Recently, Uber's business practices had me a little worried. Several issues seemed to converge: Uber was trying hard to edge out its established competition. Uber was expanding quite quickly internationally, attempting to become a market-dominating first mover. Uber had far more cash on hand than its competition, a full $1.5B in funding, which it could use to lobby against opposing regulations, and competitive pricing so as to undercut its competition. Considering these factors, it seemed as if Uber was well-poised to effectively become a monopoly. I found this prospect troublesome, because Uber's aggressive competitive tactics (e.g. sabotaging competitors like Lyft by requesting and cancelling rides) did not inspire consumer-level trust: once a monopoly for this near-essential service is established, there's nothing that would stop Uber from raising prices.

Naively, my first response was to support the competition. I decided to stop using Uber, and to instead use Lyft, which hadn't been accused of unethical business practices, and generally seemed a great deal more friendly toward the consumer. However, this was not practically feasible at peak times, as Lyft operated on a smaller capacity than Uber. So I reevaluated my position, and investigated how likely Uber actually is to become that worrisome monopoly.

In short: I think Uber will not be able to maintain a dominant market-share in the ridesharing space. This space will become divided among many ridesharing services. Here's why:

• The ridesharing space does not allow brands to differentiate themselves in a meaningful way. The vast majority of drivers I've spoken to^[0] drive for both Uber and Lyft, and simply attempt to maximize their profits. They see no difference between the services. I, as a consumer, see no difference between the services either. They are basically isomorphic. Some brands that are very similar manage to nonetheless attract significant loyalty: for example, see Coke and Pepsi. However, both of these drinks deliver consistent experiences. Rides are not consistent in this way, because there is significant variance among ride experiences because of different drivers, cars, etc. It is very unlikely for a user to have the same driver multiple times, and this variability makes it difficult for these companies to enforce consistency on customer experience.


• Thus, it seems to be a safe assumption that consumers will pick their rides simply by price. There is no obstacle for the consumer to doing this: similar to websites that try to find the cheapest flights, there are now websites and apps that will find the cheapest ride. Examples of this are Corral and whatsthefare. I expect meta-rideshares to emerge in the future that will take the user's location and destination, query all possible rideshare services and book the cheapest for the user, without the user having to do any more work than currently required to request an Uber.^[1]
  
  Similarly, I expect drivers to sign up with all rideshare companies in order to maximize their load, and to be most inclined to give rides through the best-paying ridesharing services. Perhaps meta-services analogous to those for users will emerge for drivers.


• The ridesharing space does not offer advantages to economies of scale. This is primarily because there's very little actual infrastructure involved. It could be argued that some of the more prominent services will have more drivers and more potential customers, which yields a positive-feedback loop, but this effect is unlikely to prevail in the long run, as meta-rideshares begin to emerge.


• All ridesharing services are still middlemen: they simply connect drivers with users, and take a commission. Uber currently takes a cut that is between 20 and 30 percent.^[2] The past twenty years have clearly shown that these types of businesses are easily targeted by new competitors that attempt to either significantly undercut the middleman, or to eliminate the middleman altogether. The large commission leaves an obvious opportunity for a price war between competitors.


• The barriers to entry into this market are actually very low. A key realization is that all rideshare companies operate on very simple technologies. The Uber and Lyft apps are not very sophisticated — their components that are essential to operation seem to comprise the following and not much more:
  
  
  • Payment processing
  
  
  • Maps and GPS tracking interface
  
  
  • Network messaging
  
  
  • User database
  
  
  I would wager that a team of ten reasonably competent engineers could put together a similar app in at most two months. Beyond that, some employees are needed for customer support and some are needed in major cities for screening potential drivers. Granted, more employees are needed as such a startup becomes larger, but it is quite clear that there are few large initial overhead costs. Given this and the large potential reward of starting a ridesharing service,^[3] it seems clear that we'll see more rideshare services start up.


• The only nonnegligible barrier to entry into this market takes the form of regulation. Uber is aggressively battling regulation in order to keep expanding, but it is unlikely that regulations will be amended exclusively for Uber: changes will likely legitimize all ridesharing services. Thus, this barrier can be expected to fall.


• Despite the low barriers to entry and the relative technological simplicity of rideshare services, Uber does have one nontrivial product: a review system. The element of trust is important, and possibly poses the greatest challenge to new ridesharing services as they emerge. Maybe good screening policies will ensure that the drivers of emerging services are trustworthy. Perhaps reviews of drivers will become shared across services as governments realize that the traditional taxi industry has no future, and scramble to institute minimal regulations on ridesharing services. I should note that the issue of trust will not be a direct obstacle to new services, but rather a factor that will comparatively slow their growth. This delays, but does not prevent, the fragmentation of the ridesharing market.


• One particularly important counterargument is that some ridesharing service may, by some innovative means, differentiate its product significantly from that of its competitors, and thereby capture a dominant market share. For example, Uber and Lyft are beginning to introduce carpooling, which looks to be the next big step in the industry. However, this innovation is simple for their competitors to copy. Indeed, it is difficult to conceive of a market-dominating innovation that is not feasible for others to replicate and does not involve either the creation or entering of a completely different market.

I have attempted to lay out the reasons why the rideshare market will likely become fragmented between competitors, implying that Uber will cease to hold a dominant market share within the next few years.

This does not imply that Uber's current valuation ($18B) is mistaken. The global taxi industry is notoriously hard to value, with conventional estimates often around $50B, but it is worth noting that people who previously did not take taxis are now taking Ubers. Some people are leaving their cars at home, and others are taking Ubers rather than public transit, due to a continuously diminishing price differential. Some people even use Uber to shuttle their kids to school. It is difficult to predict where this trend will go, but I would not be surprised if the rideshare market turns out to be much more valuable than the global taxi industry, which means that even if Uber were to only control some fraction of the market, it could still justify its current valuation.

[0] I've been asking Lyft and Uber drivers about their experiences when I get rides from them. I have informally collected data from approximately 40 drivers in New York, New Jersey, and Chicago in the past two months.

[1] A comparison would be to flight search engines: there exist websites that will, given a set of dates and departure/arrival cities, query many airlines, and return the cheapest flights. These are commonly used by consumers in lieu of manually researching flights.

[2] I do not know if this is consistent across different geographic areas — in New Jersey, this cut appears to currently be 31%, and I've heard of other amounts elsewhere.

[3] At least now, while the profit margins are still large.