Product Photographic Presentation and Online Consumer Behavior: The Role of Image Features in E-commerce

Description

This project aims to use image-processing techniques to identify a set of product photofeatures and examine their impact on online consumers’ behavior in B2C e-commerce.Our motivation comes from the rise of B2C e-commerce in China (e.g.,Tmall) and HongKong (e.g., Yahoo Auction), in which retailers extensively use product photos to presentproducts and attract consumers. However, prior research provides retailers with limitedguidelines for the most effective photo presentation. For example, most marketingresearch examined product presentation in physical stores and advertisements.Information systems research mainly examined the effective design of websiteinterfaces. Hence, online retailers often rely on their or others’ experiences (such ashttp://goo.gl/dkkcQ) to prepare product photos, or they outsource product photopreparation to professionals (e.g., http://goo.gl/fTJOMC and http://goo.gl/OigZpd).This project surfs the Big Data wave, using image-processing techniques to provideonline retailers with guidelines for choosing the most effective product photos toinfluence consumer behavior. Unlike prior research that focuses on subjectiveperceptions, we will study objective photo features that do not rely on individualretailers’ judgments. We will obtain clickstream data from e-commerce retailers anddevelop econometric models to rule out confounding factors (e.g., brand, quality, wordof-mouth and price) and to assess photo features’ impact on consumers’ click behavior.In order to increase the validity of our findings, we will conduct experiments with humansubjects to cross-validate the empirical findings produced by econometric models.Additionally, the experiment setup allows us to explore how the photo features attractconsumer attention. As compared with existing Big Data studies that mainly focus ontext mining, this project has the potential to show the value of product photographicpresentation in e-commerce research.In summary, we plan to: 1) use image-processing techniques to identify a set of productphoto features; 2) build econometric models to gauge the relationships between productphoto features and consumer behavior; and 3) conduct experiments to validate anddeepen the findings on how product photographic presentation influences consumers’clicks and attracts consumers’ attention. Our ultimate goal is to advance both theory andpractice on effective product presentation in B2C e-commerce as a marketing tactic,which is of great value to e-commerce development in Hong Kong and China.