Car as a Service

Ridesharing is a platform that will define the next decade of transportation

The concept of ridesharing is deceiving because it is made up of two words which conjure up clear, but static definitions. As a kid, nearly every morning I would stand at a curb waiting to flag down a big, yellow school bus to take me to school. The driver never knew my name, though I saw him every day.

A few of my friends were on my route and during our ride together, we would open our lunchboxes and share (barter or steal) the things our parents packed us.

As children, we were permitted to ride in or on certain things: booster seats, school buses, a friend’s car; but we were prohibited from getting into a stranger’s car. In the same way, we were permitted to share certain things: toys, personal space, ice cream cones; but, we were prohibited (or strongly persuaded) from charging people to use our things or sharing things we generally wanted back.

Nowadays, I’m often paying strangers equipped with a car to chauffeur me around, not minding if they pick up others along the way. Technological progress is forcing us to redefine our notions of what it means to “ride” and “share” yet the labels we’re currently using to describe the “‘sharing’ or ‘gig’ or ‘on-demand’ economy” do not capture its disruptive and transformative nature,” says Walter Isaacson. Labels tend to come with pre-packaged meanings that don’t get at the heart of entirely new concepts. Instead of a label, let me offer a definition:

Ridesharing is infrastructure for the efficient coordination of people, goods, and services.

Notice that there are three parts to this definition: infrastructure, efficiency, and coordination. Let’s address each of those parts in turn; but a bit out of order. Let’s start with “efficiency” followed by “infrastructure” and “coordination.” You’ll see that the order is important because the promise of coordination is dependent upon an infrastructure that addresses the current market inefficiencies inherent in the transport of people and commerce.

Each component could be built out but for the purposes of this post, let's use the following three statements as a rule of thumb:

  1. Inefficiencies abound because the world has too many drivers and not enough (productive) cars
  2. If a company can build an infrastructure that capitalizes on this scarce resource by harnessing the liquidity a smartphone-enabled world provides
  3. This company could reduce all friction in transport and thus, realize coordination at a magnitude we’ve never experienced before (or, in other words, all decisions regarding the transport of people, goods, and services become if/else decision profiles following a Zipf distribution curve)

There is a lot packed into this storyline but for now hold on to this: the purpose of a car is to be productive, and it is currently being underutilized with severe externalities (i.e., traffic jams, poor urban planning, wasted time). In a low-friction environment (in this case, when car access becomes abundant), a company can either serve (and cultivate) the exact needs/wants of consumers or it can become the operating system that powers itself and/or others to serve customers.

I believe the company that “wins” ridesharing will build the operating system – utilizing vehicles to service functions, which will release cars from their traditional, singular use as objects. This is the difference between saying, “I need to get to work, or buy milk, or go on a camping trip” and “I need a car, or a bus, or the tram”. Alex Danco of Social+Capital speaks fluently about this distinction in his post, “Understanding Abundance, part 3: The Next Big Thing” where he says:

[F]unction plus infrastructure architecture isn’t merely important. It’s The. Most. Important. Thing. It’s the beginning of something as disruptive as our first two paradigm shifts of tech: as meaningful as Software plus Hardware, or Web plus Internet. F+I computing, in other words, is becoming the core idea at the middle of our next wave of abundance, as we have defined it: the state when friction of consumption [of functions] approaches zero. All of the elements going into that consumption  —  AI, Cryptocurrency, IoT on the infrastructure side; Services, skills, voice, AR, bots, on-demand access on the functions side, will be organized in this paradigm. The operating system sits in the middle.

Given this thesis, I want to valuate Uber as a case study. I’ve linked my workbook here as aid.

What follows below are a few pre-read thoughts and insights so one can quickly read through the workbook.

But before we dive in, let’s do a quick pitch for the company.

Uber is a car-sharing company founded in 2009 by Travis Travis Kalanick and Garrett Camp in San Francisco with a paper valuation of $69B. Servicing 83 countries in about 674 cities on six continents, Uber’s product offerings include U4B for businesses, UberPool for carpooling, and UberRUSH for deliveries. Uber’s low capital intensity model has allowed for rapid growth as measured in cities and rides, though the rate of growth has started to slow down. Reports from Q1 2017 show increased net revenues (rising 18% from Q4 2016 to $3.4B) and reduced losses to $708M from $991M. While revenues continue to rise, competition also increases. Across all markets, competitors have not been intimidated by Uber’s scorched earth, global strategy, and instead have leveraged partnerships, local advantages, and capital infusions to establish beachheads. Competition was so severe in China that Uber exited with a 20% stake in Didi Chuxing. Uber at the same time is not helping with a string of seemingly never-ending legal, reputational, and sexual snafus which have led to public outcries like #deleteUber.

I value Uber at $44B for the following reasons:

  1. The market is the wild, wild East

    Uber is one of the early pioneers into a $195B market. The terrain includes the following markets: taxis, limos, and rental cars. In 2016, Uber generated $6.5B in net revenue and $20B in gross billings (2016), up 2X from gross receipts in 2015, according to Bloomberg. I assume Uber will generate $9B in net revenues in 2017. I doubt that Uber is profitable after covering its operating costs.

    Uber will increase total market demand for ridesharing services by 50% over the next 10 years (at an annual growth rates of 7.26% and primarly in developing countries). How much of this new market facing towards the Orient it will capture is up for debate? So far, the verdict is not good.

    Ridesharing companies are not only disrupting existing players in the market but also attracting new users into the market, both by attracting non-cab users to try ridesharing as well as by increasing the usage of car services, in general. As ridesharing becomes more widely accessible, more users will adopt, and the value of the service, according to Metcalf’s law will continue to grow. Estimates already show that firms in the ‘rides’ industry increased nationally by 63% in 2015 alone.

  2. Artificial intelligence is a tool for some, a threat for others

    However, clouds loom. Ridesharing companies are bracing themselves for the coming disruption of artificial intelligence. Autonomy will increase access to vehicle productivity and liquidity, meaning that the costs associated with the current operating equation: (driver) + (vehicle) + (fuel) + (maintenance) will reach “zero” with time. That is, we eliminate the (driver) + increase utility design of (vehicles) for multiple use cases + use alternative (energy) sources + reduce amortized (maintenance) costs due to the removal of combustion engines. As a result, we will realize cheap transport of everyone and everything.

    Uber had 40M users at the end of 2016, up from 24M users at the end of 2015, according to TechCrunch. Contribution margins in Uber’s most profitable cities range from 3-11% of gross billings and contribution margin in San Francisco, Uber’s longest standing and most mature market, is 10.1%, according to Fox Business. David Sacks of Yammer and Paypal tweeted that Uber’s success relies on a powerful virtuous cycle in which geographic density breeds a network effect.

    More geographic coverage/saturation leads to faster pickups which leads to more demand which leads to more drivers which leads to more geographic coverage/saturation, and so on. I agree with Sacks but what happens when AI arrives? Doesn't the elimination of drivers undermines Uber’s business model? As Alex Danco says, “without the moat of Uber’s driver network and without its ability to freely arbitrage liquidity as it’s used to, the barriers to competing with Uber as it is today come down.Uber will soon have to overcome an inability to scale the infrastructure of its car-as-a-service business model once it eliminates drivers.

  3. Uber will be too slow on the draw

    I conservatively predict that Uber will capture only 5% market share due to open, brutal competition in every market and the rise of AI. Once the ride sharing market stabilizes in 10 years, Uber will be left with a business model with no network effects and medium-to-low switching costs for both drivers and customers.

    Uber isn’t fixed into this business model and maybe a new CEO will explore other models; for example, becoming a large fleet owner/operator in 3 phases: purchasing or leasing out the cars of current drivers, acquiring peer-to-peer carsharing services, like Getaround, and rental services, like Avis, and retrofitting as many of these cars with AI to create a distributed network of vehicles utilized for on-demand transport of people and services.

    Uber will have to reinvest at a sales/capital ratio of at least 2% (which is the US company median) in order to stay competitive and spur continued growth. As Uber matures, it will need to invest and hold hard assets in cars and equipment like a median US company in order to scale like a software-technology company.

So what does this all mean?

There is an elephant in the room and its name is Tesla. Uber is an amazing company. No question. The company has redefined ridesharing for a smart-phone enabled world. However, the playbook Uber has used to scale thus far is not sustainable in an increasingly artificially intelligent world. It must steer with caution or risk a crash.

Remember, ridesharing is about building a digital infrastructure that maximizes the productivity of cars in order to efficiently coordinate the transport of people, goods, and services.

The quesion is whether Uber has this vision? I don't think so. But Tesla does. I haven’t mentioned Tesla explicitly until now, but it doesn't take a genius to realize that my definition of ridesharing aligns with Tesla’s 10-year journey to create an affordable, high volume car.

Elon Musk wrote in his Master Plan, Part Deux:

You will also be able to add your [Model 3] car to the Tesla shared fleet just by tapping a button on the Tesla phone app and have it generate income for you while you're at work or on vacation, significantly offsetting and at times potentially exceeding the monthly loan or lease cost. This dramatically lowers the true cost of ownership to the point where almost anyone could own a Tesla. Since most cars are only in use by their owner for 5% to 10% of the day, the fundamental economic utility of a true self-driving car is likely to be several times that of a car which is not.

Also notice what Tesla’s 2016 shareholder letter says:

Our new Autopilot hardware platform, deployed in Q4 2016, establishes a vehicle technology threshold that is unmatched by any other production car … despite the complexities of implementing Autopilot hardware on both Model S and Model X, we still produced 77% more vehicles in Q4 2016 than in Q4 2015. In Q1 2017, we began pushing over-the-air software updates to all Autopilot-equipped cars to further increase performance and safety.

Tesla is Uber’s foil – a car production company with a vision of a distributed, vehicular fleet. It seems to me that Uber primed the market, and now we're waiting to see who amongst the various competitors (i.e., Google, Uber, Lyft, Tesla) will dominate the market. I'm leaning towards Tesla but either way, we’re in for an interesting ride.