Trust & Trustworthiness

Essays in this ecommerce series:

The Meanings of Trust

Trust is a concept everybody understands at some personal level, but most people will have trouble enunciating a specific definition of the concept. Some people will have strict measures they use to evaluate their level of trust in a person or company, while others rely on a more subjective “feel” for determining whether to trust somebody.

Just as every person has his or her own way of deciding issues of trust, every role a person plays will also have its own understanding of trust based on the specific goals and priorities of that role. Empathy is great for times that call for compassion and care, but in matters of trust it can be hard to see things from another’s point of view. When you walk into a store that you have patronized for years, you have a level or trust about handing over your money in exchange for goods; the merchant may approach this transaction with an entirely different level of anxiety about your ability to pay — maybe a new employee is ringing you up — or your ownership of the credit card you proffer. Trust means different things to different people, to different roles, and in different scenarios.

Trust can mean such things as the following:

  • Do I believe that what someone says is true and factual?
    If you tell me that you will pay me back next week for what you borrow today, do I believe you?
  • Do I agree with a person or company’s goal or what they stand for?
    Is Microsoft’s dominance of the OS market really the best thing for me as a computer and software consumer?
  • Do I believe that a person or company’s goals and/or priorities match mine?
    If I registered at a commerce site, and the subsequently sent me newsletters and notices, is that really a service to me?
  • Do I believe that a site’s presentation of its informational content is objective?
    Is a site mixing advertising into its supposed editorial content?

The path people take to a level of trust can vary greatly, because some people work from the premise that trust must be earned, and some from the premise that trust is assumed but can be lost.

Earning and Losing Trust

Trust plays an important part in any transaction, but it may not be a conscious part. On a basic level, when we enter into a transaction, we make decisions about our participation: do we complete the transaction, or do we cancel the transaction?

Trust is not a characteristic that inheres in an ecommerce site; trust is a judgment made by the user, based on general experience learned from being a consumer and from the perception of the particular merchant.

In every transaction, both roles of buyer and seller must make a decision about trusting the other participant. Sometimes this decision is conscious, sometimes it’s unconscious and relies on a person’s default beliefs about trustworthiness in general. For example, some people I have talked with for this essay approach commerce with a general level of trust; these people by default trust merchants, but question a merchant’s trustworthiness if they receive worrying feedback or negative cues. In contrast, many people don’t especially trust any merchant, and must assemble much positive feedback and encounter positive cues before deciding to trust a merchant.

Those users that look for signs that they should trust a commerce web site are searching for the presence of “good” signs and the absence of “bad” signs. Good signs include professionalism of design, longevity, service, selection, positive anecdotal comments from other users, and even more points derived from personal experience. Bad signs include anything that violates expectations, as well as the obverse of the “good” points: amateurish design, newness, difficult interface, difficulty finding product, bad anecdotal comments, etc.


There is no oversight group that can objectively anoint a web site as being completely trustworthy. Trust is a subjective judgment that must be made by every user for any site, because individual goals vary and definitions of trust are unlikely to be consistent.

The merchants face the difficult task of selling themselves to the user based on the web site interface and on the advertising of the brand. Commerce sites must convince potential customers to trust the site; commerce sites have two critical messages for users, “buy from us” and “trust us”. Without the user’s trust, the merchant can’t sell.

Users look for the presence of positive cues about a site’s general trustworthiness, as well as for the absence of negative cues. Commerce sites can influence their impression of trustworthiness by carefully designing their site to avoid amateurishness, set and meet reasonable user expectations, and address common user concerns about privacy and security.

Some of the cues a user may look for while evaluating trustworthiness in a commerce site:

Is the site professional?
Have the designers taken care to create a coherent interface, or is the site the apparent product of somebody learning how to code HTML? Does the site have typos, grammatical errors, useless animation, or any of a slew of other problems that indicate that the designers or implementers don’t know what they are doing? If they can’t design a good site, why expect them to be able to handle commerce transactions and fulfillment? Personally, I won’t deal with any site that uses the Comic Sans font face because it shouts “built by an amateur” to me.
Does the information architecture make sense?
If the organization of information doesn’t appear to have been carefully designed, why expect the site designers to have thought much about fulfillment and payment systems?
Is the site easy to navigate?
Can I find the product I want to buy? Is it easy for me to browser the site looking for interesting things I might want to buy? If the designers can’t make the site easy enough to navigate around, why would I trust them to have made a backend that works?
Is the site easy to use?
Can I access and use the site and its offered functionality? Can I search, browse, and purchase from the site? If I can’t, because of my browser and preferred browser configuration, then the site designers are telling me that they don’t want my business.
Are my questions answered?
If I’m concerned about security or privacy, can I find information that addresses these concerns without having to email a faceless answer-bot? If the designers didn’t think about my possible needs before I made my purchase, why expect them to handle my needs after they have my money?
Do other people trust this site?
What’s the word about this site? Have customers received their orders correctly?
Am I familiar with this company?
Have I had good experiences with this company before, on or off the web?
Are the prices reasonable?
Is the price realistic or just too cheap? Will they hit me with an outrageous shipping charge to balance the low price?

Some users will react to a positive answer to these questions, incrementally working towards a personal measurement of a passing score for trustworthiness. Others will react to negative answers, presuming a site is trustworthy but experiencing doubt when one of these issues suddenly appears to be less than satisfactory. There are no uniform answers to these questions, and no uniform attitudes of users towards these cues. The lesson for commerce sites is to understand their targeted audiences, and address their likely concerns, and always pay attention to the implications of the site’s design.

The eCommerce Trust Study

The [broken link eCommerce Trust Study] from Studio Archetype/Sapient and Cheskin Research presents some interesting analysis on “the nature of those elements that communicate ‘trust’ in e-commerce sites, be they transactional or graphical.” According to this study:

The factors that produce a sense of trustworthiness need to be identified, in their entirety. Their interactions need to be understood, and their relative importance determined. Understanding the roles of these different factors would allow online retailers to ease consumers' concerns, and could hasten the maturation of Web retailing.

The report describes 6 main components (along with a total of 28 sub-components which I won’t mention here) of a commerce site that suggest trustworthiness:

  1. Seals of Approval -- Symbols, like VeriSign and Visa, designed to re-assure the visitor that security has been established. The companies that provide these seals of approval are referred to in this report as "security brands.
  2. Brand -- The corporation's promise to deliver specific attributes and its credibility based on reputation and the visitor's possible previous experience.
  3. Navigation -- The ease of finding what the visitor seeks.
  4. Fulfillment -- Clearly indicates how orders will be processed, and provides information on how to seek recourse if there are problems.
  5. Presentation -- Design attributes that connote quality and professionalism.
  6. Technology -- State of the art connotes professionalism, even if it's difficult to use.

This study has some valuable insights into user attitudes towards trust and ecommerce sites, and is clearly required reading for anyone interested in understanding online commerce. I do, however, have some issues with the report.

First, this analysis seems to consider online commerce in isolation. Ecommerce is a new kind of shopping experience in a new medium, but this experience is still new enough that the vast majority of people who buy online learned how to buy offline. Addressing ecommerce in isolation presumes that it can be studied as its own phenomenon with no experiential antecedents, which is unrealistic and will likely produce misunderstandings of ecommerce and user behavior.

If this study relies on an assumption that ecommerce is simply a modal variation of traditional commerce, then any analysis will be flawed. The fact is that the customer’s experience of the characteristics of traditional commerce does not map to online commerce, forcing the customer to seek imperfect analogs. The eCommerce Trust Study would be more accurate if it had addressed the chasm between what customers understand from traditional commerce, and what online commerce can possibly address.

Second, I think this report is a bit free in its lack of clear differentiation between what the merchant can do — suggest that it is trustworthy — and what the user must do — decide whether or not to trust the merchant. Granted, there is no checklist for commerce design that specifies everything that must be done a certain way to be correct, and no checklist that has a test for everything that might be done incorrectly. Commerce site designers, however, must understand that they cannot code trust into a site, they can only suggest trust. Trust is a property controlled by the customer.

A reader correctly pointed out that the second to last sentence in this essay was missing a negative; I added the word cannot on 15 February, 2000 to make the sentence retain the meaning I had intended.