Why Providing On-demand Services With A Fixed Size Fleet Is A Fool’s Errand

Consumers have been enamored with Uber and Lyft’s easy-to-use on-demand services since 2014. New startups are rushing to apply the same on-demand concept to other services or markets such as on-demand grocery delivery, on-demand laundry, and on-demand transit services. The list just keeps growing. Autonomous vehicle makers also plan to test their AV fleets as ride-hailing services, aka robo-taxi. Is on-demand really the magic bullet that delivers better services? The answer is becoming clearer each day. Not only is the answer “no”, it’s “heck no”. Delivering on-demand services comes with some steep costs.

Understanding the Real Cost of On-Demand Services

“On-demand means inefficiency to me,” a fleet owner with 25-years of experience under his belt once said to me.

On-demand is an expensive proposition. Yes, you can request an Uber nowadays in any major U.S. city and it will arrive in five minutes or so. Aren’t you impressed? That’s because you don’t know the real reason that your ride shows up so fast. Uber is keeping as many drivers as riders in your area to create that on-demand experience.

By crowd-sourcing cars and utilizing surge pricing, Uber or Lyft can mobilize tens of thousands of drivers and cars quickly and determine whether a service can be on-demand or not. If you can’t get many drivers to an area quickly, then riders experience longer wait times and the service then loses its on-demand shine.

Currently companies using these tactics can transfer the hefty cost of labor and vehicle to the driver by classifying them as contractors. This leaves the company only responsible for a fraction of the actual cost. In turn, they keep the service fees low.

Before the era of Uber and Lyft, there were only 1800 taxi cabs in San Francisco. Before COVID-19, there were almost 50,000 Uber drivers in San Francisco on any given day. Most of these drivers don’t live in the city and only in to drive for those platforms. Obviously, no company can afford to employ 50,000 drivers to serve the needs of a fraction of San Francisco’s population of 900,000. For more information on how companies like Uber transfer costs to drivers and why this business model is unsustainable, check out my article The inconvenient truth: How TNCs like Uber provide rides so cheap.

Cost Efficiency and On-Demand Services

“Will an on-demand model truly benefit my business?” is an important question that any business owner or agency must carefully consider before jumping on the on-demand bandwagon. If your business has a high fixed asset cost, such as vehicles and driver employees, you should probably stay away from the on-demand model. Why? It boils down to cost efficiency.

To better understand the concept, let’s look at an extreme case such as the airline industry. Can you imagine if United Airlines or American Airlines provided the type of on-demand services like Uber? You would simply walk to the ticket counter at the airport and demand a flight from San Francisco to New York City in the current hour. And you would only pay $100 for it.

No airline would financially survive a day in that scenario because of the significant costs of operating the business such as planes, fuel, flight crews, supplies, insurance, and so on. Those costs exist even when there is only one passenger on-board. To be financially responsible (remember unlike transit services, airlines are for-profit businesses), airlines know they must have a certain number of passengers on a flight to avoid losing money. Sometimes it makes better financial sense to cancel a flight and switch the passengers to another flight or airline.

Over the years, airlines have developed very sophisticated operating models to maximize their revenue and financial returns. This explains the myriad scheme that includes different seat classes, tiered pricing by time, over-booking rate, and loyalty programs. Southwest Airline role-models excellent operations in the airline business. Travelers do have the options of booking a flight days or weeks in advance or buying a ticket for the next available flight at the airport that day. But the prices for the same seat can vary drastically depending on when you purchased the ticket. And it’s still the airline, that sets the time and determines the price for your flight, not you. You’re only going to have truly on-demand flights when you operate your own private jet.

Airlines are the perfect example of how the on-demand model doesn’t work with high capital cost or fixed cost. If airlines don’t operate on-demand, then fleet operators need to be careful when considering on-demand service for the same reasons.

Supply Adjustment and On-Demand Services

Another factor to consider is the supply adjustment, i.e., how fast or easy is it to scale up or reduce your supply. No matter whether you’re a fleet owner or a transit agency, you have a finite number of vehicles to dispatch for a given service and you can’t add or de-commission vehicles quickly.

In a given time, if the number of riders exceeds the numbers of available vehicles, the riders who aren’t in the first group served will have to wait an additional amount of time. There are companies that attempt to apply ride-hailing technology to a fixed size fleet operation for on-demand services. They argue for the supply-model misfit by promising to pool riders in real-time and hence increase efficiency. But they can’t do this effectively in the real world.

To pool riders on-demand is not effective especially in less densely populated markets such as the suburbs where the demand is very scattered.

Imagine this: Riders A and B each book a ride at the same time. Your on-demand algorithms pool A and B into one vehicle with an optimized route and dispatch the vehicle to pick up A first.

Five minutes later, the vehicle picks up A, and is on the way to B’s location. Now C books a ride and their pickup location is the same as A’s. But the vehicle is now closer to B than A. And it will take another 15 minutes to drop-off both A and B. Now you have two options: –

1) reroute this vehicle back to pick up C after picking up B, which means both A and B are stuck in the vehicle for longer than promised and are obviously dissatisfied;

2) or you can send another vehicle to pick up C. This isn’t ideal because it’s an additional cost that’s not necessary if you could have put C in the same car as A and B.

Now, what if all your other vehicles are much farther away from C’s pickup location than the vehicle with A and B in it? You’re then stuck with the option that C will either have to wait for the vehicle to come back after it completes the trip for A and B, or another vehicle must be sent to drive a much longer distance. You will incur more cost (longer distance) and still have an unhappy customer. Overall, your operation is inefficient regardless what you do. Using a fixed size fleet to provide on-demand is really a fool’s errand if you want to operate efficiently and be financially responsible for the business.

How The Right Technology Helps

Technology must fit the business model it supports. Not every technology can be extrapolated to another business model even if the two businesses are seen to be similar, e.g., taxi and transit. Ride-hailing’s on-demand technology is designed to work with crowd-sourced cars, not a fixed size fleet.

If you’re a professional fleet operator, don’t be blinded and burned by the white-hot light of on-demand service technologies. Instead look for technologies that are designed specifically for professional fleet operation and focus on efficiency improvement, such as Duet’s Mobility-as-a-Service (MaaS) technology platform.

Founder & CEO of Dashboard Story, Inc. (dba Duet), a transportation logistics technology company, www.duetinc.com. Alum of Intel. Abbott, Cornell, Berkeley-Haas

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store