Virtual store layout: an experimental comparison in the context of grocery retail

Adam P. Vrechopoulos, Robert M. O'Keefe, Georgios I. Doukidis and George J. Siomkos

Interface design and the interaction between customer and computer are factors critical to business effectiveness over the Web. A key aspect of conventional retailing is store layout; an analogous aspect in virtual retail is virtual store layout. The layout of a conventional store is a crucial determinant in the creation of its image, and it can play a central role not only in satisfying buyers’ requirements, but also in influencing their wants and preferences. Similarly, virtual store layouts have a critical influence on traffic and sales, and their characteristics determine buyers’ willingness to purchase through the Internet. Virtual store layout design constitutes one of the core elements of the virtual retail mix, and retailers should be aware of its effects on online consumer behavior in order to be able to manipulate it accordingly. Until now, however, there has been little empirical evidence regarding the impact of the virtual store layout on consumer buying behavior. Focusing on the impact of virtual store layout effects on consumer buying behavior, this paper tests whether conventional retailing store layout theory can be applied in its present form in the context of virtual retailing. To that end, the three most common conventional retailing layout types (grid, freeform and racetrack) were transformed into virtual layouts for computer-mediated interfaces over the Web and tested against real customers within a laboratory experimental setting. Specifically, subjects in Greece and the UK participated in a laboratory experiment: they were given a planned shopping task with money to spend, and performed their shopping through a virtual store with layout as the manipulated variable.

Conventional retail store layout affects a series of consumer behavior-dependent variables in a certain manner. The research hypotheses aim to test whether the same phenomena occur in a virtual setting, focusing on testing the effects of the three most common conventional layout types on consumer behavior. The aim is to investigate whether and how the layout affects online consumer behavior in terms of: (a) planned purchases facilitation, (b) easy navigation within the store, (c) perceived entertainment, and (d) time spent within the store.

A virtual store was developed in three alternative versions - each one employing a different layout - following two basic conditions: (a) the precise concept and navigation rules that govern each layout type in conventional retailing were to be transformed to their corresponding layout patterns in the virtual store, and (b) all other variables (eg background color, promotional banners, speed, product images, buttons, product display techniques, product prices, product assortment, etc) were to be kept the same, in order for the layout to be the only differentiating factor (ie, manipulated variable) among the three alternative versions of the laboratory virtual store. Despite the fact that Web development does not constitute the research area of the present study, the development of this virtual store was a prerequisite in order to test rigorously the respective hypotheses. To that end, several methodological steps were taken towards the development of alternative virtual store layouts that simulate the relevant rules and concepts identified in conventional store layouts, while retaining distinct differences from each other. The Object Oriented Hypermedia Design Methodology was employed for the transformation process.

None of the research hypotheses was confirmed. The results show that the layout significantly affects online consumer behavior, but that predictions generated from the literature of conventional retailing about differences in the outcome of layouts do not generally hold in a virtual setting and, therefore, a relevant adaptation is required. The freeform was found to be perceived as the most useful layout for conducting planned purchases, and the grid layout as the easiest to use. In addition, the freeform was found to be also perceived as the most entertaining layout. The racetrack and freeform layouts engaged subjects for longer, as would be expected from conventional retail theory. Some of the findings can be explained by reference to research in Human Computer Interaction (HCI). A mixed grid/freeform layout pattern constitutes an emerging layout for virtual retailing – a conclusion supported both by our empirical work and its use by Sainsbury’s and Tesco in the UK.

It is clear that existing virtual grocery stores are struggling with various layouts that are not theoretically supported. Given that many business failures have occurred because conventional strategies and techniques were “copy-pasted” into the world of Internet retailing, retailers seeking to design effective virtual shopping interfaces over the Web should adopt a multidisciplinary approach and be careful when using established conventional marketing and retailing principles and guidelines.