I travel a fair bit, and I’m cheap, so I compare Uber and Lyft prices on most trips.  Do you?  (Let me know in the comments.)

How Many Consumers Engage in Comparison Shopping Uber Vs. Lyft?

According to an excellent recent research paper entitled “LEAVING MONEY ON THE DASHBOARD: PRICE DISPERSION AND SEARCH FRICTIONS ON UBER AND LYFT,” only about 16% of us compare prices between Uber and Lyft for the same ride.  (The authors suspect it might be even less than 16% due to limitations of their methodology.  But they also overlook that when I’m traveling with my wife, we may use two different devices to check prices, rather than one.  Yes, I have “forced” my saint-like wife into this lifestyle–not “the lifestyle”!)

How Much Can You Save by Comparison Shopping Uber Vs. Lyft?

Suppose you do engage in this price comparison exercise. In that case, you inevitably learn that, even though this market is a duopoly with multihoming by riders and drivers, there is still substantial price dispersion.   The market is not that efficient. In the study, the authors found an average price difference of $3.50 (14%) between Uber and Lyft on identical routes in NYC.  (Anecdotally, I see even greater dispersion across cities worldwide.)  The authors estimate “that New York City-based rideshare customers collectively leave over $300 million on the table due to a lack of price comparison.”  That’s $300 million in extra profits for the rideshare companies, driven by consumer price insensitivity and friction.

Other fun findings from the study include:

• Longer trips seem to have the same percentage price difference, making searching even more valuable.
• Neither rideshare app was consistently more expensive than the other in NYC, so searching made sense rather than relying on one service to be consistently cheaper.
• The study excluded personalized or promotional discounts. Interestingly, though, the Uber algorithm was triggered when a new account opened and started searching for rides, but never took one!  It offered hefty promotional discounts presumably to get this apparent new rider hooked on its service.  This suggests to me that checking prices might have a secondary, beneficial effect of generating even more promotional offers.
• Differences in wait time did not explain the differences in prices.  In fact, in the study, longer wait times were slightly correlated with higher prices, not lower ones.
• The percentage of consumers who comparison shop does not appear to differ much by income or age group.  There must be a cheap/smart-shopping gene.

An App to Make Comparison Shopping Uber Vs. Lyft Easier?

In researching this post, I came across an app called Obi that claims to make it easier to compare prices across rideshare services.  I wondered when this would happen and how someone would get around Uber’s API terms of use, which prohibit the use of its API to:

aggregate Uber’s data with competitors’ data; or (g) parse or scrape any of Uber’s data; in each case other than as explicitly permitted by Uber in writing. You will not share with a third party (or enable a third party to use) any operational, technical or other data obtained through the use of the Uber API Services in any manner that is competitive to Uber, including, without limitation, in connection with any application, website or other product or service that also includes, features, endorses, or otherwise supports in any way a third party that provides services competitive to Uber’s products and services.

Obi may be evading the Uber terms of use by having riders link their Uber and Lyft accounts to the application, which may mean it is acting on our behalf, but I’m not a lawyer.  (ChatGPT says it also has other ways to help it avoid the terms of use.)   Still, if you want to try the service, I’d suggest doing so soon, in case Uber decides to go after it.  (The last innovative rideshare company, Empower, I wrote about has been in big trouble with the city government!)

The Larger Implications

As someone fascinated by two-sided markets, platform economics, and antitrust considerations, this little study reminds us, as the authors point out:

even small sources of friction, such as app design, information frictions, or hassle costs, are sufficient to sustain substantial price differences and limit competitive pressure. These patterns imply that the modest search barriers in ridesharing markets are profitable for platforms but reduce potential gains to consumers.

It’s yet another reminder of behavioral economics at work and a reminder of how inefficient markets can be as a result.