The Web is bad; really bad.
My estimate is that at least 90% of all commercial websites are overly difficult to use due to problems like:
Bloated page design that takes forever to download .
internally focused design that hypes products without giving any real info about them.
Obscure site structures that either have no logic or are based on the company's org.chart.
Lack of navigation support, making it very hard to find things when combined with an obscure structure.
Narrative writing style optimized for print and linear reading; not for the way users read online (they don't; they scan ).
As discussed in my previous column on user testing of websites, the average outcome of Web usability studies is that test users fail when they try to perform a test task on the Web. Thus, when you try something new on the Web, the expected outcome is failure .
Even when the site works, the total user experience often remains miserable . For example, I recently had to buy a new PC and tried to do so through Dell's website, following my own rule that you must live a "Web lifestyle" yourself if you want to be an Internet pundit. The Dell site had some weaknesses, but it was reasonably easy to use and allowed me to order the desired high-end machine. Three days later I received a confirmation email stating that the machine was expected to ship 6 weeks later. This was obviously not satisfactory: when you order on the Internet, Amazon.com has trained users to expect a confirmation email within a few minutes and the product within a few days, unless the website has warned them about shipping delays.
When I called up Dell, I was told that the late delivery was because my requested tape drive was out of stock. How about integrating your inventory system with your website, folks? Customers need to be told about delays and inventory problems while they are still researching their purchase online and can consider alternative options. Outcome: Dell lost a $3,035 order because their website delivered poor customer service.
Why Do They Like It?
There are several reasons why people keep using the Web despite its many problems:
Even though 90% of sites are bad, users don't spend 90% of their time at bad sites . People only visit a bad site once but become loyal users of the good sites. Thus, any individual user may spend 90% of his or her time at good sites and only 10% checking out bad sites. For example, Yahoo! is the most-visited site on the Web, partly because they have one of the fastest and most minimalist designs.
Most people don't know how much better the Web could be . Think back to the age of the Ford T: people bought this car in droves since it was better than riding a horse. But if you had to chose between a Model T and a Mercedes E430, you would pick the Mercedes. To me, using the Web is like having to drive a Model T every day when I used to drive a Mercedes: I know how much better hypertext systems can be, based on all the research we did back in the 1980s and early 90s.
Sometimes the Web does work and is better than reality . For example, it is much easier to bookmark a list of stock quotes at a website than it is to look up the stock price in the newspaper. It is even easier to search for company-related news online than to scan the business pages.
Because they do get good service on the Web every now and then, users behave somewhat like Skinner's rats who would keep pressing a lever in their cage as long as it gave them a food pellet at rare intervals. In fact, the rat would keep going longer if the food pellets came at random intervals: with randomness, there is always the hope that next time will bring the reward - exactly as when you click a link on the Web.
There is one happy conclusion from the fact that most websites are bad and that users spend most of their time and money at the 10% good ones: it is relatively easy to succeed on the Web because most of your competitors are clueless. In the physical world, it is difficult for large companies to give good customer service since most of their staff is unqualified. In contrast, on the Web, good service is a small matter of programming and can scale to ever more customers and product selection.
Luckily, usability has improved since I wrote this article in 1998: success is now more common than failure when we test users trying to perform a task on the Web. But most of my other conclusions from 1998 continue to hold.