The industrial chain is deeply cultivated to control transport capacity, and diversified exploration to build ecology

The total capacity is the total constraint of all the above factors, and the control capacity is the core barrier: the spatio-temporal matching ability of the capacity determines the ordering efficiency and service experience of the online car booking platform, and the current capacity improvement is mainly achieved by providing drivers with incentives.

First-mover advantage needs to precipitant cross-edge network effect, which is obvious for scale and efficiency improvement: in the early stage of industry development, due to low demand and supply density, it is often difficult to match the time and space of demand and supply, and the ordering time is long and the distance is long, which seriously affects the supply and demand experience. After years of market education, the network car head platform has accumulated a large enough cross-side network effect, which has brought about the improvement of demand and supply density, which directly improves the efficiency of supply and demand spatio-temporal matching, and brings a better taxi experience.

For passengers, a waiting time of 3-5 minutes requires a high-density, extensive network of transport capacity and a dispatch system supported by strong computing power to quickly calculate the optimal route. For the driver, the order includes the receiving distance and the travel distance, and the long sending distance will lead to the reduction of the driver's order completion rate per unit time, and then affect the unit time income.

In the early days of online car booking, the linear distance (without considering traffic jams and one-way street factors) dispatching system was adopted, and the main dispatching logic was distance priority. With the development of mapping technology, the Estimated Arriving Time (ETA) system was gradually changed. Based on the real-time map, the estimated arrival time was calculated by taking into account traffic jams, one-way streets and other factors. ETA ordering technology has extremely high requirements for maps and algorithms, so the ordering system (algorithm) is the core technology of the ride-hailing industry.

The order dispatching logic of Didi is the global optimal solution of GMV maximization, but the transport capacity in peak demand is the only constraint, and the function of order dispatching logic is not obvious: supply and demand forecasting, route planning, service sub-(driver service evaluation system), cloud computing and machine learning, and the objective function is the global optimal solution of GMV maximization. However, in the case of limited transport capacity, the effect of this strategy is not obvious: the vehicle profit ratio reaches the peak, the supply side of the transport capacity becomes the core constraint, and the analysis and matching of supply and demand is not important.

The core of the network car construction barrier lies in the control ability of transport capacity, and there are two ways to improve the supply side transport capacity control:

1) Diversified products supplement peak capacity;

2) Redo and deepen: lengthen the supply chain length and increase complexity, deeply participate in the upstream of the industrial chain, and improve transport capacity control through vehicle control when users and drivers are difficult to control.

Path 1: Free ride and carpooling are effective products to solve the shortage of capacity and improve the control of capacity, while also lowering the price threshold and harvesting price sensitive customers. Carpooling aggregates demand in time and space, and concentrates passenger demand on the main trunk line for matching, which greatly improves the single loading capacity of the driver. In the process of carpooling, there will be a combination effect: the more overlapping routes, the more obvious the combination effect.

Uber's Uber Express Pool product is based on the above principle: Passengers who choose express pool need to wait for 2-3 minutes first, the system calculates the passenger's boarding point, and passengers walk to the boarding point (2-3 blocks) according to the map instructions to get on the bus. When they arrive near the destination, passengers get off the bus and walk several blocks to reach the destination.

Uber Express Pool attracts more passengers to use ride-sharing through low price and improves order intensity; Through passengers walking to the boarding/alight point, the demand is aggregated in space, and the driver only needs to pick up passengers on the main road, which greatly reduces the detour time and improves the proportion of route coincidence.

The actual operation is more complicated, and Hitch faces regulatory and profitability challenges. In the actual operation of Hitch business, the low acceptance of carpooling, the changing of actual road conditions, and the low accuracy of maps lead to the weak effectiveness of pooling, and both drivers and riders are in need of subsidies for a long time. Uber and Didi's carpooling business has not yet achieved profit.

Path two: Deepening the supply side industrial chain. Enrich the depth of the supply chain, and enhance the B-end stickiness and capacity control through value-added service price value creation. By continuously enriching the supply chain on the supply side, the platform improves the richness of the product side and the complexity of the supply chain.