Communications of the ACM

Virtual extension

Using Insurance to Create Trust on the Internet

By Fang-Fang Tang, Marcel G. Thom, Liang Toon Wang, Joai Ching Tan, Wai Yee Chow, Xiaoyun Tang
Communications of the ACM, December 2003, Vol. 46 No. 12, Pages 337-344
10.1145/953460.953519
Comments

The idea that trust is a stumbling block in the growth of Internet activities is not a new one. As many researchers have recognized, lack of trust can create problems for commerce in general, and for e-commerce in particular [3]. In 1997, one of the first e-commerce market surveys showed that consumers had strong privacy concerns about shopping on the Internet [12]. Subsequent studies unanimously demonstrated that lack of trust continues to be an impediment to e-commerce growth [2, 4, 5, 7, and 10]. As written in the November 2000 issue of the Communications, "The managers of Internet stores should, rather than focusing on shopping convenience, begin thinking about how to reduce the risk perception" and "look for mechanisms . . . to signal the security of financial transactions on the Web" [1].

To negate the effects of mistrust, the Internet industry devised various ways to cope with security concerns. Many companies began using firewalls and digital encryption technologies, and some companies collaborated in building shared centers for cyber security and in using public key infrastructures. The Trust Certification was developed to assure consumers that certified Web sites meet detailed security requirements. Although progress is being made, these attempts at allaying the Internet security concerns of consumers and businesses only go a limited way.

In this article, we focus on another angle—using insurance as a proxy in order to create trust for e-commerce activities on the Internet. Already suggested in the industry press [2], this method for creating trust may also reduce the possibilities of government intervention due to recent "spectacular frauds" [9].

Back to Top

Why Insurance?

The security issue is more than a technology problem, so it cannot be resolved by technological advancements alone. This is where insurance can help.

For example, in traditional international trading, the credibility of both the buyer and the seller is critical to a successful trading relationship. Business is conducted via each party's merchant banks, and trust is built slowly over time. A small, nonestablished player is at a great disadvantage. E-commerce can level the playing field in terms of pricing and size, but that credibility must still be established. This is especially true for global trading, because of the diminished role of government in the transaction. With the proper insurance, these concerns can be effectively allayed (Figure 1).

Back to Top

Types of Trust

There are three basic forms of trust in trading, and any insurance solution must be able to address all of them:

• Marketspace trust. The trust that both seller and buyer must have in the marketspace where the transaction will occur.
• Buyer's trust. The trust that the buyer has that the goods will be delivered as agreed.
• Seller's trust. The trust that the seller has that he or she will be paid for the delivered goods.

Proposed marketspace trust solution. To resolve marketspace trust issues, we propose a Comprehensive Marketspace Policy, which provides coverage in the following areas:

• Web site security insurance provides assurances to your trading partners that your Web site is secure, that confidential data they post is directed to the appropriate persons and is kept confidential, and so on. The insurance company may evaluate the strength and weakness of any particular Web site before accepting any risk.
• Errors & Omissions (E&O) content insurance provides assurances to your trading partners that your Web site's content is accurate and reliable. Accidental errors and omissions are covered by this policy; intentional errors and omissions may allow the insurance company the right of compensation after making payment to the aggrieved party.
• Trading partner identity insurance provides assurances to your trading partners that you are who you say you are. All participants in a given marketspace should be scrutinized, which is difficult to do when most (95%) Web users decline to provide personal information to Web sites when asked, and 40% of those who do provide demographic data, fabricate it [6]. A third party, such as an insurance company with a large market capitalization, is in a better position to acquire such information about individuals or companies because of the principle of utmost good faith. Utmost good faith dictates that all insurance transactions are conducted on this basis. That means insured and insurer must declare all necessary information even when they are not asked, if it is deemed important enough for the other party to make a decision based upon it. Failure to exercise utmost good faith in any insurance transaction allows the aggrieved party to seek damages or even void the contract altogether.

Proposed buyer's trust solution. For buyer's trust, we propose a Comprehensive Guaranteed Delivery Policy. The ordered goods must be delivered as promised, in the agreed-upon place, in the agreed-upon manner.

Under contract laws, if the delivered goods differ from the ordered goods, there is no contract because there was no consensus ad idem. Here, an insurance company might conduct product inspections before they are shipped or even before they are sold. In this way, trust can be created and the online buying experience may even be better than the traditional commercial transaction. The insurance company can include shipping insurance to cover the usual risks associated with transporting the product. All this can be done seamlessly to ensure that the provided guarantee is comprehensive.

Proposed seller's trust solution. For seller's trust, we propose a Comprehensive Guaranteed Payment Policy that secures payment for the shipped goods. Since the insurance company is already collecting information to verify the participant identities in the marketspace, it is also suitably placed to check the credit rating of buyers and, if necessary, collect payment.

Since most consumers use credit cards to pay for online transactions and the credit card companies ensure that payment is made to the sellers, seller trust is not as big an issue as marketspace trust and buyer's trust. However, in countries where credit card penetration is weak, this type of insurance policy may be particularly attractive to establish seller trust.

Collectively, we call these three forms of insurance the Internet Trust Insurance (ITI). We have given only basic descriptions of these products, as our intent here is to engage in conceptualizing a framework to enhance trust and promote e-commerce in the marketspace.

Back to Top

Interactive Surveys

We conducted surveys in Singapore and South Korea to measure feedback on respondents' experiences with the Internet, views on trust, views on the use of insurance to protect online transactions, and the appropriate pricing of the proposed ITI. We chose these two countries, both of which rely heavily on electronics and IT industries that exhibit great potential for e-commerce development, so that we could build a comparative study between them. We designed an extensive questionnaire (25 statements under four sections). Respondents were asked to indicate the extent to which they agreed or disagreed with each statement based on a 5-point Likert scale ranging from "Strongly Agree" to "Strongly Disagree."

As ITI is a new concept, we felt the survey would be better achieved if the respondents fully understood the objective of our study and how the proposed ITI works. Therefore, we conducted our surveys on an interactive basis, personally approaching and briefing each target respondent before asking the respondent to complete the questionnaire. This proactive approach gave respondents the opportunity to ask questions on the spot, and respondents often offered additional, valuable comments.

The surveys were conducted in June 2001 with CEOs, general managers, IT managers, sales, and purchasing departments. We surveyed 35 Singapore companies and 38 South Korea companies, broken out by industry as follows:

• Insurance companies: 3% of companies surveyed in Singapore, 34% in South Korea.
• Electronics and IT firms: 54% in Singapore, 47% in South Korea.
• Other industries: 43% in Singapore, 18% in South Korea.

To further assess the feasibility of the proposed ITI from the perspectives of potential industry players, we conducted more in-depth interviews with seven insurance and technology organizations (three in Singapore and four in South Korea, the leaders in their respective industries). Here, we first present empirical results of the interactive surveys, and then offer some insights from the in-depth interviews.

Back to Top

Singapore

Using the Kolmogorov-Smirnov one-sample test, a non-parametric test, we reject the null hypothesis that the distribution of responses for the majority statements (18 out of 25) is identical at the 5% significance level, indicating that the respondents did not behave randomly when expressing an opinion on most of the statements.

From the results, which are statistically different from a random distribution, we found that all (100%) of the Singapore companies surveyed were familiar with using the Internet but only 63% were confident of their abilities to handle online transactions. Security measures such as VeriSign and eTrust appear to enhance confidence for online transactions as respondents generally or strongly agreed (68%) that they would be more inclined to transact when these measures were in place. Eighty percent of the respondents were also more inclined to conduct e-commerce with sites that protect buyer interests.

A finding of particular importance is the high statistical significance of the responses in Section 3, that all statements pertaining to the use of insurance as a proxy for trust in the cyberspace garnered positive responses (at least 69% indicating "agreed" or "strongly agreed" for each statement). The idea of using insurance for transaction protection seems well received among the respondents, which may indicate a good potential market. However, there was strong feedback that the seller should be the one paying the premium to encourage buyers to transact online with them. When it comes to the pricing of ITI, around 70% of the respondents were uncertain of the appropriate premium, probably because ITI is new.

Back to Top

South Korea

The same test rejects the null hypothesis at 5% for all except five statements. The South Korean survey revealed that South Korean companies were generally familiar with the use of the Internet to procure or sell goods. Interestingly, despite this familiarity, only 40% of the respondents saw the Internet as a safe medium for transacting business. Most of the companies would safeguard their interests by conducting rigorous checks on the background of the site before transacting business (66%). They were also more inclined to transact with sites that protect purchasers (76%).

The respondents were in general agreeable to the use of insurance as a proxy for trust, and contrary to their Singapore counterparts, the South Korean respondents were more willing to pay a small premium to transact with sites that offer insurance protection (66% "agreed" or "strongly agreed"). Most (61%) also believed that insurance would potentially translate into cost savings from reduced search and appraisal efforts.

Back to Top

Comparison

The Kolmogorov-Smirnov two-sample test shows that the null hypothesis (no difference in the responses given by the two groups) cannot be rejected at 5% for 22 out of the 25 statements. Only the responses for Statements 3, 5, and 17 are statistically different. In respect of Statement 3, an absolute majority (84%) of the South Korea sample indicated that they had used the Internet to purchase or sell goods on the Internet more than once, whereas less than half (49%) of their Singapore counterparts had indicated the same. In fact, about one-third of the Singapore respondents had never bought or sold merchandise over the Internet. Such results are not unexpected, given that South Korea ranks among the most wired nations in the world. While South Koreans enjoy the highest broadband penetration rate worldwide, Singaporeans still rely mainly on modem access [11]. A good infrastructure, coupled with the intensive support of South Korean government towards IT developments, logically means that companies in South Korea are more ready to embrace e-commerce than the Singapore companies.

Following the higher level of e-commerce activities, for Statement 5, the South Korea sample also exhibits more familiarity with reputable sites that offer goods for sale (66%), compared to the Singapore sample (54%). This familiarity with reputable portals might be the exact reason why almost half (47%) of the South Korea respondents were indifferent to whether goods sold over the Internet are insured against fraudulent misrepresentation, whereas 77% of their Singapore counterparts felt such insurance would be important, for Statement 17.

Back to Top

In-Depth Interviews

The interviewees with whom we spoke at length included senior/executive vice presidents, general managers, venture capitalists, and senior researchers. These interviews offered valuable insights on e-commerce trust issues and the viability of the proposed insurance product. The salient points discussed are summarized as follows.

• Trust Issues on the Internet
  All interviewees agreed that trust is a crucial consideration for online transactions. In spite of the support for and confidence in e-commerce growth, some of them did not feel secure transacting business over the Internet. Interestingly, on a personal level, most preferred to buy offline and if online purchases were inevitable, they would take extra precautions such as setting their credit card limits. One interviewee noted that purchasers generally associate brand names with trustworthiness. Even in South Korea where people are more inclined to purchase from domestic firms, having a strong international brand name would certainly help in the Internet landscape.
• Critical Success Factors for ITI
  Most interviewees regarded the ITI idea an interesting concept that carried certain potential. In particular, the coverage on mismatched quality of goods and non-delivery was thought to be novel. In their view, its viability hinges on several critical success factors: the backing of a strong brand name to underwrite the project or the collaboration of big insurance companies with the requisite e-commerce experiences, accumulation of critical mass (that is, there must be a sufficiently high demand for the product to make it profitable), right timing for the launch, careful control of the variables in the product to be offered, and appropriate pricing of the product.
• Potential Implementation Issues
  At the infancy stage of e-commerce development, there could be difficulties in developing off-the-shelf kind of insurance products for online transactions. Risks involved in electronic transactions are complex and hence, hard to quantify. While insurance products that involve too many variables are unlikely to succeed on the Internet, an attempt to create comprehensive insurance schemes that encompass all possible risks would lead to exorbitant premiums. A related issue is that purchasers transacting on the Internet tend to be more price-sensitive. The pricing of the premium would thus be critical: insurers would only launch the product if profits could be achieved, but the premium has to be small compared to the transaction value. One interviewee opined that a good estimate of the premium for the proposed ITI would be in the range of 0.5% to 1.5% of the transaction value.

Back to Top

Implications

The central issue is trust. There was unanimous agreement that trust is very important in offline transactions and may be even more important in online transactions. In short, trust is the foundation upon which commerce is built, and in the virtual world it may be the fuel for locomotion.

Most of the people surveyed and interviewed were enthusiastic about using insurance as a proxy for effecting trust on the Internet. However, there were reservations about paying for such a product. Part of the appeal of doing business over the Internet was the idea of reducing intermediaries between the buyer and seller, thus reducing costs and prices. Introducing new intermediaries like insurance companies would actually increase the cost of doing business on the Internet, and these costs would likely be passed on to the price-conscious consumers. Because using insurance is a less conventional way of solving the lack of trust in the virtual world, its value proposition is not clearly understood and appreciated.

We imagine that if we were to conduct market research on the feasibility of bottled water—before its advent and wide acceptance today—we would be told that water is very important but most respondents would be unwilling to pay a premium for it to be bottled and conveniently sold. Consumers would tell us that they are comfortable with the way they receive it (by piping it to their homes). Yet today, bottled water is a billion dollar business, and some companies have even gone as far as branding it. Higher product price is not necessarily a barrier to marketing success [8].

Insurance serves a need in the old economy somewhat like the need for water at home does. The new economy is about greater mobility, as evidenced by the bottled water that allows people to carry this precious commodity with them wherever they go. In time, insurance will probably exert its influence in the new economy much as it has played a pivotal role in promoting commerce in the old economy.

Still of major concern is the basic principle of supply and demand. Insurance companies are concerned about the growth and capacity of e-commerce, the key demand driver on the demand side. Once the market is attractive enough and other obstacles are removed (for example, getting enough support from reinsurance companies), it is likely that insurance companies will supply some kinds of products to meet the market demand.

The importance of e-commerce cannot be ignored. Business-to-business (B2B) commerce is generally believed to be the dominant player in marketspace. The use of insurance will promote greater end-user confidence, which will in turn translate into a higher volume of transactions on the Internet, and foster the growth and development of both business-to-consumer and B2B e-commerce. Enhancing trust on the Internet and broadening the revenue model for the insurance market is a reciprocal process from which all involved parties will benefit.

Back to Top

Back to Top

Back to Top

Back to top

The Digital Library is published by the Association for Computing Machinery. Copyright © 2003 ACM, Inc.

---

No entries found