Applying a bidding mechanism to shared scooters and parking spaces
This is the English version of my article originally written in Chinese. If you prefer to read the original article, you can find it here.
After reading “Radical Markets,” I realized that the contents of the first chapter are quite similar to the product logic that I had previously contemplated during my time at GoShare. I attempted to organize this business structure and even extended it to government transportation policies.
Firstly, in the traditional Chinese version of “Radical Markets,” the final section of the first chapter, “Insights from the Ground,” explicitly mentions two Taiwanese shared scooter brands, GoShare and WeMo. Shared scooters, in essence, aim to replace ownership with the right to use, reducing the public’s reliance on private transportation.
However, under the current operating model, there are two major pain points that have yet to be thoroughly addressed. In simple terms, these are 1) a lack of available vehicles when renting and 2) a lack of parking spaces when returning the vehicles. Let’s break them down further:
1. A lack of available vehicles when renting
The simple reason for this problem is “insufficient vehicle deployment,” or sometimes “insufficient dispatching capacity.”
2. A lack of parking spaces when returning the vehicles.
The number of scooter parking spaces in Taipei City is far fewer than the number of scooters, especially in certain areas during specific times (such as weekday mornings in business districts or weekends in Ximending), making it extremely difficult to find a parking spot.
One business strategy adopted by operators is to negotiate with private parking lot operators. WeMo has already partnered with numerous parking lots in Greater Taipei, using license plate recognition for entry and exit.
Public bicycle sharing systems
In the case of the YouBike system, which operates at fixed rental and return points, the above two problems also exist. It is common to see Giant’s trucks dispatching YouBikes to stations with low inventory or staff temporarily locking bicycles on the sidewalks near the stations to make it easier for more people to return their bikes at popular stations. It is speculated that the labor costs of staff and the external costs of occupying sidewalks have not been taken seriously from a commercial perspective, possibly due to the public bicycle sharing system’s policy support and subsidies.
Shared car services
Taking iRent’s operations in Greater Taipei as an example, it offers both roadside rental & return as well as same-station rental & return models. For the former, in addition to public roadside parking spaces, operators also sign contracts with private parking lots in an attempt to solve the problem of returning the cars.
However, we can still observe that many shared cars are concentrated in areas with more available public parking spaces, such as riverbanks or industrial areas, and the situation of not being able to rent a car is likely to be prevalent.
How a bidding system can solve the problem of not being able to rent a vehicle
Create an open, real-time bidding platform for vehicle ownership.
In concept, first split the operator’s operating scope based on the distance that general users are willing to walk to pick up a vehicle. At the same time, split the time period into small units, preferably using the industry-standard practice of activating a vehicle within 10 minutes of booking as the minimum unit. A day is divided into 144 “time units,” and usage rights are redistributed at every 00, 10, 20, 30, 40, and 50-minute mark.
Users can bid for the right to use vehicles within the split “zones” and “time units,” and the allocation to users will be based on the number of vehicles in that “zone” and will be distributed from highest to lowest bid.
User journey description
Open the app > GPS automatically locates the user’s position > The app prompts the user with the number of available rental vehicles within a 300-meter radius and real-time bidding for vehicle reservations.
The user can set the radius of the search area, and in theory, the smaller the radius, the fewer vehicle options and the higher the bid. Users can bid in real-time for the right to use the vehicles within the split “zones” and “time units,” and allocation to users will be based on the number of vehicles in that “zone,” distributed from highest to lowest bid.
Scenario Example
For example, a few employees who are about to get off work at a company’s headquarters usually clock out at 6:00 pm. They open the shared scooter app at around 5:51 pm and bid for the right to use the vehicles in the “zone” to which the headquarters belongs, within the time unit of 6:00–6:10 pm.
- When the user opens the app at 5:51 pm, the system defaults to the user’s GPS location and the next time unit for bidding, which is 6:00 pm.
- If there are idle vehicles available for reservation in the “zone” during the 5:50–6:00 pm period, or if new vehicles have entered the area, the user can reserve the vehicle for free. However, if the vehicle is not activated before 6:00 pm, it will be put back into a new time unit for bidding.
- If there are three available rental vehicles in the “zone,” the system will collect bids from the top three users who bid for the vehicle right to use. These three users can activate their vehicles before 6:10 pm.
- If the user is one of the top three successful bidders for the 6:00–6:10 pm time unit but expects to pick up the vehicle at 6:15 pm, the user can continue to participate in the bidding for the 6:10–6:20 pm time unit. However, since the available rental vehicles may not be three anymore when the system collects bids at 6:10 pm, whether the user can be allocated the vehicle right to use will be re-ordered based on the idle vehicles and the bid amount.
- If there is a shortage of vehicles in a specific time unit in the “zone,” the user does not need to pay the bid amount.
Summary
Users can also bid to reserve vehicles for a future date and time, and the platform can provide users with historical data for that “zone” and time unit’s available rental vehicles and successful bid price ranges.
If a large number of users reserve vehicles for a specific area but no vehicles flow in, the platform can create a real-time information section, encouraging users who return their vehicles to that “zone” to earn additional rewards. These rewards come from the additional bid amount that users are willing to pay to reserve their vehicles. Thus, we create an open, real-time supply and demand market for vehicle dispatching.
How a bidding system can solve the problem of not being able to return a car
Create a real-time bidding platform for parking space usage rights
In concept, the rate for each parking space will fluctuate during each time period, and we will use one hour as a time period. Let’s assume the starting bid for each hour of scooter parking space is NT$1.
Suppose there is more supply than demand for parking spaces in the area, that is, there is no bidding situation, then the fee for parking for 24 hours will be the base price of NT$24.
On the platform, users can reserve the use of parking spaces at their destination in advance and pay the corresponding bidding price. The bidding logic is the same as for scooter usage rights.
User journey summary
- Open the app > Select destination location > The app prompts the number of available parking spaces within 300 meters, as well as the real-time bidding for reserving parking spaces.
Users can set the radius in meters themselves. The smaller the radius, the fewer parking space choices and the higher the bidding price. Bids can be changed in real time until the vehicle is parked.
4 parking space status
As the number of parking spaces in the area decreases, the platform will gradually adjust the bidding base price for parking spaces. For example, there are 30 parking spaces in the area and currently 30 people have reserved them, with the lowest bid being $3 per hour. If you are willing to pay $4 per hour for parking, you can obtain the usage right. The platform will also inform you of all the bidding rates for the 30 parking spaces currently being bid on, letting the public know the level of competition (e.g., are all 30 parking spaces priced at $4 or has a certain proportion of parking spaces already been raised to $5 or even $6).
To avoid trouble, users who have already parked their vehicles can continue to park at the original rate, but they also need to indicate the length of time they plan to lock in that rate and pay the accumulated fee for that length of time in advance for the platform to calculate potential parking space supply in the future. However, the longest lock-in period should be set separately depending on the characteristics of the area.
If the original reserved time is exceeded, the subsequent fee will be charged based on the market bidding price per hour and collected when the vehicle is retrieved. If the parking space supply exceeds demand in the area for an extended period of time, the fee may still be NT$1 per hour.
So the usage right for a parking space will actually have 4 status:
- Idle parking space
- Idle parking space (reserved): The platform will display the current bid amount for the reservation.
- Occupied parking space (locked at a certain rate): The platform will display the expiration time of the lock.
- Occupied parking space (paying the market rate)
Usage scenarios and interactive relationships
For private scooter riders, they can generally predict their parking duration and choose a suitable rate locking period. For commuting riders with fixed start and end points, the charges in different areas at different times should also be similar in the long term, allowing for advanced planning.
For shared rental platforms, another set of geographic data can be developed to estimate the parking duration that the platform should lock in when users returning vehicles to different areas, i.e. how long the vehicle is expected to be ridden away by the next user. However, theoretically, as the turnover rate of shared rental vehicles is higher, i.e. they are more likely to leave the parking space, the platform can also choose not to lock in parking rates and let parking fees follow market rates, and encourage users to ride away vehicles located in hot parking zones that are being charged high parking fees through rate deduction or even reward mechanisms.
For shared scooter riders, they need to consider whether the platform will pass on the cost of parking fees to them due to different parking spaces when returning the vehicle. One simple approach is to charge riders the market rate for one hour of parking in the parking space.
In summary, as with vehicle usage rights, the platform can also prompt users with historical data on the number of available parking spaces and the price range for successfully reserving a parking space. This creates a public real-time supply and demand market for parking spaces.
Contradiction with “What Money Can’t Buy”
I once read Michael J. Sandel’s book “What Money Can’t Buy: The Moral Limits of Markets”, which describes how the ethics of “first come, first served” is gradually being replaced by the ethics of “paying customers enjoying services”. In airports, theme parks, highways, and hospitals, the influence of money and the market is increasingly penetrating all aspects of life that were originally regulated by non-market mechanisms.
Should shared scooters and parking spaces follow the ethics of “first come, first served” or “paying customers enjoying services”?
In my opinion, following the concept described in the book “Radical Markets”, as long as the additional bidding prices of shared scooters flow back to the platform as rewards for users to help redispatch vehicles, and the additional bidding prices of parking spaces flow back to the government’s public transportation construction, the ultimate benefit of paying customers enjoying services is still shared by everyone.
