Online Trust: An International Study Of Subjects’ Willingness to Shop at Online Merchants, Including The Effects of Promises and of Third Party Guarantees
Our expectation was that promises alone would have little impact on customers’ WTP. Our thinking was that if customers trusted a website, then promises would not be necessary, and that if customers did not trust a website they would not trust its promises either; either way promises would have no effect. This turned out to be incorrect. Our expectation was that third party guarantees would be most effective. This turned out to be only partly supported. Our expectation was that China would be a significant outlier in one direc-tion, and Singapore would be a significant outlier in the opposite direction. Again, this turned out not to be entirely true. All of this is described more fully in our hypotheses (section 4) and our analyses (sections 7 through 11). Code Shoppy The US online market is the most mature online market in the world, and the behavior of online shop-pers in the US is remarkably similar to the behavior of shoppers in traditional physical venues. Shoppers at a range of physical stores, from Wal-Mart and Rite-Aid to Tiffany’s and Brooks Brothers, make their purchas-es without requesting promises that the goods they buy are authentic, and without requiring third party guaran-tees. The online behavior of US shoppers seems simi-lar to their offline behavior. In contrast, promises and third party guarantees are significant in other markets, with vendor type and nature of assurances interacting in different ways in different countries. Promises were not always without value, and third party guarantees did not always increase the value of promises.
Hypotheses Our discussions with executives at 360buy, Taobao, and Yihaodian suggested the following: (1) vendors’ reputational capital may be the greatest de-terminant of consumers’ WTP; (2) vendors’ promises of product quality are of little value from low quality vendors because buyers may not view these promises as binding in the absence of reputational capital; (3) third-party guarantees would help create greater trust and faster growth in online markets, (4) Chinese con-sumers may be less trusting and Chinese vendors may encounter greater trust penalties; and (5) different risk mitigation mechanisms may have different effective-ness in China, due to different consumer experience with online commerce generally, with third party war-ranties and performance bonds, and with the legal system when trying to follow up on complaints. These led directly to our six hypotheses.Hypothesis 1: In every country, with purchases aggregated across all product types, as vendor quality decreases, consumer’s WTP will decrease. The idea that trust is directly observable through subjects’ expressed WTP is central to our experiments. This is critical to establish, because if we cannot ob-serve this in the laboratory then our experiments can-not test any of our more important hypotheses. Hy-pothesis 1 attempts to determine the extent to which reputation matters to subjects’ WTP for items from different categories of seller. Hypothesis 2: In every country, with purchases aggregated across all product types, promises alone will not increase consumers’ WTP compared to these consumers’ WTP for the same products purchased from the vendors but without promises. (Promises alone refers to promises offered without 3rd party guarantees.) Hypothesis 2 suggests that promises from low quality vendors will not be credible. The value of promises alone is directly observable through differ-ences in subjects’ WTP for products online, from a range of merchant types, with and without promises. The extent to which differences in WTP vary by mer-chant type and merchant reputation will indicate the role of reputational capital in determining the value of promises. A promise from Carrefour or Wal-Mart is credible because they have reputations that they do not want to damage; they are likely to back up their prom-ises. Alternatively, if the merchant is unreliable then the promise will probably be seen as unreliable as well. Hypothesis 3: In every country, and across all product types, consumers’ WTP for products from vendors who offer promises backed by a trusted third party will be greater than consumers’ WTP for the same products from the same quality vendors who offer promises not backed up by a trusted third party. Hypothesis 3 suggests that promises will be effec-tive only if they are backed up by third party guaran-tees. This hypothesis interacts in complex ways with hypothesis 2. In particular, this suggests that promises from low quality vendors will be credible if they are backed up by third party guarantees, even if these promises alone are not credible (consistent with hy-pothesis 2). Additionally, hypothesis 3 suggests that promises from high quality vendors are also more effective if backed up by third party guarantees, even if these promises alone are not credible (which is not consistent with hypothesis 2). Combining hypothesis 2 with hypothesis 3 suggests that vendors without established reputations could benefit early in their online evolution through the use of third party guaran-tees. The value of third party guarantees is directly observable through differences in subjects’ WTP for products online, from a range of merchant types, with and without third party guarantees. Hypothesis 4: For each vendor quality level, the decreases in consumers’ WTP associated with ven-dors’ quality level will vary across countries. Hypothesis 4 suggests that the impact of trust and reputation on WTP will vary by country. Based on discussions with online retailers, we expected trust to be less of a problem among consumers in Singapore, with consumers in Germany and the United States exhibiting intermediate behavior. We expected online trust to be the greatest problem in China, given the incidence of problems with counterfeit and defective merchandise, and with product tampering in China. We test this using subjects’ stated WTP for merchan-dise from all vendor types under the baseline condi-tion, with no explicit promises or third party guaran-tees. Hypothesis 5: Across vendor quality levels, ven-dor promises without 3rd party backing will increase consumers’ WTP in different degrees across countries. Hypothesis 6: Across vendor quality levels, ven-dor promises with 3rd party backing will increase consumers’ WTP in different degrees across countries. Hypotheses 5 and 6 suggest that mechanisms to increase consumers’ confidence and to increase con-sumers’ WTP will vary across nations. We expected variation across nations, based on national differences in experience with online shopping, the severity of problems with traditional physical shopping, and the effectiveness of national legal codes in creating a sense of consumer protection.