时间的价值:基于共享出行平台的实证研究

The Value of Time:An Empirical Study on a Ride-Sharing Platform

消费活动往往伴随着一定的时间要求.有些消费行为本身就具有时间维度,如咨询服务的时长、交通方式的快慢.在另外一些场景中,购买商品需要消耗一定的等待时间,如游乐场和商店门口排长队的时间、网购的物流时间、网约车等待接单的时间等.随着消费场景的多样化的出现,商家在设计产品组合的时候,也会考虑对应的与时间相关的产品策略,如"绿色通道票""次日送达""优先呼叫"等.本文通过使用经脱敏和标准化处理后的出行平台用户浏览和出行订单数据,研究打车等待时间如何影响消费者的出行消费决策,测量出行场景中消费者的时间价值.我们发现,经标准化处理后的时间和价格的边际替代率为0.97,即节省1个时间单位的等待时间对乘客来说等价于0.97个价格单位的价格优惠,而相比之下,车型升级等价于0.4个价格单位的优惠.同时,等待时间的边际效用是非线性的,并且存在拐点:当等待时间少于1.5个时间单位时,时间的边际效用随时长递减;然而,当等待时间多于1.5个时间单位时,时间的边际效用随时长递增.通过进一步研究,我们发现拐点会随着出行时段变化而改变.在工作日和高峰时段的出行中,拐点发生的时间更晚.本研究通过使用真实交易数据对用户时间价值进行的量化以及对其非线性属性进行的刻画,既丰富了时间价值领域的实证研究,同时还可以为企业进行时间维度的产品优化提供丰富的管理策略启示.

Consumption is usually associated with time spending. Sometimes the length of time itself is part of the products, for example, the length of consulting services or time duration of transportation. In other scenarios, there is a minimum requirement of time, or waiting time, that customers have to spend during their consumption process. Examples include the long queue outside an amusement park, shipping time after placing an order on Amazon, and waiting time when hailing for a ride on the road. Marketers have started to offer various types of time-related service packages correspondingly, such as premium pass, 1-day delivery, and priority order. In this paper, we use anonymized and normalized browsing and order data from Didi--an online ride-hailing company, and estimate the value of waiting time in riding scenarios. We find that the value of saving one standardized unit of waiting time is equivalent to lowering 0.97 standardized unit of price for an average passenger. As a reference, upgrading to a higher level of car type is equivalent to lowering prices by only 0.4 standardized units. The marginal value is decreasing when the time horizon is short, but there exists a turning point at about 1.5 standardized time units. When the waiting time is less than 1.5 units, the marginal utility is decreasing;whereas when it is longer than 1.5 units, passengers exhibit an increasing marginal value in waiting time. Furthermore, we find a larger turning point in peak hour orders(relative to off-peak hour orders),and in weekday orders(relative to weekend orders). Our study provides new insights on understanding the value of time and is also practically relevant for managerial decisions on time-related product design and pricing.