I am a PhD candidate at INSEAD in the Technology and Operations Management area. I am excited about the potential of new business models for smarter cities, be it bike-share systems, ride-hailing platforms, or delivery based models. In my research work I integrate consumer behavior, location and availability related operational concerns and their interrelation with environmental aspects. I have an interest in applying structural empirical models as well as advancing the methodology.
This paper compares the financial and environmental performance of two revenue models for the online retailing of groceries: the per-order model, where customers pay for each delivery, and the subscription model, where customers pay a set fee and receive free deliveries. We build a stylized model that incorporates (i) customers with ongoing uncertain grocery needs and who choose between shopping offline or online and (ii) an online retailer that makes deliveries through a proprietary distribution network. We find that subscription incentivizes smaller and more frequent grocery orders, which reduces food waste and creates more value for the customer; the result is higher retailer revenues, lower grocery costs, and potentially higher adoption rates. These advantages are countered by greater delivery-related travel and expenses, which are moderated by area geography and routing-related scale economies. Subscription also leads to lower food waste–related emissions but to higher delivery-related emissions. Ceteris paribus, the per-order model is preferable for higher-margin retailers with higher-consumption product assortments that are sold in sparsely populated markets spread over large, irregular areas with high delivery costs. Geographic and demographic data indicate that the subscription model is almost always environmentally preferable because lower food waste emissions dominate higher delivery emissions.
The cities of Paris, London, Chicago, and New York (among many others) have set up largescale bike-share systems to facilitate the use of bicycles for urban commuting. This paper estimates the impact on bike-share ridership of two facets of system performance: accessibility (how far the user must walk to reach stations) and bike-availability (the likelihood of finding a bicycle). Our analysis is based on a structural demand model for spatially differentiated products that includes distinct mechanisms for the short and long-term effects of bike-availability (via lost sales and increased user-interest, respectively). The bike-share context, and the distinct mechanisms require us to go beyond past work in incorporating real time changes in product (bike)-availability information, and including much finer data on potential demand sources. These enhancements render traditional estimation methods computationally infeasible; we transform our estimation from the time domain to the “local-stockout- state” domain to address this. Our estimates for the Vélib’ bike-share system in Paris suggest that a 10% increase in station density would increase ridership by 5.09% (±0.45%), while a 10% increase in bike-availability would increase ridership by 12.29% (±0.39%), three-fourths of which comes from fewer lost-sales, and the rest from increased user interest. We illustrate the use of our estimates in identifying neighborhoods and times to target for improvements, and in comparing alternate operational improvements and station networks.
Achieving scale is key to the efficacy, survival and eventual domination of peer- to-peer marketplaces. Marketplace operators often run aggressive promotions and incentive schemes to attract new users or increase the usage of existing users. This study quantifies and compares the effect of incentives given to the “buyer” side and “seller” side of the marketplace. Specifically, using data from one of the leading ride-hailing marketplace, we estimate the effect of passenger incentives and driver incentives on number of trips arranged through the marketplace. Further, the incentives can be designed in different formats, i.e. they could be given for every use (linear incentives) or could be given only upon a certain level of use (threshold incentives). We build a structural model to accurately capture the driver and passenger response to incentives, and the nature of incentives. We take into account the effect of service levels on passenger and driver behavior and their endogenous realization in the system, as well as the cross externalities and economies of scale effects. Driver effort on the platform is unobserved, for which we devise a novel local matching model based imputation method. We find that in short-term (current week) passenger incentives are more effective than similar driver incentives. In long-term (next 3 months), the opposite is true; driver incentives are more effective than passenger incentives. This change of effects is determined by the differential stickiness of passengers and drivers to the platform, as well as differential response to evolving service levels. When structuring driver incentives, it is more effective to use threshold incentives compared to linear incentives. The marketplace exhibits substantial economies of scale. For every doubling of passenger and driver numbers, the number of trips more than doubles, it increases by further 25%.
Cities differ on the nature of ride-hailing service demand flow in different directions and over time, creating different degree of asymmetry in demand. Using data from a ride-hailing marketplace, we estimate driver location decision behavior as a function of the evolving spatio-temporal demand and quantify the impact of coordinated location decisions for drivers on the efficiency of the system.
Demand flow directions in a bike-sharing System is asymmetric. Cities differ on the amount of asymmetry in demand because of city structure and commuter flow patterns. In this study we derive a measure of demand asymmetry for a system. We then characterize the effect of this demand asymmetry on the decisions of number of bikes and number of docks as well as the frequency with which these bikes will have to be reallocated from full stations to empty ones. This analysis should help guide the design of new bike-sharing systems.