The Ride Share Economy
I needed a ride into Manhattan Saturday evening and while scanning the different ride-share apps noticed that UberPOOL fares were remarkably cheap; $4.53 from Williamsburg to East Village — more than reasonable compared to a $2.75 subway ride in the cold. I soon found out why. Uber, beginning in September, now automatically pools with as many people they can fit in the car, up to 3 separate parties. When the third passenger climbed into the car, I turned to the driver who scowled and shook his head. Drivers were now forced to all rides Uber assigned them. Fares were cheaper because we were sharing a single ride with more and more people.
In the sharing economy where distributors (Uber) have been able to commoditize suppliers (drivers), ride-share apps are essentially competing to provide the best product for the consumer (rider). Given that the product — drivers — are usually the same for all apps as the same drivers are servicing all the apps, the differentiator essentially comes down to price. Many apps now exist to compete through differentiated pricing — Via exclusively offers ride-sharing at set prices, Gett is optimized for peak hours with no-surge — however, the market leader is unquestionably Uber who has been able to strategically anticipate optimal products to suit the market before anyone else and undercut its competitors.
A key to Uber’s success has been its huge wealth of data — not only data about rider behavior, but crucially, pricing. Uber captures capture the time, place, price, and surge factor of each ride, but also every occasion where a customer declines the offered price or chooses one Uber service over another. What this has given Uber is a huge database of supply and demand curves. As Bill Gurley, on the board at Uber and partner at Benchmark, writes:
“Demand for its transportation services is HIGHLY elastic. As the company achieved lower and lower per-ride price points, the demand for rides increased dramatically… The “math department” and management realized that if they could increase driver utilization (the number of rides per hour for a driver), then they could lower the price for the end user while maintaining earnings quality for the driver. Higher efficiencies through higher volumes and better algorithms could help deliver the desired lower price points and better cash flows. Interestingly, these lower price points would lead to more demand, even more liquidity, an even higher utilization, and then another incremental price decrease”
It’s frightening how much of a market leader Uber has become in transportation such that it has a high enough volume of riders to make UberPOOL possible, even with 3 passengers per pool — how many rides per second must Uber be getting in order to pool 3 separate passengers who coincidentally happen to be traveling in the same direction at the same time, not to mention the processing scale required to compute the required matching and pricing algorithms?
Uber is slowly moving towards CEO Travis Kalanick’s vision for “the Perpetual Ride” whereby a driver will always have a customer in the car. UberPOOL’s rationale is ostensibly simple: lower the price for the rider through maximizing driver utilization, while allow drivers to maximize driving time, hence maximize profits, for both the driver and Uber. Increase the number of rides per hour for a driver while lowering cost to the end user. As a business strategy, this sounds highly reasonable. The math and economics seems to work out.
For riders, the inconvenience of sitting in a car with strangers is often justifiable enough to save up to 50% on the cost of an UberX ride. Fares are so cheap that it often makes sense to take an UberPOOL over taking the subway. In the last 3 months, I’ve spent $15.70 on NYC Taxis. In comparison, I’ve spent $266.45 on ride-share apps, half of which went to Uber. I wouldn’t normally spend that much on cabs, but the low price point of Uber makes more sense and convenience than taking alternative forms of transportation, i.e., the subway. The vision of Kalanick has always been, “it’s not about the market that exists, it’s about the market we’re creating,” and that ambition has translated case-in-point. Uber is no longer about stealing market share from the taxi industry, but rather, the industry of all transportation — private cars, taxis, public transportation, and even car ownership. Even people who own cars often take Ubers because it’s cheaper to get an Uber than pay for parking. However, at whose expense are we able to enjoy these low price points?
I never really thought deeply as to how Uber compensates drivers beyond articles decrying how little they made. After researching ‘The Ride Share Guy’ Blog, a blog written by a full-time Uber driver in LA, I began to understand more specifics. The driver is paid a set rate for the pool ride from pick up of first passenger to drop off of last passenger. Regardless of what price the passenger is charged, the driver will receive a fee equal to the time and distance the full trip took. The rate, however, is lower than the UberX pay rate ($0.75/mile vs $0.85/mile in SF). This justification is based on the fact that UberPOOL provides drivers continuously longer trips. An empty car makes the driver nothing — hence it is still worth it for drivers to drive with passengers in a car at a lower rate, vs none at all. If the trip is unmatched, the driver is still paid the set rate for the trip, but at the lower UberPOOL rate. It is presumably in the driver’s interest for the trip to pool as the extra miles and time going out of the way to pick up additional passengers equates to more money. The driver does not, however, make any additional money for the duration of the trip that is common to all pooled passengers — that is, drivers are paid a flat rate from A to B, no matter whether they carried 1 person or 4 passengers in their car for any part of the trip.
Here is where Uber really is mining their monetization strategy — Uber make more commission from pooled rides than UberX — significantly more, because it is able to charge passengers multiple times for the same trip, yet pays the driver the same rate regardless of how many people are pooled. UberPOOL pricing is notoriously unpublicized, for both the rider and the driver, but here is what a fare looked like for a real driver in SF, contributing to ‘The Ride Share Guy’ who compared his earnings to the price charged to his passengers (that he asked them for during the ride).
This is a large increase from the commission Uber normally takes from UberX, which is approximately 20%. Had Pax 1 opted for an UberX, Uber would have taken a smaller commission, probably closer in the range of $3 for the one trip alone. However, under UberX, the driver would have made slightly less — he or she would have earned only the trip covered for Pax 1 at a slightly higher rate, but would have missed out on the incremental distance added by traveling further for Pax 2 and 3.
UberPOOL purports to be in the drivers’ self-interest as it encourages maximal driver utility, but the reality is more complex than that. A central problem is the rate paid to the driver; Uber assumes the same rate for the driver regardless of how many people are in his or her car. During the shared route period where the driver is ferrying two parties along the same route, he or she is being paid the same $0.75/mile rate whether having 2 or 4 passengers in their car. Uber seems to think that filling a car with more people both maximizes utility with no difference to the driver and believes this justifies them to pay them even less.
In fact, drivers highly disagree and believe they do double the work for the same pay. This is a common and highly vocalized grievance from drivers. They liken an UberPOOL to doing the same work but for two bosses instead of one. Both passengers will rate the driver, both may have requests (music/charging phone), both are required to be separately picked up and dropped off. In fact, many drivers feel like the pick-up is the most stressful part of the trip: frantically waiting for the absent passenger, holding up traffic, and receiving cancelations enroute routine problems which cause drivers angst. The driver is paid the extra money for the wait time, but often feel like the marginal few cents (0.15c/min) is not worthwhile given the added inconvenience to the driver, particularly if there is already a waiting passenger in the car — hence why UberPOOL is often called UberFOOL in the driving community. Payout is optimized if the overlap between multiple passengers is small and drivers need to go further out of their way to pick up and drop off the additional riders. However, Uber now has guaranteed drop off time windows per ride, meaning that the dynamic matching will likely optimize to select rides with the greatest overlap.
Herein lies the central problem with Uber — it treats their drivers like commodities and scales their service like a cloud service. Empty seats in a car is wasted, unused space — why not fill them with other riders who can optimize unused utility with minimal added labor required. A friend, speaking to me about the disadvantages of working in consulting vs for a software company, said, “You can’t scale human effort, but you can scale a product.” Uber has realized that, at least for a regular sized vehicle, you can in fact theoretically scale a driver’s effort, by a maximum of four. Yet, in practice, it doesn’t quite work that way. Drivers aren’t machines, and until Uber replaces drivers completely with self-driving cars, Uber ought to acknowledge that its success is dependent on people rather than machines.
Uber pushes its drivers to use UberPOOL. It used to deactivate drivers who did not accept up to a threshold of ride requests (80–90%), but has slightly softened its punishment to “time out” periods of 2–30 minutes. Moreover, now when you take an UberPOOL ride, the driver will automatically be assigned a second or third rider if Uber can find one; previously the driver had to manually accept the ride. This causes driving difficulties when a ride changes mid-journey, annoyance from the passenger who finds their trip delayed, and potential safety issues if the rider has a low rating or is attempting to go to an area that the driver feels is unsafe. In these aspects, an Uber driver is much less an independent contractor as Uber purports, who chooses who to pick up at their liking, but rather, a cab driver under behavioral regulation from a commission.
Exasperation from Uber drivers is unquenched. Is driving two people the same as driving one? Uber says yes — and argues that the more people utilizing a car, the better. Drivers, who actually have to deal with the experience, are, on the other hand, highly censorious. As one driver puts it: “Is this a Travis Kalanick fetish, or is there some other gaggle of idiots sitting around in a board room trying to figure out the most epic ways to screw their drivers?” Unfortunately, the truth is, Uber one day will have screwed drivers out of a job completely. Uber has already cut pay to drivers by 35% since it began in New York in 2014. It can get away with this kind of behavior because Uber is not an employer — it’s a distributer of transportation. As it attempts to cut out the most expensive element of the supply chain — the driver — Uber moves closer and closer to its vision of changing the entire transportation industry.
As I’ve recently become more aware of the ethics of my technology usage, I’ve been thinking more and more about how the actual suppliers are treated in the sharing economy. Juno, an Uber competitor, is banking on differentiating itself through moral superiority: it heavily markets how their drivers are treated more equitably, are paid more, and can earn stock in the company. Lyft ought to market the same — Lyft is more generous to its drivers than Uber, has played fairly by adhering to city regulations, and generally have much happier drivers. Yet, one could argue that Lyft, evaluated at $5.5 billion compared to Uber’s $62.5 billion, may have been disadvantaged by its more considerate approach to scaling.
Low or unfairly paid workers are not by any means a new phenomenon in America, but the slight difference between the criticism against Uber and the criticism against corporations like McDonalds is that Uber drivers are not employees of Uber. When it was revealed that many McDonald’s workers live below the poverty line, McDonalds as a corporation was chided for not being a fair employer. Uber, on the other hand, washes itself of employer responsibility, as they do not employ any of their drivers and no one is making a driver participate. Clearly, the cost-benefit analysis must be in the drivers favor, or otherwise they wouldn’t do the job. Yet, in America, where there is always someone available to do a job at almost any price, the standard is so low that the gig economy, as a whole, and the end consumers, are able to take advantage of low prices at the workers’ expense. I encourage users to utilize the ride-share apps that at least attempts to reward their drivers more generously. Juno is one, and Lyft is another. Yet, I’m under no illusion that Uber, as by far the market leader, has an unparalleled advantage in the market with greater liquidity, more data, and the conspiring of strategic minds that have imagined an economy that previously didn’t exist.