Abstract
Researchers and practitioners have long been interested in the determinants of online consumer shopping behavior. Product presentations directly aid consumer understanding of products in an e-commerce context, and retailers often exert significant effort in polishing these presentations. For example, application (hereafter referred to as app) developers design impressive icons to convey brand impressions and attract consumers to download their apps. Photographers of online products also strive in improving their skills to take aesthetic photos, resulting in well-presented products that attract consumers to click. Despite such endeavors, research on the effect of product presentations on shopping behavior is limited. Most previous studies utilize an experiment-based approach, which deliver strict theories but cover only a small range of photo characteristics. Through advanced image-processing techniques, a large set of photo characteristics can be investigated simultaneously in an empirical study. This thesis exploits image-processing techniques to study how visual aesthetics of product presentation affect consumer behavior from three e-commerce environments. It investigates the effects of the aesthetic design of icons on app downloads, effects of photo aesthetics on consumer click-through behavior, and effects of photo background as well as the interaction effects of foreground and background on consumer click-through behavior.The first part of this thesis focuses on the effects of app icon on app sales. With the rapid development of the mobile app market, understanding the determinants of mobile app success has become vital to researchers and mobile app developers. Extant research on mobile apps primarily focused on the numerical and textual attributes of apps. Minimal attention has been provided on how the visual attributes of apps affect the download behavior of users. Among the features of app “appearance,” this part focuses on the effects of app icon on sales. With aesthetic product and interface design theories, icons are analyzed from three aspects (i.e., color, complexity, and symmetry) through image-processing technology. A dataset collected from one of the largest Chinese Android websites is used, and results show that icon appearance influences the download behavior of users. Particularly, apps with icons featuring high colorfulness, proper complexity, and slight asymmetry lead to many downloads. These findings can help developers to design their apps.
The second part studies the aesthetic effect of product photos on consumer click-through behavior. To rule out possible confounding factors, real company dataset from a social shopping website is used. This part employs a two-stage nested logit model and examines product photo characteristics such as color, composition, complexity, and model face. A differences-in-differences approach is also utilized to rule out unobservable factors such as style, popularity, and brand name. Results demonstrate that the three aesthetic aspects of photos (i.e., color, complexity, and spatial layout) affect the online browsing behavior of consumers. Particularly, consumers prefer to click product photos with warm colors, large key objects, and appropriate complexities. These findings can also provide guidelines to properly manipulate the four dimensions of photo aesthetics based on the click-through behavior of consumers.
The third part explores the aesthetic effect of the product photo background and interaction effects of foreground and background on consumer click-through behavior. To understand the effects caused by the background of product photos, the whole product photo is separated into major object and background areas. According to aesthetic theories, two facets (i.e., color and complexity) are identified to study the effect of background on the click-through behavior of online consumers. The random coefficient logit model is used to estimate consumer demand for photo characteristics. Results show that the aesthetic aspects of photo background and its interaction with foreground affect demand for clicking action, demonstrating the importance of arranging photo background aesthetically. Particular categories of products (e.g., clothes, shoes) are suggested to be presented on background with low lightness, low saturation, high colorfulness, high cool color ratio, and high clear color ratio. Products from different categories are suggested to be displayed on background with a different level of complexity. Clothes are suggested to be shown on a background with a low number of key points, high kurtosis, and high skewness. Shoes are suggested to be presented on a background with a high number of key points, low kurtosis, and low skewness.
The main contributions of this thesis are summarized as follows. 1) This thesis incorporates the recent trend of big data research by integrating heterogeneous data (multimedia and sales data) to address the problem of e-commerce demand. It integrates multimedia data from the e-commerce environment with transactional business data to address a marketing problem with significant effect, that is, online consumer shopping behavior. 2) This thesis provides a quantitative assessment of app icons and product photos that can be measured using image-processing software. It builds a framework that maps photo features extracted using image-processing techniques with theories developed in HCI and psychology. It also uses strict econometric analyses to establish a causal relationship. The interaction of economic and marketing hybrid models with image-processing technology from computer vision can benefit both research fields. 3) This thesis also has significant implications for practicing e-commerce, especially for developers and retailers. Results of this research can be used to predict the marketing effectiveness of provided product photos and direct the product presentations in e-commerce websites. They may also contribute to the development of software that can assist vendors in selecting icons and photos that can increase sales.
| Date of Award | 9 Sept 2016 |
|---|---|
| Original language | English |
| Awarding Institution |
|
| Supervisor | Xin LI (Supervisor) |