Omnichannel Fulfillment Dilemmas: Customer Preferences and Manager Perceptions

Abstract

Problem Definition : Omnichannel retailers face hard choices when they decide how to improve the way they serve their customers. They use such levers as improving delivery and return policies, providing access to detailed product information and reviews, and offering lower prices. Because it is difficult for firms to excel on all dimensions simultaneously, it is crucial for them to have a profound understanding of the trade-offs consumers make when evaluating their offering.

Academic/Practical Relevance : If managers have biased perceptions of consumer trade-offs, these biases can represent serious obstacles to designing a winning operational strategy. We investigate the degree and nature of such managerial biases and propose a remedy for them.

Methodology : We use a state-of-the-art empirical approach to measure consumer prefer- ences for five dimensions of online channels, part of a larger omnichannel ecosystem (delivery policy, return policy, product information, branding, and pricing), and four product categories.

Results: We reveal that managers with experience in these categories have biased perceptions of what consumers prefer. Our analyses also show that it is very hard to identify experts with more accurate knowledge based on characteristics observable to the firm (e.g., experience, tenure, gender). Averaging predictions across individual managers show benefits from the so-called “wisdom of the crowd” and help to overcome individual biases.

Managerial Implications: Across the four categories, the crowd’s predictions outperform more than 96 percent of the individual managers’ predictions, resulting in more than 17 percent increased accuracy over the average manager. Our results also show that groups of as few as 5 to 10 managers already make a smart crowd, enhancing the feasibility of this strategy to overcome individual managerial biases in omnichannel retailing.

Keywords: omnichannel, marketing-operations interface, managerial biases, conjoint analysis