Why Web Sites Fail: Part III - Usability

Information Architecture

To understand what navigation techniques are effective today, it's helpful to view the larger issue from an historical perspective.

The graphic designer Richard Saul Wurman coined the term “information architecture” back in 1976 to describe the process of disseminating and displaying information in an easy to understand manner. This term was not widely used until the Internet spawned Web sites that required navigation systems to access content. And while many like to think that the principles of Web site navigation were self created or evolved from the software industry, there is a more extensive history that includes numerous graphic systems that are often taken for granted. Books, libraries, airports and highways all employ navigation systems that have been refined over the centuries. All are based on establishing conventions, utilizing familiarity and memorized mental maps as the basis for optimizing a search method.

For instance, books employ a mostly linear navigation system. It's easy to figure out how to navigate a book and where you are in it. But Web sites are not linear and are often accessed randomly, creating a great need for designing a consistent navigation system. When Web sites employ unique navigation schemas that require users to learn a new system, they are ignoring conventional practices and typically serve only to obfuscate the user's goal.

A recent study conducted by the Software Usability Research Laboratory at Wichita State University (http://psychology.wichita.edu/surl/usabilitynews/41/web_object-ecom.htm) quantifies commonly accepted navigation schemas, illustrating where on the page users expect certain navigational elements to be located. Understanding these conventions and effectively implementing them forms the basis of an easy to use site.

However, information architecture refers to more than just navigation. The foundation for effective information architecture is the ability to clearly communicate the purpose and contents of a particular page and how it relates to the larger Web site as a whole. If a site is serving too many interests then a navigation scheme, however well thought out, will never solve this fundamental problem. One way that this “multi-objective” quagmire can be solved, is to create multiple micro-sites, segmented by the different audiences to which the site appeals. Breaking down the site into smaller sections and providing simple hub-gateways (ie, home pages) to the micro-sites reduces the pressure from which home pages often suffer.

User Experience

Assuming that people will change their behavior because doing it online is faster/better/cheaper was one of the biggest failed assumptions consultants made during the Internet boom. There were numerous examples from Furniture.com to Pets.com that failed to understand the limitations of the online user experience. Even online banking took considerably longer to achieve modest penetration rates when initial research indicated consumers would flock to such a new service. Even the term “online user experience” can be misleading and its overuse often smacks of high priced snake oil propagated by over paid consultants left over from the Internet boom. The reality for most people is that the “real world” is more compelling than a computer screen (at least as it exists presently) and the Web is at best a very intricate and compelling information utility, not a replacement for many types of human interaction. Keeping in mind the limitations of the online user experience can help in developing a Web site that does not try to overreach its abilities. For instance, whenever migrating business practices online or developing a Web application meant to replace traditional offline behavior, make sure it is thoroughly tested first. Changing behavior, as the not uncommon failure of customer relationship management software has shown, is a slow, arduous and expensive endeavor.

Usability Testing

Whether developing a 10-page brochure Web site or a 1,000-page corporate portal, every Web site will benefit from usability testing. At a minimum, this testing can be as simple as asking a user to find something on your site and listening as they think out loud in pursuit of the information. It's amazing the weaknesses testing can uncover and often how easy it is to correct these navigation problems. Usability testing should be a part of every phase of a Web initiative – the earlier the better. It doesn't have to be done by professionals, though that will help and it doesn't have to cost very much. Allotting a modest percentage of the overall budget, in the 1%-3% range can increase the effectiveness of a site exponentially.

A recent survey by the consulting company GuideStar, asked Web site users “What makes a good Web site?” Numbers one and two was “content” related and number three was “easy to use.” Understanding what users want and delivering it in an easy to use manner will always be the foundation of a successful Web site.

A Note to Our Readers:

I would like to qualify slightly my discussion of Judge Campbell's decision in the Ashley Albright decision, which I addressed in my June article entitled “The Ashley Albright Case: A Mixed Message About LLC Business Asset Protection.” As you'll recall, that decision essentially held that single-member LLCs lack the critical LLC statutory business organization law advantage of business asset protection.

In the course of my article, I made the statement that although the holding of Judge Campbell's opinion in Albright was unquestionably correct, his apparent rationale for this holding – namely, that the transfer of an LLC member's economic rights carries with it a transfer of the member's management rights – was, as set forth early on in his opinion, technically incorrect, since under the Colorado LLC Act (the governing act in Albright), no transfer of management rights is valid without the approval of the members. Obviously, Ms. Albright would never have approved a transfer of her management rights to the bankruptcy trustee in that case.

I want to qualify the above statement by pointing out that later in the same opinion, Judge Campbell presented a rather sophisticated reading of the Colorado Act in which he noted that in the case of transfers of economic interests in single-member LLCs, no statutorily required approval of such transfers by other members of the LLC is even possible.

The point I've made above is admittedly a narrow one, but it may be important to the extent that in other courts, Judge Campbell's decision becomes an issue in bankruptcy and creditors' rights litigation involving single-member LLCs. Maybe, if you read his opinion as a whole, you can legitimately conclude that even his seemingly overbroad statement that LLC management rights automatically follow economic rights can be viewed as correct for single-member LLCs (though clearly not for multi-member LLCs).

- John Cunningham, Ransmeier & Spellman, P.C., Concord, NH.

While there are as many reasons for failure as there are different kinds of Web sites, this three-part series is based on experiences I have had over the past 5 years as a Web site builder and refers to Web sites that are used for client acquisition, customer relationship management and e-commerce. This third part addresses how people use Web sites and why they often leave before finding what they want.

Site Flight

According to Steve Krug in his usability book “Don't Make Me Think”, 55% of users abandon Web sites before they find what they are looking for. Frequently, Web pages, particularly home pages, are often under intense pressure to serve too many interests, leading to the adage that more is less. When confronted with unexpected complexity (ie, too many choices) users typically abandon the intended path for an easier route. Since the Web is a seemingly unending source of information, it's not surprising that users abandon sites so frequently.

Information Architecture

To understand what navigation techniques are effective today, it's helpful to view the larger issue from an historical perspective.

The graphic designer Richard Saul Wurman coined the term “information architecture” back in 1976 to describe the process of disseminating and displaying information in an easy to understand manner. This term was not widely used until the Internet spawned Web sites that required navigation systems to access content. And while many like to think that the principles of Web site navigation were self created or evolved from the software industry, there is a more extensive history that includes numerous graphic systems that are often taken for granted. Books, libraries, airports and highways all employ navigation systems that have been refined over the centuries. All are based on establishing conventions, utilizing familiarity and memorized mental maps as the basis for optimizing a search method.

For instance, books employ a mostly linear navigation system. It's easy to figure out how to navigate a book and where you are in it. But Web sites are not linear and are often accessed randomly, creating a great need for designing a consistent navigation system. When Web sites employ unique navigation schemas that require users to learn a new system, they are ignoring conventional practices and typically serve only to obfuscate the user's goal.

A recent study conducted by the Software Usability Research Laboratory at Wichita State University (http://psychology.wichita.edu/surl/usabilitynews/41/web_object-ecom.htm) quantifies commonly accepted navigation schemas, illustrating where on the page users expect certain navigational elements to be located. Understanding these conventions and effectively implementing them forms the basis of an easy to use site.

However, information architecture refers to more than just navigation. The foundation for effective information architecture is the ability to clearly communicate the purpose and contents of a particular page and how it relates to the larger Web site as a whole. If a site is serving too many interests then a navigation scheme, however well thought out, will never solve this fundamental problem. One way that this “multi-objective” quagmire can be solved, is to create multiple micro-sites, segmented by the different audiences to which the site appeals. Breaking down the site into smaller sections and providing simple hub-gateways (ie, home pages) to the micro-sites reduces the pressure from which home pages often suffer.

User Experience

Assuming that people will change their behavior because doing it online is faster/better/cheaper was one of the biggest failed assumptions consultants made during the Internet boom. There were numerous examples from Furniture.com to Pets.com that failed to understand the limitations of the online user experience. Even online banking took considerably longer to achieve modest penetration rates when initial research indicated consumers would flock to such a new service. Even the term “online user experience” can be misleading and its overuse often smacks of high priced snake oil propagated by over paid consultants left over from the Internet boom. The reality for most people is that the “real world” is more compelling than a computer screen (at least as it exists presently) and the Web is at best a very intricate and compelling information utility, not a replacement for many types of human interaction. Keeping in mind the limitations of the online user experience can help in developing a Web site that does not try to overreach its abilities. For instance, whenever migrating business practices online or developing a Web application meant to replace traditional offline behavior, make sure it is thoroughly tested first. Changing behavior, as the not uncommon failure of customer relationship management software has shown, is a slow, arduous and expensive endeavor.

Usability Testing

Whether developing a 10-page brochure Web site or a 1,000-page corporate portal, every Web site will benefit from usability testing. At a minimum, this testing can be as simple as asking a user to find something on your site and listening as they think out loud in pursuit of the information. It's amazing the weaknesses testing can uncover and often how easy it is to correct these navigation problems. Usability testing should be a part of every phase of a Web initiative – the earlier the better. It doesn't have to be done by professionals, though that will help and it doesn't have to cost very much. Allotting a modest percentage of the overall budget, in the 1%-3% range can increase the effectiveness of a site exponentially.

A recent survey by the consulting company GuideStar, asked Web site users “What makes a good Web site?” Numbers one and two was “content” related and number three was “easy to use.” Understanding what users want and delivering it in an easy to use manner will always be the foundation of a successful Web site.

A Note to Our Readers:

I would like to qualify slightly my discussion of Judge Campbell's decision in the Ashley Albright decision, which I addressed in my June article entitled “The Ashley Albright Case: A Mixed Message About LLC Business Asset Protection.” As you'll recall, that decision essentially held that single-member LLCs lack the critical LLC statutory business organization law advantage of business asset protection.

In the course of my article, I made the statement that although the holding of Judge Campbell's opinion in Albright was unquestionably correct, his apparent rationale for this holding – namely, that the transfer of an LLC member's economic rights carries with it a transfer of the member's management rights – was, as set forth early on in his opinion, technically incorrect, since under the Colorado LLC Act (the governing act in Albright), no transfer of management rights is valid without the approval of the members. Obviously, Ms. Albright would never have approved a transfer of her management rights to the bankruptcy trustee in that case.

I want to qualify the above statement by pointing out that later in the same opinion, Judge Campbell presented a rather sophisticated reading of the Colorado Act in which he noted that in the case of transfers of economic interests in single-member LLCs, no statutorily required approval of such transfers by other members of the LLC is even possible.

The point I've made above is admittedly a narrow one, but it may be important to the extent that in other courts, Judge Campbell's decision becomes an issue in bankruptcy and creditors' rights litigation involving single-member LLCs. Maybe, if you read his opinion as a whole, you can legitimately conclude that even his seemingly overbroad statement that LLC management rights automatically follow economic rights can be viewed as correct for single-member LLCs (though clearly not for multi-member LLCs).

- John Cunningham, Ransmeier & Spellman, P.C., Concord, NH.