Scene Sells: Why Spatial Backgrounds Outperform Isolated Product Depictions Online

ABSTRACT

Online retailers work to optimize product presentations; this study relies on scene perception theory to predict that product images with spatial backgrounds lead to more favorable consumer responses than isolated product images, in which products appear to be floating in the air, because spatial backgrounds facilitate visual processing and aesthetic experiences. Three experiments test these predictions and reveal that consumers evaluate products more favorably, express higher purchase intentions, are willing to pay more, and are more likely to choose a product depicted on a spatial background, even when product characteristics were considered. This effect is robust to short and long exposure times and for durable and consumable products. It is mediated by processing fluency and aesthetic experience and moderated by two variables: shoppers’ purchase involvement and online experience. These findings challenge an established industry practice to use isolated images and suggest that retailers instead should depict products in scenes to facilitate consumers’ perceptions and enhance their own revenues. Moreover, this study highlights the importance of scene perception for efficient and effective information processing in online retailing.

KEYWORDSAND PHRASES:

• E-retail
• online aesthetic experience
• online product depictions
• online scene
• online spatial background
• processing fluency
• scene perception

Notes

1. In line with other studies on perception and fluency [42, 45, 53], and given that it is difficult to conceptually delineate the different types of fluency detailed in the pertinent literature (which display subtle differences in meaning in scope) we use the general term processing fluency.

2. Prolific combines rigorous recruitment standards with relatively low costs. Moreover, it has some advantages compared with other crowdsourcing platforms such as MTurk; for example, participants are more naïve, less dishonest, and more diverse, and data quality is superior [55].

3. Such a choice situation arises when consumers browse for products using search engines, and so on, or different online shops simultaneously. The product presentations might differ markedly, even if the search is limited to one online shop, because certain products might be highlighted or displays vary across specific product collections. This choice situation also exists when consumers form consideration sets and compare several products from different retailers across multiple browser windows. They then must decide which windows to close, which products to compare further, and finally which product to buy.

4. Another pretest of the manipulation determined whether participants perceived products depicted in isolation differently than products on a spatial background. We avoided including this manipulation check in the main study because previous studies of aesthetics and fluency show that making people aware of the manipulation can hinder or even reverse its effect [75]. In this pretest, we randomly assigned 25 students to isolated or spatial background product depictions displayed on laptop computers in a laboratory setting. Each participant viewed four products (two chairs, two bags) and rated their perceptions with one item, from 1 (the displayed product seems to rest on a surface) to 7 (the displayed product seems to float in the air) [5]. The mixed ANOVA confirmed that participants perceived products depicted in isolation as floating in the air (M = 3.03) more so than products depicted on a spatial background (M = 1.72), F(1, 23) = 41.08, p = .004. We conducted similar pretests for Studies 2 and 3; the detailed results are available on request.

5. Including the outlier does not change the pattern of results.

6. We also compared processing effort in Study 2 across the high and low involvement products, using the time the participant took to answer the questions as measures. On average, they needed 642 seconds to answer questions about the high involvement product, compared with 423 seconds for the low involvement product. Wilcoxon tests confirm that time needed to answer differs significantly, Z(173) = −1.98, p = .048, offering further confirmation of the successful manipulation of purchase involvement [3].

Additional information

Notes on contributors

Daniel Brylla

Daniel Brylla ([email protected]; corresponding author) is a Ph.D. candidate in the faculty of economics and business administration of Friedrich-Schiller University Jena, Germany. He received a master’s degree in computer science and has 21 years of industry experience as a designer in the e-commerce domain. His current research interests focus on online product presentations and the design of e-commerce platforms. His research was published in several journals and proceedings, such as Electronic Markets, International Conference on Information Systems, and European Conference on Information Systems.

Gianfranco Walsh

Gianfranco Walsh ([email protected]) is a professor in the faculty of economics and business administration of Friedrich-Schiller University Jena. He received his Ph.D. and Habilitation from the University of Hannover, Germany. His work has been published in, among others, Academy of Management Journal, British Journal of Management, Information & Organization, Journal of the Academy of Marketing Science, Journal of International Marketing, Journal of Consumer Psychology, Journal of Service Research, and Psychology & Marketing.