Balancing Profit and Traveller Acceptance in Ride-Pooling Personalised Fares

Abstract

In a ride-pooling system, travellers experience discomfort associated with a detour and a longer travel time, which is compensated with a sharing discount. Most studies assume travellers receive either a flat discount or, in rare cases, a proportional to the inconvenience. We show the system benefits from individually tailored fares. We argue that fares that optimise an expected profit of an operator also improve system-wide performance if they include travellers' acceptance. Our pricing method is set in a heterogeneous population, where travellers have varying levels of value-of-time and willingness-to-share, unknown to the operator. A high fare discourages clients from the service, while a low fare reduces the profit margin. Notably, a shared ride is only realised if accepted by all co-travellers (decision is driven by the latent behavioural factors). Our method reveals intriguing properties of the shareability topology. Not only identifies rides efficient for the system and supports them with reduced fares (to increase their realisation probability), but also identifies travellers unattractive for the system (e.g. due to incompatibility with other travellers) and effectively shifts them to private rides via high fares. Unlike in previous methods, such approach naturally balances the travellers satisfaction and the profit maximisation. With an experiment set in NYC, we show that this leads to significant improvements over the flat discount baseline: the mileage (proxy for environmental externalities) is reduced by 4.5% and the operator generates more profit per mile (over 20% improvement). We argue that ride pooling systems with fares that maximise profitability are more sustainable and efficient if they include travellers' satisfaction. Keywords: ride-pooling, personalised pricing, individual discounts

Figures (7)

• Figure 1: Method at glance: We use a batch of trip requests (origin, destination and time) from the population with a known distribution of behavioural parameters. We construct a dense shareability graph to identify all potential pooled rides. Next, we conduct calculations for each ride. We construct the acceptance probability as a function of a fare for each traveller. To calculate the expected profitability (at the ride level), we recognise all acceptance/reject configurations with corresponding weights (probability). We optimise the fare considering the trip characteristics and acceptance probability functions to maximise the expected profitability (objective). Finally, we select an offer, i.e. subset of pooled rides that maximises the objective.
• Figure 2: Density distribution of discounts resulting from the personalised pricing algorithm. We separately aggregate for all rides in the shareability graph (blue) and those selected for the offer (orange). Each bin represents a $5\%$ interval. While most rides in the shareability graph receive only the guaranteed minimum ($5\%$), those selected for the offer are associated with higher discounts.
• Figure 3: Degree distribution in offers constructed with 3 pricing strategies: personalised, flat sharing discount of $0.15$ and of $0.2$. The personalised approach favours more complex rides: only $14$ of $150$ travellers are offered a non-shared ride.
• Figure 4: The expected profitability under pricing strategies: personalised and two flat discounts. At each vertical line, there are three points: blue, orange and green. They represent the expected profitability for the same shared ride under the three strategies, respectively. The red dotted line denotes $1.425$ threshold: the expected profitability of a discounted ($5\%$ guaranteed discount) private ride. The personalised pricing leverages the local optimum, hence, for each feasible ride, the expected profitability is improved over the baselines.
• Figure 5: Distance reduction if a shared ride is realised (x-axis) against the expected profitability (y-axis). Dots represent the individual ride's characteristics: in blue all rides in the shareability graph and in red those selected for the offer. To contribute to the system performance, rides need to be well aligned and offer mileage reduction. First efficient combinations (over $1.5$ profitability) appear around $10\%$ distance saved. Upward the threshold, the trend resembles an exponential growth where each small increment in the mileage reduction results in a significant boost in the expected profitability. Around half of rides in the shareability graph provide at least $20\%$ distance reduction.

Theorems & Definitions (2)

• Theorem 1
• proof