On-Demand Mobility Services, and particularly ride-hailing, have emerged as a strong option for consumer urban transportation. In the process, ride-hailing has disrupted the taxi and limo industries and could next disrupt public transportation and last-mile package delivery. Nowhere is this more evident than in cities such as New York and San Francisco. Other mobility services such as shared ride-hailing, and micromobility, as well as various forms of microtransit and car sharing are also showing robust growth. In a previous post I organized automotive OEMs into five categories. I’ve tried to create a similar structure with a small set of categories for the companies offering on-demand mobility services, focusing particularly on ride-hailing services. It proved to be a harder task because of the complex interactions among the dimensions I used. In this post I present my first attempt to organize these companies into five categories based on how they approach next-generation mobility and the value they offer to their customers.
On-demand mobility services is not new a new concept. Car rental, as offered by companies like Hertz and Avis, and public transportation provide early manifestations of mobility services. However, over the past 10 years we have seen new types of mobility services entering the market. From Zipcar’s reinterpretation of car rental to the car sharing offerings of Car2Go and Maven, and the various forms of ride-hailing and ride-sharing popularized by Uber, Lyft, Via, and Chariot. For example, consider the changes to the car rental operating model. Traditionally, car rental companies like Hertz, and Avis required consumers to go to their own lots to pick up a vehicle. Zipcar, among other innovations, changed the model by determining where to preposition vehicles in locations such as parking lots, in order to better match and optimize consumer demand for vehicles with vehicle supply and thus increase each vehicle’s utilization and the revenue per vehicle.
DriveNow, Car2Go, and Maven took Zipcar’s model even further. They allow cars to be picked up from any location they are parked. Companies like Turo made a further refinement to the car rental model. They enable individuals to offer their privately-owned vehicles for rental. The evolution of the car rental model was driven by recognizing the need to offer consumers better pricing options, improved convenience and overall better customer experience without degrading the safety of the vehicles rented out.
Similarly, the first incarnation of the ride-hailing model, as introduced by companies like Uber, Lyft and many others, can be considered an evolution of the taxi and limo service model. Shared ride-hailing, e.g., UberPool, and microtransit models, e.g., Via and Chariot, are considered important evolutions of the original model. Price, convenience, customer experience, and safety are dimensions driving the success of all these services.
Analyzing On-Demand Mobility from A Consumer Perspective
We can analyze on-demand mobility services from a consumer’s perspective using the following four dimensions:
- Convenience: This is determined by the availability of a desired mobility service between a point of origin and a destination. For example, if a consumer wants to go from San Francisco’s Caltrain train station to the Moscone convention center they can choose to use a ride-hailing service offering point-to-point, single passenger, transportation, or a fixed route microtransit service offering multi-passenger transportation. The first will get the consumer exactly at the desired destination. The second will likely drop off the consumer nearby the destination.
- Passenger experience: This is a more complex dimension. It is determined by taking several factors into consideration. These factors include the overall user experience while interacting with a particular mobility service in order to accomplish a transportation goal, how complex of transportation goals the mobility service company can satisfy, the in-vehicle personalization experience while being transported, how quickly the vehicle will arrive once the service is ordered, how fast the transportation will be completed, and how accurate these estimates were. For example, getting from the Caltrain station to San Francisco’s Transamerica Tower during the early afternoon hours in the middle of a week may be faster if the mobility service user is willing to combine ride-hailing with bike-sharing, instead of using ride-hailing alone. A company offering such a multi-modal transportation option to a consumer who can utilize such an option, i.e., the company must be able to determine whether the consumer is able to bike and willing to bike, provides a better passenger experience.
- Price: Price was an important factor for the success of ride-hailing services over city taxis. While in many cities price parity between taxis and ride-hailing services has been achieved, for many consumers today low prices combined with convenience represent the main factors for using certain on-demand mobility services.
- Safety: For mobility services that are provided using human-driven vehicles, safety is established both by the condition and safety record of the vehicle used by the mobility service, as well as the driver’s safety record when using any form of human-driven ride-hailing services. Autonomous vehicles are considered to be safer than human-driven vehicles. However, based on published results one can conclude that the companies developing autonomous vehicles still need to complete significant work before we can consider this issue fully addressed.
Autonomous vehicles are viewed as critical to the ongoing success of on-demand mobility services. Their use is expected to significantly reduce the operating costs of such services thus improving their profitability prospects, but also reducing the price of service thus benefiting users. In addition, autonomous vehicles are considered safer than human-driven vehicles, can provide better passenger experience, and, if optimized effectively, can improve the user’s overall convenience.
Analyzing On-Demand Mobility from A Business Perspective
Analyzing on-demand mobility services from a business perspective requires more dimensions and a more complex set of criteria. After evaluating startups, established private companies, and the newly established mobility services operations of automotive incumbents like GM’s Maven, Daimler’s Car2Go and Moovel, and others, I decided to use the following six dimensions for this analysis:
- Vehicle type: A large variety of vehicles will be used in the context of next-generation mobility for passenger transportation and last-mile goods delivery. They include conventional ground vehicles with Internal Combustion Engines (ICE) (car, minivan, van), autonomous electrified (ACE) ground vehicles (car, minivan, van), Unmanned Aerial Vehicles (UAVs) including air taxi, delivery drone, and long-haul Vertical Takeoff and Landing (VTOL) vehicles, bicycles (conventional or electric), as well as scooters (conventional or electric, 2- or 3-wheeled). Each vehicle type has its own economics and can be used for different mobility services and in specific territories (see below).
- Mobility service type: We have already mentioned the growing variety of mobility services that have been introduced. From single passenger point-to-point ride-hailing, to ride-sharing/microtransit with fixed routing, ride-sharing with dynamically determined routing, and micromobility. We are now seeing the introduction of services offering multimodal transportation, e.g., combining car transportation with bike, escooter, and bus. Each service can be delivered by specific vehicle types impacting the business model and cargo type (see below).
- Company type: Two mobility services company models are emerging:
- Ride coordinators. These are the ride-hailing companies we know today. They coordinate rides in vehicles owned or leased by individuals.
- Fleet managers. We will see two types of fleet managers. First, managers of fleets that are owned by another company. For example, in the partnerships that Ford is developing with Postmates and Domino’s Pizza, Ford will manage the Domino-owned fleet. Second, companies that both own and manage their fleets. Examples of this model will range from pureplays like Waymo, Zoox, and Amazon, to automotive OEMs like GM. Bike sharing companies like Mobike, Ofo, Limebike, and others, as well as escooter sharing companies like Bird and Spin also fall in this category.
- Business model: Ride-hailing companies are starting to test subscription-, and advertising-based business models as complements, or in addition, to the transaction-based models we are familiar with. We will also see the emergence of Passenger Commerce-based models, and loyalty program-based models in the near future. Subscription models are already offered by some automakers. They can be particularly well-suited for car sharing services. In the near future we may also see models that combine a subscription with transactions, similar to what we find today in fractional ownership of private jets.
- Operating territory: The service’s operating territory may be one or more of inner-city, suburbs, intercity medium or long-haul transportation, or country to country transportation.
- Cargo type: We see several options in this area from single passenger-only services, such as what will be offered by various types of ground-based or aerial pods, multiple passengers-only that will be offered using cars, minivans, and air taxis, goods-only, that will be offered using vans and trucks, and mixed-mode, i.e., the transportation of both passengers and goods, that will be offered using cars, vans, and UAVs. The last type will be the one most likely used for inner-city and suburb transportation as mobility services companies will look to maximize the revenue-miles of each autonomous vehicle.
Five Categories of On-Demand Mobility Services
When considering the relations between the four passenger-centric dimensions with the six business-centric dimensions that organize the on-demand mobility services companies we first observe that convenience is influenced by the type of mobility services offered, the cargo type that can be transported, the business model used, and the vehicle type(s) employed by the service. Passenger experience is influenced by the type of mobility services offered, the company type, and the vehicle type(s) employed by the service. Price is influenced by the business model used, the types of mobility services utilized, the vehicle types(s) employed by the service, the operating territory, and the type of cargo being transported. Safety is influenced by the vehicle type(s) employed by the service, and the company type because of the level of control it will have on the vehicles providing the service.
Based on this analysis I have created five categories of on-demand mobility services providers. In some cases, there already exist mobility services companies that belong to these categories. In others we expect such services to be created with the introduction of various autonomous vehicle types. To further emphasize the previously mentioned similarities between the airline value chain and the Fleet-Based On-Demand Mobility Value Chain, together with the characteristics of each category and example mobility services companies that possess these characteristics, I also provide examples of airlines that have similar characteristics. The five categories are:
- Category A: Mobility services companies offering high-end service but limited coverage. These are characterized by high price, superior passenger experience, medium convenience because of their limited territorial coverage, and high safety. The mobility services companies created by high-end automotive OEMs, e.g., Daimler, BMW belong in this category based on the vehicles they use and the customer segments they primarily target. This is the Singapore Airlines model.
- Category B: Mobility services offering basic service, limited coverage, and average passenger experience. They are characterized by moderate price, average to low passenger experience, medium convenience because of their limited coverage, and high safety. Mobility services companies operating regionally with strong knowledge of the region’s characteristics, such as Grab in Southeast Asia, and Easy Taxi in several Latin American countries, belong in this category. These companies will likely continue to use conventional vehicles and remain ride coordinators. They may partner with the companies in Category E, or they may be acquired by them. This is the Skywest Airlines model.
- Category C: Mobility services companies offering basic service, limited coverage, and great passenger experience. They are characterized by moderate price, superior passenger experience, medium convenience because they will use autonomous vehicles to operate in specific territories, e.g., inner-city passenger and goods transportation, and high safety because they use autonomous vehicles. Ride-hailing pureplays such as Waymo, and Zoox, as well as the divisions of automotive OEMs such as GM Cruise, and Ford will belong in this category. Last-mile delivery pureplays, such as the one that Amazon can develop using autonomous drones, will also be part of this category. Depending on the territories they will cover and the types of mobility services they will offer, these companies may use a small variety of vehicles, e.g., minivans, SUVs, shuttles, rather than a single vehicle type. This is the Virgin America model. Some of the companies in this category may be able to support larger territories while maintaining their other characteristics. In this case their model will be more analogous to that of Southwest Airlines rather than Virgin America.
- Category D: Mobility services companies offering budget service belong in this category. They are characterized by low price, average passenger experience, medium convenience because often times the user has to go to specific locations to pick up the vehicle, and medium safety because of the vehicles that are often used by these services. Microtransit services, that may or may not use autonomous vehicles, such as May Mobility, and Voyage, as well as companies that use scooters, motorcycles of various types, such as Go-Jek, and even bike-sharing companies belong in this category. This is the Frontier Airlines model.
- Category E: Mobility services companies offering high convenience at a price. They are characterized by high convenience because they operate globally, moderate to high price because of the fleets they use and the regulations they adhere to in each country they operate, average passenger experience, and medium safety as demonstrated by the incidents that have been reported to date involving drivers associated with these companies. The companies in this category will adopt various types of autonomous vehicles, e.g., minivans, SUVs, shuttles, in order to improve their profitability while reducing price, improve the customer experience and their safety record. In order to be able to operate globally at scale and effectively address demand elasticity (in other words, so that they can respond to peak demand without building overcapacity), these companies will most likely use mixed fleets that include both conventional privately-owned vehicles together with the autonomous vehicles they operate. In addition, they will offer their service for consumer transportation and last-mile goods delivery. Global ride-hailing companies like Uber, Lyft, and Didi are the best examples of companies that belong in this category. This is the United Airlines model.
The characteristics of these categories are summarized in Figure 1.
Figure 1: The five categories of on-demand mobility services companies
By analyzing these categories, we arrive at the following observations:
Observation 1: On-demand mobility services companies must continue investing heavily for the foreseeable future. To date, particularly the companies in Category E have been investing heavily in driver and passenger acquisition, demand generation, and technology development for ride coordination. However, operating fleets of autonomous vehicles will require larger investments for vehicle acquisition, technology development and acquisition, as well as talent acquisition, i.e., hiring people with specialized skillsets such as fleet management. As companies like Uber and Lyft are quickly finding out, their software prowess that allowed them to develop state-of-the-art systems for ride coordination does not automatically translate to leadership in autonomous vehicles technology. They needed to hire new teams of autonomous system specialists, sometimes via acquisitions, and continue to expand them aggressively. Uber’s autonomous vehicle team now exceeds 1000. These companies will need to continue ramping up their technology investments to develop systems for fleet optimization, e.g., balance vehicle demand with the supply available vehicles including autonomous vehicles, revenue optimization, e.g., maximize passenger revenue-miles, etc.
The cost of available capital will determine the rate at which these services continue to proliferate and adopt autonomous vehicles. Private companies offering mobility services had little trouble raising large sums of capital to accelerate their geographic expansion plans but also their technology investments. Similarly, incumbent automakers and their suppliers have used profits from certain operations to fund their autonomous vehicle efforts. However, as the cost of capital increases, mobility services companies of every category will be forced to make choices. They will have to determine in which territories they will be able to operate profitably, what vehicles to use, and what business models to use, implying that they may need to constrain their expansion plans.
Observation 2: Transportation, communication, and electrification infrastructures will require large investments to address the needs of autonomous vehicle fleets. The breadth of the autonomous vehicle adoption by companies offering on-demand mobility services will be dependent on the state of the transportation (roads, bridges, etc.), communication, and electrification (power generation, power distribution grid) infrastructures in the areas where such services are deployed. Places with poor infrastructures will not be selected by the companies that want to offer such services using autonomous vehicles even if these areas have many of the other important characteristics such as high population density, favorable weather, etc.
Companies in categories C, D, and even E can draw consumers away from public transportation. Under such conditions cities may have to select how much to invest in their transportation infrastructures and how much on their public transportation networks. Public transportation systems globally are perennially subsidized and underfunded. Additional funding reductions will lead to major degradation of already sparse service. As public transportation services degrade, certain consumer groups, for example those living in outlying suburbs or rural areas, may find themselves with no options as they may not be able to afford on-demand mobility services while public transportation does not serve them.
Observation 3: The business models of on-demand mobility services companies are still in flux. The business models that rely on the economics of autonomous vehicles are evolving and will continue to do so for the foreseeable future. Many are incomplete, have many open issues, and make assumptions that have not yet been validated. The costs of many on-demand mobility services will remain uncertain or unknown until the autonomous vehicles expected by these services are deployed at some scale. This is not expected before the middle of the next decade and will likely vary by location.
Observation 4: OEMs offering on-demand mobility services must carefully select their fleets. In the previous post of this series I identified two categories of automotive OEMs that offer on-demand mobility services. The mobility services companies these OEMs set up, like GM’s Maven and VW’s Moia, will be required (or at least be pressured) to utilize exclusively vehicles that are in the parent OEM’s model lineup. For example, Maven uses a variety GM models from its various lines in its carsharing program (Chevy, Cadillac, etc.). Cruise, GM’s ride-hailing business, will utilize Chevy Bolt subcompacts for its autonomous fleet. Companies utilizing a variety of models will be able to operate in more settings, e.g., small cities, congested cities, spread-out metropolitan areas. Companies utilizing a single model, e.g., Nio, could have certain operating advantages (easier to maintain a single vehicle type rather than training mechanics on multiple models, lower parts inventory, etc.) as we are already seeing in the airline industry. For example, Southwest Airlines achieves strong financial performance by operating exclusively Boeing 737 aircraft. Therefore, the configuration of the fleets that OEMs decide to create will have a direct impact on the on-demand mobility services category they will belong.
Observation 5: Dealing with congestion will require more sharing. As the popularity of on-demand mobility services continues to increase, it often leads to increasing congestion, as we are already seeing in cities like NY and San Francisco. Under heavy congestion conditions autonomous vehicles may be programmed to automatically re-route themselves in order to avoid congested areas. But such decisions could result in rides that take longer than expected. Longer rides, whether because the vehicle is stuck in a congested area, or because of re-routing to avoid congestion, will negatively impact passenger convenience and the overall transportation experience. On-demand mobility services companies may be forced to offer more shared services using autonomous vehicles rather than single passenger services. Finally, even ride sharing may still prove inadequate for addressing congestion. This may necessitate other types of regulations to limit traffic, including various forms of surcharges and taxes that will obviously negatively impact the business models of these companies.
Observation 6: Consumer education will be necessary. Consumers across many segments and geographies will need to become convinced of the benefits offered by autonomous vehicles in general and the spectrum of on-demand mobility services that utilize such vehicles in particular. For example, while Millennials may feel more comfortable with autonomous vehicles, Baby Boomers do not. This education must address consumer anxieties relating to the use of such vehicles (the technology works correctly all the time, not most of the time; there are both short-term and long-term benefits from using autonomous vehicles), and any issues relating to the potential job losses because of the employing autonomous vehicles for a variety of services.
The on-demand mobility services companies, together with the autonomous vehicle manufacturers and suppliers, must also address the public’s anxiety relating to cybersecurity and data privacy risks. The public must understand how safe the vehicle is. For example, can the vehicle be hijacked by hackers while a consumer is being transported? It must also understand what data is being collected and for what purpose it is being used. This means that the public may also need to understand in some detail a company’s business model, and how data impacts it.
Only once the public feels positive about how these issues are addressed they may feel comfortable consistently utilizing autonomous vehicles for on-demand mobility services.
There is no question that today many consumers are intrigued by autonomous vehicles. Their curiosity will lead them to try out such vehicles for at least a few times. The challenge for the on-demand mobility services providers will be to convert them into regular users while charging them rates that will ultimately result in real and profitable businesses. The use of autonomous vehicles for the transportation of goods is an easier case to accept and get behind. Companies providing on-demand consumer transportation must constantly evaluate whether their service can be offered safely and reliably in the cities that have strong financial opportunity for such a business, e.g., San Francisco, NY. As it turns out, these are also the cities with the most challenging traffic environments where autonomous vehicles have the most trouble.
The next article in the series.
The previous article in the series.