• by Gavin O'Malley  @mp_gavin, March 15, 2012

Despite modest advances, brands still have a lot to learn about Web site design and execution.

From 2010 to 2012, just 1% of brand sites were given a user experience score of “very good” -- while just 17% were deemed “good,” according to Forrester. 

Since 1999, Forrester has used its Website User Experience Review methodology to evaluate the user experience of 1,500 sites spanning business-to-consumer and business-to-business, and across disparate industries and geographies. To gauge site quality, Forrester employs a type of methodology known as a “heuristic” or “scenario” review across 25 distinct criteria. 

However, what’s good for one brand’s Web site isn’t necessarily good for another. “Web sites aren’t inherently good or bad,” according to Forrester analyst Adele Sage. “They can only be judged by how effectively they meet specific users’ needs. For example, “when buying a TV online, a recent college grad on a shoestring budget living in a tiny first apartment has different needs than a 45-year-old setting up a state-of-the-art in-home theater.”

In general terms, however, brand sites seem to be making the most progress in the key categories of value and consumer trust. These improvements have been achieved by a greater focus on privacy policies in the context of requests for personal information, along with clearer error messages, site reliability and simple location cues.

Where sites most often go wrong is with illegible text, inefficient task flows, poor use of space, and missing links to privacy and security policies in context. 

“Two of those -- illegible text and inefficient task flows -- have been the top two problems for almost five years,” according to Sage.

Luckily, many common design flaws are understood -- as are the ways to fix or even avoid them, according to Forrester. Text legibility problems boil down to type that’s too small, too low in contrast, or both -- and the definitions of “too small” and “too low in contrast” are backed by years of scientific studies.

That said, some design flaws can’t be fixed just by following universal guidelines. The right information architecture depends on what information is being architected and for whom.

But even though these specific problems are unique to each site, human-factor specialists have developed standard methods for resolving them with a high degree of certainty.

To find and fix the problems on their Web sites, Forrester suggests applying a four-step process beginning with the identification of target users and their goals. Once brands have goals in mind, they should test them, evaluate their site experience against Forrester’s 25 criteria, and finally, fix any problems.