Placing Trust at the Center of Your Internet Strategy

When consumers visit a retail Web site, how do they know that the information describing the products or services they want to buy is accurate and unbiased? When they order and pay for a product online, how do they know that their financial records will be protected, that the product will be delivered on time, or that they can return something that is damaged or fails to meet their expectations? The answer is they often don’t know.

Consumers must make these and many other online research and purchasing decisions almost solely on the basis of trust. Yet most Web sites provide consumers with scanty information on which to base their trust. Some Web retailers are start-ups with little or no track record of fulfillment. Some may be on shaky financial footing and unable to meet their service and delivery guarantees. Some secretly collect data about each customer’s Web activity and then sell this information to third-party marketing firms. Even well-regarded companies like AOL have suffered embarrassing security breaches, while auction sites such as eBay have been scrutinized for their failure to effectively police self-serving “customer reviews” posted by the sellers and their friends. It’s no exaggeration to say that as consumers become more sophisticated about the Internet, Web-site trust is going to become a key differentiator that will determine the success or failure of many retail Web companies.

Trust has always been a key element in successful marketing.1 In industrial marketing, the 20% of the sales force that sells 80% of the volume owes much of its success to building trust-based relationships with clients. To preserve trust and confidence in the relationship, a smart salesperson will even recommend a competitor’s product if it better serves the customer’s needs. In consumer marketing, brands such as Coca-Cola, Tide and Disney act as trust marks, signaling consumers that they will get the quality they expect. For financial and insurance services, successful selling is heavily based on trust perceptions. Consider the recent Saint Paul Companies advertisements that show a girl on the Serengeti touching the horn of a rhinoceros. The ad copy reads: “Trust is not being afraid even if you are vulnerable.”

In recent years, the Internet has established itself as a new medium for marketing consumer and industrial goods and services. However, most sites on the Internet today do not focus on building trust as part of an ongoing relationship with their customers. Many Web sites act merely as self-service catalogs: If you know what you’re looking for, you can find and order the product or service. Such sites are commonly characterized by their crowded format, flashing banner ads and off-price promotions. Pursuing the hard sell, these sites do not give customers much information or help in making buying decisions. Not surprisingly, they convert few of their visitors into purchasers, suffer low customer retention and generate meager profits. Many companies have failed with such an approach to marketing on the Internet, primarily because they have failed to build trust.

For the Internet, trust-based marketing is the key to success. Companies can use the Internet to provide customers with a secure, private and calming experience during which they converse with an on-site, trusted personal shopping advisor who is dedicated to helping them make the best decision. Trust-based Web sites provide customers with accurate, up-to-date, complete and unbiased information, not only on their own products, but on all the competitive products available in the market. Their smooth, easy-to-use navigation makes searching, shopping and comparing a pleasure. Moreover, they preserve and build trust through faultless fulfillment and satisfaction guarantees. It is not surprising that trust-based Web sites can enjoy higher rates of customer conversion and retention than sites that do not engender loyalty. Trusted Web sites actively promote deep customer loyalty, thus greatly enhancing the lifetime value of their customers. By mastering trust-based strategies, companies can build a positive relationship with their customers while increasing their market share and profits.

Leading-edge sites are just beginning to demonstrate some trust-building components. In this article, we will review the emerging trust cues that can be found on some Web sites and propose virtual advisors as a new element in building trust into the customer’s shopping experience. Virtual-advisor software that mimics the behavior of a personal shopping assistant can become a powerful and cost-effective part of almost any company’s Internet strategy. We have tested consumer reaction to such software at Truck Town, a Web site we designed to advise consumers on truck buying options. As you will see, most consumers who used the site trusted Truck Town’s virtual advisors more than the dealer from whom they last bought a vehicle.

Building Trust at Each Phase of the Acceptance Process

Before we consider the trust-building potential of virtual advisors, let’s review some of the more established methods by which trust can be generated on the Web. (See “The Keys To Building Web-Site Trust.”) Trust is built in a three-stage, cumulative process that establishes (1) trust in the Internet and the specific Web site, (2) trust in the information displayed and (3) trust in delivery fulfillment and service. Trust in the information cannot be established until the Web site itself is trusted, and trust in fulfillment requires prior trust in the Web site and in the information it provides. Web trust cannot be established unless all three elements are well executed.

Web-site trust can be enhanced by ensuring consumer privacy.2 Customers increasingly demand that their personal data be kept private. For this reason, Web sites should not employ cookies unless their use is specifically allowed by the individual customer. Internet businesses often use cookies to record a customer’s activity on Web pages that display banner ads. In this way, the companies build customer profiles that can be sold to other marketing organizations. Some sites secretly collect even more granular information about their customers by using Web “bugs” to collect click-stream data on every step of their Web-site activity, even on pages that have no banner ads. Many privacy advocates and customers consider these powerful marketing tools to be an invasion of privacy and worry that they are being used to build personal dossiers on individual users. Companies that are serious about building trust do not employ such methods unless the customer explicitly approves their use.

Third-party seals of approval can provide an important cue to consumers that they can trust a particular site. For example, TRUSTe grants its seal to sites that adopt its standards for privacy and comply with its audits. Similarly, sites can build trust by assuring customers that their online payments are secure and can be executed only with proper authorization. VeriSign grants its seal of approval to sites that use its encryption and authentication services. An authoritative security policy with no embarrassing failures is a great trust builder. In the event security is breached or hackers disable the Web site, communicating with customers via toll-free telephone service can do much to preserve their confidence. Above all, Web sites that want to build trust must live by the privacy and security policies they endorse. The highly publicized attempt by bankrupt e-tailer ToySmart.com to sell its customer database to a subsidiary of the Walt Disney Co., it’s largest investor — or to auction if off to the highest bidder — is an example of how a company can be tempted to violate its own privacy policy. Failure to uphold privacy policies can affect not only the Web company itself, but its investors and business partners as well.

Building or transferring brand equity can also enhance Web-site trust. A brand name can provide an important trust cue connoting a Web site’s credibility. Companies may be able to transfer brand equity from their existing brick-and-mortar business to their Web site. For example, Barnes & Noble has attempted to capitalize on its reputation as a quality book seller by transferring its established brand and its attributes (selection, convenience, service attractive prices) to the company’s Web site. Web sites can also build new brands that generate confidence. Amazon.com has built a Web-trust brand by satisfying customers with the widest selection, thorough information (reviews, ratings), low prices, rapid ordering and delivery, easy exchange and quick credits.

Another way to establish Web-site trust is by creating customer communities that present user feedback to reduce the customer’s perception of risk.3 For example, eBay posts the number of positive and negative customer evaluations of each person or store that offers an item at auction and also allows bidders to contact past customers by e-mail. Although customer feedback is a potential trust builder, there are real limitations. Abuse by supposedly impartial reviewers can bias the input. An author’s friends may write rave reviews on Amazon.com. Or, more insidiously, a manufacturer may hire a firm to create favorable comments about its products and unfavorable comments about competitors’ products. Consumer feedback is useful when the sites supplying such information couple effective policing policies4 with a warning to their customers that such anonymous reviews may be unreliable. Other cues, such as best-site awards and celebrity endorsements, can be useful. Explicit pricing, return guarantees, telephone and e-mail support, and publication of the customer’s legal rights help build confidence. Intuitive navigation that allows customers to control their Web experiences also builds trust.

After establishing trust in the site, the next task is to engender belief in the information. Information on the Web site must be complete and accurate. Sites that ask customers to make a purchase should provide all the information needed to make an informed decision: product specifications, prices, in-stock availability, delivery time and reliability (see BizRate.com for customer ratings of sites based on timely delivery), magazine reviews, customer recommendations (see Epinions.com for customer evaluations by self-designated experts), and return guarantees. All third-party information (e.g., content from magazine articles and consumer feedback) should be free of bias. At minimum, the Web site should make its bias and information shortcomings explicit. For example, companies should warn customers if only their retail partners are providing the delivery and service ratings or if prices listed after completing a Web search are not arranged from lowest price to highest. Whenever possible, companies should allow customers to view information in the order they choose. Other powerful trust builders include posting reviews by credible sources, such as consumer-advocacy organization Consumer Reports, or audits of site information by reputable independent parties; frequently updating changes in products, prices and availability; and allowing customers to personalize and specify the information displayed.

It is also important to provide customers with unbiased information about competing products, including those made by other manufacturers, thus enabling consumers to make fair comparisons.5 Many Web intermediaries provide impartial information in virtually every product and service category. They pose a key challenge to incumbent companies unwilling to do the same. For example, CarPoint.com and CarsDirect.com supply information on all makes and models of autos. While it may seem radical for branded manufacturers like General Motors or Ford to supply fair comparisons, the Internet will inevitably force such companies to match the breadth of information offered by the new intermediaries — or “infomediaries.” Companies resisting this trend will sacrifice their customers’ trust.

The most important element of trust is fulfillment. Quite simply, trust is earned by meeting expectations. As small commitments are met, customer confidence grows in the belief that companies will also fulfill larger expectations. Critical functions include shipping the right product at the right time, effective installation, service, support, error-free billing and credits on returned items. Automated tracking services and telephone-based customer fulfillment hotlines can maintain confidence when inevitable delivery problems occur. Trust is difficult to earn and easy to lose. Failing to meet customer expectations is the quickest way to destroy trust.

Virtual Advisors As Effective Trust Builders

Web-site advisors are a powerful new way to build Web trust. Although a real-time human advisor might be cost-prohibitive, a virtual advisor enabled by software is able to mimic the behavior of a personal shopping consultant. A software program not a live person, the virtual advisor is programmed to behave like an experienced human advisor. It asks questions, records responses and proposes recommendations on the basis of the customer’s responses.

In 1999, pickup trucks accounted for more than 15% of all consumer vehicles purchased in the United States. Each of the more than 90 product alternatives sold for more than $10,000. Clearly, consumers could benefit when making such complex and risky decisions. Imagine having the world’s best expert on pickup trucks help you select the best truck. Such an expert would ask questions about your needs and preferences, provide you with full information on the variety of trucks available and recommend vehicles that suit your needs. With this in mind, we designed a site called “Truck Town” to demonstrate how virtual advisors can create trust on the Internet. We used the analytical tools of utility theory to estimate customer preferences for each product alternative and Bayesian decision theory to revise these preferences as customers told the advisor more about their needs.6

Truck Town visitors can choose to be guided completely by the virtual advisor or may navigate the site independently. Truck Town’s architecture allows the user to exercise considerable control over information acquisition, which is an important trust cue. Truck Town’s welcome screen displays a map to help customers locate dealers, the bank, a newsstand, coffee shop, city hall and customer-advice offices. An avatar in the form of a friendly, intelligent owl guides the customer to site features and answers questions about information sources. The owl explains that Truck Town helps customers make informed purchasing decisions and then refers customers to the town hall and the personal advisors for specific product information and recommendations. The second screen takes the customer to the town hall where Truck Town’s mayor explains that all the information is accurate and up to date. He also answers queries about the satisfaction guarantee, return policy and system audits.

In a national survey of consumers, respondents were asked whom they trusted most when buying automobiles: dealers, salesmen, mechanics, contractors, bankers, neighbors or magazine editors.7 Respondents said they would most trust an auto mechanic, a retired editor of an auto magazine or a contractor who has purchased many trucks. On the basis of this ranking by consumers, Truck Town presents these three advisors, who are not modeled after specific individuals but are composites of the best practices of people serving in those roles.

The virtual auto mechanic in our example is named Craig, a middle-aged man dressed in a mechanic’s uniform. (See “Meet the Auto Mechanic.”) After introducing himself, he answers queries about how he is paid, his strengths and weaknesses, and past customer comments. He then starts the conversation by asking the customer the size of the truck he or she prefers. (See “The Dialogue Begins.”) After conversation on the truck’s intended use — for example, if it will be used to haul a trailer — Craig probes for other preferences, such as relative preferences for low price, performance, fuel efficiency, power and style. Craig then creates a personalized showroom with four trucks that he thinks best fit the customer’s needs. He bases his recommendations on how well the trucks match the customer’s stated preferences and what similar customers have purchased.

The customer can then obtain complete information on the trucks’ specifications (horsepower, towing capacity, load-carrying volume, fuel efficiency, bed length), configuration (selection of options), competitive comparisons with other truck specifications in a matrix table, and evaluations by other users. Customers can also obtain magazine articles and advertisements about all the trucks presented. Additional options are to “meet other people like me” or to be transported to the coffee shop in Truck Town where they can engage in a live chat session with other customers who have similar needs. The dialogue ends by scheduling a test drive or requesting a price and delivery quotation from an actual dealer.

Truck Town’s Trusted Advisors

In June 1997, we conducted an empirical concept-and-usage study of the Truck Town prototype to assess the viability of a personal advisor. A sample of 280 Boston-area respondents who had bought a truck in the previous 18 months evaluated Truck Town in terms of trust, quality of recommendations and their willingness to use and pay for the service. Respondents spent an average of 30 minutes on the Truck Town site prior to answering the survey. (See “What Customers See on the Truck Town Web Site.”)

The results indicate that the virtual advisor developed during this research is able to generate trust in consumers. In answer to the question “Did you trust the advisor?” 82% of the respondents answered “yes,” 76% agreed that the information provided was trustworthy, and 88% agreed that the advisor recommended trucks that fit their needs. More importantly, 60% agreed that the advisor suggested alternatives they would not have considered otherwise. In terms of purchasing, 88% of respondents would consider buying a vehicle through Truck Town. On average, they would be willing to pay an additional $40 for this service.

In fact, 82% of respondents considered the Internet experience more trustworthy than an in-person dealer experience. Most respondents said Truck Town’s information quantity (87%) and quality (83%) was better than that available from a dealer. Whereas approximately 80% would recommend Internet shopping on the basis of testing Truck Town, only 20% would recommend the dealer from whom they last purchased a vehicle.

These are encouraging results and suggest that the advisor-based Truck Town site successfully engendered acceptance of the advice and information provided. This is a key element in trust building. Assuming that trust correlates with sales, the Truck Town site demonstrates how establishing trust can enhance buying through this type of advisor-based system.

The Buyer’s Need for Information or Advice

An advisor is not necessary for all products. Trusted advisors provide great value if the purchase decision involves certain attributes such as high price (e.g., trucks), complexity (e.g., financial planning), learning (e.g., how to use digital cameras), rapid change (e.g., adapting to a new PC) or risk (e.g., health care). An advisor can help build trust and thus sales on the Internet for products having one or more of these attributes, though not necessarily all of them. For example, customers regularly pay more for branded pain relievers than for equivalent generic tablets, because they do not want to risk their health on unfamiliar products. A virtual pharmacist advisor for cough-and-cold problems could help customers make better decisions by supplying relevant information on medicines and their ingredients.

Not all people prefer to use an advisor. In our extended concept test, we compared the Truck Town advisor to an information-intensive site, Microsoft’s Carpoint.com, which provides detailed information on all types of trucks. Approximately half of those who preferred the Internet to a dealer prefer the advisor-based site; the other half preferred the information-intensive site. A regression analysis indicated a significantly higher strength of preference for Truck Town among consumers who rated their knowledge of trucks lower, visited more dealers, were younger and were more frequent users of the Internet.

These results suggest that advice is more valuable to those who are less knowledgeable and confident. Open-ended responses indicated that people who liked Truck Town valued its advisory capability, the personal quality of the advisor and the easy-to-use and fun nature of the site. Those who disliked the site felt the town analogy was like a game, childlike and unprofessional. Those who preferred CarPoint. com liked the depth of information, professionalism, direct access and control. Criticisms of CarPoint.com focused on the selling intensity of the site and its sponsorship. Both sites gained trust, but one did it through an advisor and the other providing information.

This research suggests that there are two different segments of buyers. One segment comprises those who have confidence and knowledge about what they want and therefore need an easy, direct information-search capability. The other segment consist of those who have less knowledge and want help or advice in making their choices. If one wanted to design a site to appeal to both segments, a direct-search format could be supplemented by a trusted-advisor button that provides a personal-shopping advisor. Alternatively, an advisor format could be adopted that becomes a direct-search site if the customer wants to access specific product data. It is important to match site navigation to the customer’s cognitive decision process. It should be possible to fill the needs of both decision styles on the same site.

If you propose to use an advisor of some type, there are several technical approaches to choose from. They range from using intensive market-research questionnaires that obtain customer preferences to accessing toll-free telephone and e-mail connections with real people who answer questions. Of course, there are trade-offs in cost, analytic content and trust. (See “Which Advisor Strategy Is Right for You?”) The importance you place on each of these factors will determine the best approach for you, but a virtual advisor may be an attractive alternative if the cost of human advisors is prohibitive.

Selecting High Pressure or Complete Trust

Manufacturers and retailers must decide how much trust they should design into their Web sites. Strategies range from pressure selling to developing a full-trust relationship. (See “Choosing Between High Pressure and Complete Trust.”) Many sites on the Internet today are designed as high-pressure sales environments. A pressure-selling strategy tries to build business by promoting only the seller’s products. Information is slanted to win sales; advertising employs flashing banners; messages are pushed along the digital network on the basis of the user’s cookie and personal characteristics. Heavy promotion is used to move inventory or stimulate sales of high-margin items. Service is minimal. Results are measured in the short term.

In contrast, a full-trust relationship strategy advocates for the customer across all product alternatives, including competitors’ offerings. Full and accurate information is presented; advertising is displayed only if the customer requests it; a calm, consultative atmosphere is maintained. Premium prices are justified on the basis of value added, “high touch” service guarantees user satisfaction over the product’s life cycle and benefits are measured by the customer’s long-term loyalty to the franchise.

The third alternative, an intermediate trust-building strategy, can be implemented by offering honest information on some, but not all, products in the market. Manufacturers pursuing intermediate trust building would present an unbiased view of their own products, but would not provide information about competing products. However, their site would provide links to competitors’ Web sites. Customers would be presented with advertisements, but they would be given an explicit option to remove them from their computer screens. The Web site would also offer value-based pricing and guaranteed service to ensure customer satisfaction. Intermediate trust building is a good step on the way to a complete trust relationship and may well be the most politically viable, near-term solution for executives who see great risk at either end of the trust-strategy spectrum.

Forces that move our strategic attention to the full-trust end of the strategy spectrum include competition from new infomediaries who give complete information on all brands, the desire to preempt other branded competitors, and the desire to better serve customers through innovation. However, conflict with existing non-Internet channels, traditional reliance on high-pressure sales and marketing, and the difficulty of changing old views in the organization all tend to direct our attention to the high-pressure end of the strategy spectrum.

Consider the dilemma of today’s automakers as they try to choose between high-pressure selling and building customer trust. They want to respond to competition from new firms like CarsDirect.com, Autobytel.com and more than 20 other Internet auto aggregators, but much of the existing dealer structure is resistant. Why be innovative with distribution channels if the existing forces can gain protection through franchise laws? Consumers can find competitive brand information on the Web, but many advertising and marketing executives cannot imagine providing fair and unbiased information about competitive brands on their own sites. What if their brand does not win in the comparison? Why should they undermine their own products? It would take an innovative and courageous executive to say: “If our brand is not the best, let’s redesign it and make superior products.” Indeed, a trust-based strategy creates powerful incentives to deliver the best products and services to earn customer loyalty. Of course, this is easier said than done, because the stakes are enormous. It can take several years and cost billions of dollars to design and build a new or improved automobile or truck. Nevertheless, if the auto companies do not meet their customers’ desire for a trustworthy environment, they will delegate the marketing function to these new Internet aggregators and become suppliers to the new channel. Clearly, the automakers are impaled on the horns of a dilemma. Automakers — and other companies in a similar position — must move now to adopt an intermediate trust strategy. They could even explore the benefits and risks of embracing Web trust by launching a pilot program around a particular product line or division — one they are confident would do well in competitive comparisons. The biggest risk is to do nothing and stick with pressure selling while the rest of the world evolves toward trust.

Resisting, Following or Leading the Trust-Building Trend

The trade-offs that automakers face today will soon confront most established firms in other industries. Trust-based marketing presents enormous opportunities for gain, but the threat it poses to existing interests — and to all inferior product lines — will be a major deterrent to innovation. It is tempting for an executive to avoid arousing conflict by waiting until it is obvious to everyone that the only choice is to use trust as a strategy. Unfortunately, by that time it may be too late to gain the momentum necessary to join the leaders in the world of trust.

Enduring customer relationships can generate enormous gains in both sales and profits. Loyal customers are less price-sensitive, and their loyalty reduces the threat of commoditization that is faced by many firms selling on the Internet.8 Moreover, strong customer relationships open the door to cross-selling of services, an important avenue for business growth. For example, a trusted automaker could provide finance and insurance services for cars and perhaps extend into peripheral arenas such as property and casualty insurance. Trust on the Internet can leverage a firm’s existing consultative selling skills. A bank that provides personal-financial advice to wealthy clients with more than $1 million dollars in investment assets could provide a virtual advisor to customers with $250,000 in assets. The result: large gains in profit at little increase in cost. Similarly, an advisor-based Internet presence could allow a firm to reach global segments that would otherwise be prohibitively expensive or impossible to reach through personal services.

Managers who consider moving toward an advisor-based system will undoubtedly worry that their decision will arouse negative reactions from existing channels and marketing power sources in their organizations.9 They can overcome such resistance by converting channel members into allies in the trust-building process rather than adversaries that oppose change. Sharing the rewards of increased trust with the channel may help. For example, if trust widens the customer relationship, channel members may sell additional products in volumes that more than compensate for sales lost by competitive comparisons.

Companies that enjoy strong brand equity and high levels of customer loyalty may be able to follow a “wait and see” strategy. In this case, trust in the brand can substitute for site trust earned by providing competitive comparisons. However, if customers have less than perfect loyalty and need data on other products before making a purchase, brand equity may not be enough to compete successfully with intermediary sites that provide full information. For this reason, even companies with strong brands should begin exploring intermediate trust building.

Other factors conducive to a “wait and see” approach are the costs involved in implementing a full-trust information site and the initial desire to deliver basic functionality to an Internet site. Investing in trust building and competitive comparisons may also cause revenues to slump and short-term profits to plummet if the company’s products are inferior to those of competitors. This prospect is not trivial given the stock market’s propensity to examine quarterly results carefully. Profits would also be lost if the company were forced to invest heavily in redesigning its inferior products. Companies may also have to forgo revenue from the sale to other firms of customer demographic and preference information.

Ultimately, each executive must balance the potential risks and rewards to determine if a trust-based strategy is worth the considerable investment. Clearly, there is a big opportunity now. The pioneers in trust-based marketing on the Internet stand to gain significant long-term competitive advantage. But let there be no mistake: building a full- trust Web site entails big risks — non-competitive products will not survive the process. Unfortunately, doing nothing has an even bigger downside. Those who wait too long to adopt trust building will be marginalized by existing firms that have learned how to earn consumers’ trust as well as by entirely new competitors such as the infomediaries if they attack the auto industry with unbiased, honest advice.

Trust as the Currency of the Web

A few years ago, a wise observer said: “The Internet changes everything.” The same could now be said about Web trust. Fundamentally, we are experiencing an unparalleled growth in customer power. Today, customers can buy the best products with full information by using price-bots, such as DealTime.com, which displays all product prices available on the Internet. Consumers can form communities to exchange recommendations by means of sites like Deja.com, which provides consumer ratings on products. Industrial buyers can collaborate using Web sites such as VerticalNet.com. Consumers can also get peer input from vertical sites, such as iVillage.com, which provides specialized information for women. There are even signs of customers grouping together to get lower prices on Web sites like Mercata.com, which lowers prices as more customers agree to buy something. Similar exchanges are being formed in the business-to-business space to organize industrial buyers. Although to date these group-buying sites have failed to capitalize on Web trust in general, all the pieces exist to combine community, trust and group-buying power to give customers dominance in the producer/customer relationship.

Indeed, the Internet puts such power in the hands of consumers that a new term is needed to describe the paradigm shift: we call it “consumer to business” (C2B) marketing. In C2B marketing, customers will demand the best products at the lowest prices. They will demand and receive trust-based relationships with preferred vendors. In the future, we fully expect that consumers will band together to design the products they want and then put out bids to trusted manufacturers that will be clamoring to supply them. Trust will be the key to survival in this C2B marketing future.

In our opinion, trust will soon become the currency of the Internet. One need only read the daily business press — replete with stories of security breakdowns, fulfillment debacles and the misuse of customer information — to realize that trust, or the lack of it, has already become an issue requiring the attention of senior managers. Companies that want to do business in the Web world must learn to communicate and sell their products and services by providing trusted information, advice and service. Consumers will enter into a trusting and enduring relationship with suppliers on the basis of the exchange of this information. They will give their loyalty if their expectations are fulfilled. Moreover, they will pay a premium price to the companies they trust. The companies that earn real profits in the rough-and-tumble world of Internet marketing will be trust generators selling products that deliver the best value in a complete, unbiased, competitive comparison.