The Role of Sellers’ Reputation in China’s C2C Market: Theory and Evidence from Taobao

By JIN WANG

Over the past couple of decades, China’s online consumer-to-consumer (C2C) market has expanded dramatically with transactions volume exceeding 40 billion Yuan in 2010. This unique case of extremely fast development in China’s C2C market provides researchers with the perfect opportunity to study a number of interesting research questions. In traditional markets, buyers can get information regarding the quality of goods through repeated personal interactions with sellers and merchandise. The electronic markets, by contrast, tend to be more information-asymmetric, especially when goods are sold anonymously in the C2C markets. Online buyers usually have no tangible interaction with the products or sellers. So it would be interesting to study how to achieve efficiency in these markets with better quality goods receiving more sales.

My research subject is Taobao.com, the largest online C2C website in China. Taobao doesn’t sell product itself. It mainly serves as a platform in which sellers and buyers can trade with each other. People who want to sell goods can apply for an account on Taobao. Although Taobao usually asks applicants to provide the authorized ID and other personal information for registration, this information is private and the buyers can not observe it during their shopping. To solve the asymmetric information problem, Taobao offers a rating system in which sellers and buyers can evaluate each other after the trading. The evaluation for sellers includes the quality of the goods, whether the descriptions of goods on sellers’ websit are consistent with the real ones, service quality of the sellers and so on. The evaluation for buyers is mainly about whether buyers have paid money in time. Each seller or buyer’s overall reputation, which is calculated from all rating reports he has received, is shown to all participants on the website. Buyers, although cannot directly observe the quality of goods, can use sellers’ reputation to judge sellers’ honesty and the quality of goods, and therefore decide their purchase decision. In other words, sellers’ reputation here serves as a signal which can help buyers to distinguish sellers who provide high quality goods with those who provide low quality goods. 

This reputation system is not unique to Taobao. There are many US online C2C websites including eBay that have adopted similar reputation systems. The studies done on these websites presented controversial results regarding the relationship between sellers’ reputation and sales. Some find that bid amounts barely increase as sellers improve their reputations, if there is any increase at all (Eaton (2002), Houser and Wooders (2001), Lucking-Reiley et al. (2000), Mc- Donald and Slawson (2002), Melnik and Aim (2002), Resnick and Zeckhauser (2002)). While a recent study by Livingston (2005) shows that sellers on eBay are strongly rewarded for the first few reports that they have behaved honestly. Currently little is known about the role of the reputation system in the Chinese C2C markets, and my study plans to fill in this blank area.

The study on Taobao’s reputation system also has practical applications. Due to differences in culture, social structure, income distribution and consumer preferences, the consumer behavior in online market could be quite different in China and the U.S. Investigating the role of the reputation system in Taobao provides a way to understand the consumer behaviors in the online C2C markets of China. This would help U.S. companies who want to participate in this market to adopt proper business strategies.

The key question I want to answer in this project is the relationship between seller’s reputation and sales including price and quantity. To better understand this question, I begin by writing a simple theoretical model to illustrate the mechanism of reputation on sales. I use the framework of signaling model by Spence (1973) and extend it to the setting of online market. I assume that there are two types of sellers: high type seller and low type seller who are different in their abilities to fulfill the online transactions. The high type seller is more capable than the low type seller to fulfill the transactions successfully, for example, to deliver the goods in time. The sellers’ type is private information which is only known to sellers. A seller can signal its type through reputation which is based on the reviews the seller receives from previous transactions. The mechanism is as follows: we assume establishing reputation costs sellers’ efforts. For instance, a fast delivery of goods usually makes a seller to get a higher rating from buyers. To guarantee the speed of delivery, a seller needs to pay a higher delivery fee or work more efficiently. Since the high type seller has a lower marginal cost. It is much easies for them to establish a good reputation than the low type. For the buyers, when they observe a seller with a higher reputation, they are more likely to believe this seller is of high type, and thus are more willing to pay a higher price. We can show that, under some conditions, there exists a separating equilibrium that the high type seller chooses to build a high level of reputation and thus be able to sell his goods at a high price, while the low type seller gets low reputation and sell at a low price.

In this model, the reputation works as a signal that helps the buyers to distinguish different types of sellers. Buyers believe that the high type seller is more capable to build a good reputation than the low type sellers, therefore buyers would like to pay more for a high reputation. This motivates sellers to build a high reputation. But low type sellers find that it would cost too much if they try to establish the same reputation as high type sellers do. So the game finally ends up with high type seller getting a higher reputation and prices than the low type. An implication from this model that can be empirically tested is whether a seller with higher reputation can sell his goods at higher prices or sell at a larger quantity if the prices are competitively similar. 

To carry out the empirical analysis, I need to get data from Taobao.com about seller’s reputation, prices, sale amount and other characteristics of the seller. With the support from US-China Institute, I was able to go back to Beijing, China in July for data collection. The first place I visited is the research center of Alibaba Group in Beijing. Alibaba Group is the parent company of Taobao. Its research center manages the research and data support related to Taobao. There I met a manager of research center and explained to her the purpose and detailed plan of our project. She provided me lots of useful suggestions and information. For example, she suggested me using online shopping data instead of auction data for the reason that the data of online shopping is much richer than auction data on Taobao. My research results would be more robust and convincing if it is based on the data with large sample size. She also introduced a lot of applications that Taobao adopts to facilitate the transactions between seller and buyer. This makes me to consider investigating how these applications interact with sellers’ reputation to influence the prices and sales in my project. As for data support, unfortunately, Alibaba research center temporarily cannot provide the data I requested due to the concern of business privacy.  Although I had tried several times, it still didn’t work. During a talk to some professors in Peking University, I got to know that it is possible to write a computer program to record the information on the website of Taobao. I then consulted several students who are majored in computer science for the feasibility of this task. Besides, I spent some time in the library of Peking University to look up the books about computer programming. Due to the complication of code programming and time constraint, I finally decided to hire an experienced person to write the code. Now the programming work has started and it is expected to be finished before December.  With the help of this program, I will be able to get a panel data about the evolvement of sellers’ reputation and their sales information on Taobao. These data will help me to analyze the role of sellers’ reputation on the online shopping.

I have got lots of benefits from this summer field project. I am deeply indebt to the generous support from USC US-China Institute. I also thank Alibaba Research Center and professors at USC and Peking University for helpful comments and suggestions.

Click here to view projects of other 2011-2012 USCI Graduate Summer Fieldwork Grant receipients.