یک رویکرد مبتنی بر خطر برای مدل سازی تأثیرات خرید آنلاین در مدت زمان خرید اینترنتی A hazard-based approach to modelling the effects of online shopping on intershopping duration

Introduction Information and communication technologies (ICT) are transforming the retail sector with a proliferation of new channels and virtual alternatives. This change is observed in spending patterns: online retailing in the UK accounted for 14% in 2014 of all sales in 2014 up from 2.1% in 2002, Office for National Statistics 2015, and in the USA share of online retailing reached 7% of all sales up from only 2.8% in 2006 (Bucchioni et al. 2015). The growing prevalence of online shopping is driving major disruption in retail with poorly understood impacts on personal mobility and delivery operations. Historically, it has been difficult to study such impacts due to lack of sufficiently detailed data. In this paper, we present the use of consumer panel data from Kantar Worldpanel (KWP), which is a relatively new data source in this context. Consumer panels are attractive as they capture individual shopping episodes via both online and in-store channels. Further, market research companies that collect consumer panel data (e.g., Kantar, Nielsen, GfK, Ipsos) operate in multiple international markets, and hence this type of data is available for many countries and similar analyses can be repeated in different contexts and times. The focus of this paper is on modelling inter-shopping durations for grocery shopping using consumer panel data incorporating both online and in-store activities. The overall aim is to gain insights into how shopping patterns are changing with the increasing use of online channels for shopping. This will provide insights to substitution question between physical and virtual activity in the context of grocery shopping, and also contribute to multiday activity generation modelling work. The remainder of this paper is organised as follows. In the next sub-sections, we briefly review earlier studies on ICT’s impact on shopping related personal travel and present specific objectives and rationale behind using hazard-based models as well as our focus on grocery shopping. In the following sections, we lay out the methodology describing analysis methods and empirical data used, followed by presentation of the results and conclusions.