches a certain level of liquidity (e.g., 5-minute wait times or less), the network effects no longer accrue further.

On the other hand, if you are providing a long tail of differentiated supply as, for example, with the Amazon marketplace or with Airbnb, the network effects continue to accrue. The greater the supply, the more likely that there is strong price competition among the suppliers which makes the marketplace even more attractive to consumers. Today you can go to the Amazon marketplace and find virtually any item making the marketplace value proposition very strong for consumers.