Readers' Comments on Web Trust

by Jakob Nielsen on March 8, 1999

Sidebar to Jakob Nielsen's column on trust on the Web.

I have received some interesting comments on my March Alertbox with on trust and how to communicate trustworthiness in Web design.

Cheating Customers Eliminates Hope of Long-Term Relationships

David Jao from Harvard University writes:

By a freak of coincidence, I read your March 7, 1999 Alertbox right after I read a New York Times article on how more e-commerce sites are hoping to turn browsers into buyers. Judging from the Times article, it looks like real world e-commerce sites aren't listening to a word you say. Here are some choice quotes:
Ken Cassar, an e-commerce analyst for Jupiter Communications, a New York-based research firm, said a big problem was shipping and handling costs, which many sites do not disclose until late in the buying process. "It's sticker shock," Cassar said. "A lot of people just aren't willing to pay what is essentially a 30 to 40 percent markup."

The obvious solution is to disclose the shipping cost at the outset.

Perhaps the most aggressive approach to the problem is represented by Cybershop, an online department store that actually pursues shoppers who abandon large purchases.... "if it's a nice-size order, we'll call or e-mail them and try to close the sale,"
It is exactly these kinds of abuses that make customers wary of giving away personal information such as e-mail addresses. Ironically, in the very same article merchants complain that shoppers balk at filling out the registration forms which the merchants so love.
[Andy Halliday from Excite said that] the Internet has been a medium for impulse spending. "You have to make a consumer's purchase decision as slippery a slope as possible," referring to the speed with which sites should finalize sales.
I could not disagree more. Swindling customers into a sale does not make for loyal customers. Instead, e-commerce sites should try to improve the user experience for the 95% of users who visit the site without making a purchase. I know for a fact that whenever I buy from a web site, it is always from a site that has provided me with helpful information in the past.

Jakob's reply: I agree with your analysis. It is interesting to note how the received wisdom in almost all conference speeches these days is to aim for:

  • branding
  • 1-to-1 marketing
  • life-term value of a long-term customer relationship

And yet, when it comes down to it, many sites totally abandon this long-term approach and go for very short-sighted gains. True, you can trick people to give out their email address for mailings they don't want, and you can trick people to buy something they don't want. But long-term relationships and a trusted brand? Lost the second you trick the customer.

One of the worst examples of the short-term thinking is the "trick banners" that have become popular where the advertisement looks like a dialog box. These banners probably do get high click-through rates because some users think that they are actual dialog boxes that require a response. But the second the user discovers the trick he or she is out of the destination site. And it doesn't seem to be a beginning for a customer relationship to start out by tricking the prospect.

Reputation Primary Asset for Small Publisher

Adam C. Engst, publisher of TidBITS writes:

We pay a great deal of attention to our reputation with TidBITS, since as a small publisher, our reputation is perhaps our primary asset. We've been publishing every week on the Internet for nine years now, and I'd offer some additional suggestions for increasing trust when working online.
  • When possible, be up front and honest with details. Explain what you're doing and why you're doing it and readers will respond positively, especially if the activity might seem questionable. We interpret this as "self-documentation" as well, and regularly write articles about how and why we use certain technologies in certain ways.
  • Interact with users on an individual basis. People respond to other people, not to companies. To achieve this, we write from the first person singular in TidBITS, and we always identify individuals whose writing appears in TidBITS. On a more generic level, I'll bet most people far prefer receiving customer support responses when the response is signed by a real person. And a good experience with an individual can increase customer loyalty immensely. As an example, we shop with, a Seattle-based Internet grocery service, and the fact that they've gone to the trouble of hiring smart, communicative drivers (who are the people you meet) has significantly helped our opinion of them.
  • Write a clear and comprehensive privacy policy. People may not choose to believe it, but having one is better than not having one. Ours is here .
  • Set down delivery guidelines and stick to them. With TidBITS, that has meant publishing on exactly the same day every week for 9 years, and for the last 5 or 6 years, sticking to a self-imposed 30,000, character limit. People trust regularity.
  • Strive for professionalism and quality at all times. It's easy to be sloppy on the Internet, but if you take the time to ensure all public communications are spelled correctly, use proper grammar, and (if in email) have at least the rudiments of formatting, people will regard you more highly. This may be easier for us than for many people, since we're professional writers and editors, but it is worth the effort. We even go to the extent of spell-checking and doing basic editing on all posts sent to TidBITS Talk, the moderated mailing list we run alongside TidBITS. See the Talk archive for examples of this (plus a discussion interface we're quite proud of).
  • Respect your users. People aren't dumb, and if you respond quickly to comments and suggestions with explanations of why you might or might not implement them, that goes a long way toward helping your reputation.

I won't pretend that these suggestions are all that easy to implement, since for the most part, they require changes on the part of people, not pages. But if you take into account the economic advantages of increased trust and reputation, there's probably a business case that could be made. Plus, the more "reputation points" you've built up, the less damaging blunders will be.

Worst-Ever Low-Trust Design

Azeem Azhar writes:

Interesting piece on trust on Websites. One of the worst examples of this is This is a Macintosh store online. You are presented with a blank screen demanding (with no other identification, or reasoning provided) for an e-mail address... and no way of verifying you put in the correct one! You can't proceed to the site without answering that question.

Fredrik Warg follows up:

I just read the "Readers' Comments" section on your Alertbox "Web Trust". There was a link to cwonline (Worst-ever...) which I immediately followed, and sure enough, without further explanation they asked for my e-mail. I killed the window.

Then, in your reply, you wrote: "Apparently the site has realized the errors of its ways because it had changed when I revisited it"

Uh, did I miss something. I saw exactly the white screen described by Azeem Azhar. I went back to cwonline to see if I did miss something, and behold, the main page appeared.

As it turns out, not only do they have the "Worst-Ever Low-Trust Design", they also implemented it badly. The first time you visit they send you to the e-mail page and set a cookie saying you have been there. The second time around they detect the cookie and send you too the main page directly, whether or not you actually typed in anything on the e-mail page on your first visit.

Indeed, a strange design.

Jakob's reply: This is indeed the worst site I have ever seen with respect to "greeting" the user. At first, I thought that the site had realized the errors of its ways because it seemed to have changed when I revisited it recently. But as Fredrik Warg discovered, the changing interface is due to a cookie. Leaving aside the issue of the overly aggressive request for the user's email address, it is also bad for usability to change the interface this drastically based on a cookie. Cookies are fine for saving log-in information (for sites that don't require high security) and for recording user preferences and personalization settings. But the interface shouldn't change drastically simply because of the number of visits: doing so violates the basic design rule of "perceived stability." (It's perceived stability because individual pages obviously can change to reflect new stories, new products, and other updates.)

Trust and New Classes of Users

John Collins from the Ghana Media Gateway writes:

I agree completely with you about "trick banner ads." I am training journalists how to use the Internet in Ghana, West Africa. The Web can be difficult enough to explain to people who have little or no computer exposure, without having advertisers intentionally misleading them. I have made it a policy to tell my 'students' to just ignore banner ads, as they are generally just going to be led astray. (This doesn't bode well for e-commerce in W. Africa!)

In addition, with regards to operating system usability, I'm amazed at the badly thought-out metaphors that Microsoft has introduced into Windows. "Recycle Bin"? Who's ever heared of this outside the Western world? Many people have no idea what this is for. In addition, the web-based file management introduced in Win '98 has done nothing but confuse people who have no access to the Internet in the first place. Why would these people want to see their computer as a web-page? What's a web page?

Anyway, I know that Africans must only account for <0.01% of computer users worldwide, but there's my cry in the dark!

Jakob's reply: John Collins raises an interesting point in that it is even more necessary to be absolutely honest and straightforward with new classes of users who are coming online for the first time and who are not as hardened yet as some of us old-timers who have seen it all. If people's first impression of the Web is that they are being taken advantage of, then it will be harder to grow this new medium into its full potential.

Share this article: Twitter | LinkedIn | Google+ | Email