|کد مقاله||کد نشریه||سال انتشار||مقاله انگلیسی||ترجمه فارسی||نسخه تمام متن|
|5035003||1471785||2017||13 صفحه PDF||سفارش دهید||دانلود رایگان|
- We model online shoppers' visit and purchase behaviors across websites.
- We find that shoppers' cross-site visit patterns tend to be clustered.
- The website choice probability systematically varies within a visit cluster.
- The purchase propensity is significantly higher at later visits within a cluster.
- The cluster-based analysis can help retailers develop online marketing strategies.
Low transportation costs online allow shoppers to visit multiple e-commerce sites for a purchase decision. This research investigates online shoppers' visit and purchase behaviors across competing websites. To consider that shoppers' longitudinal cross-site visit data may consist of several unobserved shopping episodes, we propose a modeling approach to probabilistically clustering and relating online visits to latent shopping episodes, based on the temporal patterns of the visit events. The inferences are then used to examine shoppers' visit-to-purchase behavior across websites. Using Internet clickstream data on individual-level browsing and transaction records at major air travel sites, we find that online shoppers' cross-site visit patterns tend to be clustered and the purchase propensity is significantly higher at later visits within a visit cluster, compared to earlier visits. As our results suggest the possibility that visit clusters can serve as a reasonable proxy for shopping episodes, we look further into shoppers' website choice and purchase behaviors within a cluster. We discuss how the cluster-based analysis can help managers tailor online marketing and advertising strategies based on shoppers' cross-site visit and purchase patterns.
Journal: Journal of Retailing - Volume 93, Issue 3, September 2017, Pages 253-265