Essays on the elements of consumer journeys in retailing

Abstract

"In two essays, I examine how elements in online and offline consumer journeys can be used to study consumer behaviour. In the first essay, I bring in the network theory perspective as an element to study online consumer behaviour. Using the clickstream data, I show how consumer browsing can be interpreted as a network to gain richer insights. I hypothesize using the tenets of network theory, flow of information and orchestration of the flow. I show that when consumers browse, they visit websites multiple times, revisit websites, and sometimes visit only one website to make their decisions. This behaviour is depicted by using three network properties, i.e., total connections (TCs), betweenness centralization (BC), and network density (ND), to determine their average basket value. Further, I examine the moderating role of contingent variables, variety and volume of information on consumer online search behaviour. I find that the ABV decreases when the TCs and BCs increase. However, ABV increases when ND increases. To validate the results, I conduct several robustness checks (involving endogeneity, benchmark models, sample selection bias, omitted variable bias, alternative dependent and independent variables, and four product categories). In the second essay, I explore the elements contributing to offline in-store consumer journeys. I hypothesize using the stimulus-load theory from the store environment literature. By combining data from multiple sources, such as store sales and blueprints, I create a novel dataset and validate the importance of inter-department proximity (IDP). Results show that as the IDP of two departments increases, joint sales (combined sales of those two departments) first increase and then decrease, depicting an ‘inverted-U’ relationship. I consider three (external and internal) factors for the department in establishing boundary conditions for IDP. Results show that the type of departments and department layout positively impact the relationship between IDP and joint sales among external factors. Finally, with the internal factor, category differential, I find that the as the difference in the number of categories between two departments increases, the relationship between IDP and joint sales strengthens. I conduct several robustness checks (endogeneity, unobserved heterogeneity, benchmark models, multicollinearity, sample-selection bias) to establish the results."