Working Paper

Auctions versus Posted Price Internet Channels: A Seller’s Perspective on When to Make and When to Take

Report Summary

In today’s marketplace, firms face a dizzying array of routes to market. Sellers in multi-generational, heterogeneous product markets must choose the appropriate channel or sequence of multiple channels to maximize their margins and probabilities of sale. Such markets might include, but are not limited to, goods such as used products, collectibles, homes, and even dating partners.

Here, Ernan Haruvy, Sandy Jap, and Robert Zeithammer develop a model to explain how firms should choose between auction- versus Internet-based channel formats. Based on data from the wholesale used car market, they propose a theoretical model by which sellers can maximize their profits by managing their product placement choices between an auction channel (which yields higher probability of sale with lower margins) versus an Internet channel (which yields lower probability of sale but significantly higher margins against a high cost of delay).

This multichannel context raises a number of intriguing issues. How should the seller choose among these channel options and what are the circumstances under which the choice of each channel dominates? In other words, how can the seller optimally manage the tradeoffs between the probability of sale and expected margins against a high cost of delay? Does it make sense for the seller to use a combination or sequence of channels in its market approach?

The authors advance a structural model to inform how the seller must trade-off the costs and benefits of one channel over another and identifies their options within the context of a three-period game. The model predicts the impact of changing channel listing fees and ascertains the period in which the seller will switch its selling strategies across the channel options as well as the timing and duration that is optimal for each offering. The model also improves the seller’s price premium and sales probabilities across these channels by identifying and incorporating systematic channel and product characteristics (such as car condition, seasonality, and the role of past sale failures).

Finally, the authors use a proprietary dataset to show that the theoretical model resonates with the real institutional context and offers empirical support for its various assumptions. Their investigation suggests that such models will allow sellers to develop selling strategies based on channel selling costs, derive implications for market and channel design, prescribe and identify key switching points in a go-to-market strategy, and connect channel costs to the selling price across channels. It offers a starting point for further investigations into optimal channel strategies for forward-looking, profit-maximizing sellers faced with multiple routes to market.

Ernan Haruvy is Associate Professor of Marketing, University of Texas at Dallas. Sandy D. Jap is Goizueta Term Chair Professor of Marketing, Emory University. Robert Zeithammer is Assistant Professor, University of California At Los Angeles.

Acknowledgments
This research was supported by a financial grant from the MSI-WCAI (Wharton Customer Analytics Initiative) “Modeling ‘Multichannel’ Customer Behavior Competition,” #4-1644 and the Mack Center for Technological Innovation at the Wharton School at the University of Pennsylvania. We thank the anonymous market maker for data support.

Comments from members