For the last several years, CIOs have been deluged with analyst reports and research linking fast-loading Web pages with increased customer satisfaction, retention, and spending.

The mantra has been that if a customer has to wait more than eight seconds for a page to download, they will bolt from the site in favor of another vendor with faster pages (a.k.a. better customer service).

This thinking has led many enterprises, specifically e-commerce ventures, to pour dollars into performance-enhancing products and services like load-balancing switches, caching devices, and content-delivery networks. But is that all that can be done to guarantee happy customers and clients? And, when it comes to e-commerce, will those enhancements prompt visitors to spend more money on the site?

Not necessarily.

While good performance is essential, there’s a growing contention in the industry that other factors may also play significant roles when it comes to customer satisfaction, retention, and spending. Specifically, current research shows that users will often abandon sites if the site’s organization makes it difficult for users to find what they need.

The impact of site usability
User Interface Engineering (UIE), a research group in Bradford, MA, conducted what they term a “7-Eleven milk” experiment to determine the impact Web design has on user satisfaction.

“The idea is that if we gave you a person with a poured bowl of cereal and no milk and you went to that person’s house, picked them up, drove them to a 7-Eleven, gave them enough money to buy milk, and sent them into the 7-Eleven, 100 percent of the time, that person would successfully buy milk,” explained Jared M. Spool, UIE’s founding principal. “We tried something similar with Web sites.”

In its experiment, UIE recruited users who perform certain tasks, like buying books or electronics equipment online, gave them money, and instructed them to buy products.

The surprising result was that only 30 percent of the test volunteers successfully completed their task—obviously, a much lower percentage than the hypothetical 7-11 milk experiment result.

So, who or what was to blame? Site performance had very little to do with successful transactions, said Spool. UIE found that, many times, people simply could not find what they were looking for and that poor site organization hindered the buying process.

The UIE study, called “Getting Them to What They Want,” was published in October 2001. It reports the various problems users had visiting Web sites and offers insight on making Web sites easier to use so visitors can find what they what.

Perception is reality for satisfied users
The idea of improving site-visitor satisfaction is just beginning to take hold. In the past, many CIOs relied on time-to-download performance measurements to ensure that sites were user friendly. Now, IT and business-strategy management are seeking new metrics and analysis tools for evaluating performance to determine if their sites are meeting visitor needs.

One such tool is Webtrends Analysis Suite, Advanced Edition from San Jose, CA, vendor NetIQ Corp. This tool helps CIOs understand visitor activities and explains why a user left a site before completing a purchase, providing clues as to whether a user left because of navigational problems or slow page-download times. It also tracks user navigation within the site to see what content grabs a user’s attention.

Analysis Suite, Advanced Edition takes into account things like failed transactions, links that lead nowhere, and basically anything that causes a problem for a visitor. The UIE research indicates that remedying such problems can yield positive user perceptions of a site.

For example, the UIE experiment had users visit several sites they normally would patronize and perform a series of tasks, including searches, finding an item to buy, and making purchases.

Surprisingly, the study found no correlation between actual download speeds and perceived download speeds. While no two users performed exactly the same set of tasks, all rated the sites in a consistent manner. So there was no disagreement on which sites performed the best and which ones performed the worst.

But when UIE measured the actual page download times, they did not match up to the users’ performance ratings. Although some sites took three to four times longer to download, users rated these sites among the best performers.

There was, however, a strong correlation between perceived download time and whether users successfully completed their tasks on a site.

The point, according to UIE’s Spool, is that time is relative. “When people are not happy with a situation, times seems to pass slowly,” explained Spool. “It seems that when people accomplish what they set out to do on a site, they perceive that site to be fast,” he added.

So while companies have spent lots of money improving the download times of their pages, some simple organizational changes may very likely yield a more favorable impression in a visitor’s mind. Better organized content, more efficient search engines, and simple ways to find specific products (e.g., by letting the user enter a catalog number) could increase user satisfaction as much as any hardware improvement.

How fast is your site?

Have you used any performance tools or enhanced site logistics to improve the user experience? Write and tell us, and we’ll share your input in a future story.