Usability Methods in E-Commerce Solutions

 

Brian LaLonde

Final report for EECS 547, Fall 2001

Submitted 12/12/2001

 

Abstract

 

E-Commerce solutions must be usable, but developers have difficulty in spending resource on usability. Discount Usability Engineering is a good solution, but it is not sufficient because available resources are often hard to use themselves.  The argument for usability will be summarized and a useable summary of practical applications will be provided.  The summary will be published on the Internet, along with tools that have been implemented to aid in Discount Usability Engineering.

 

Introduction

 

E-Commerce is a term referring to the migration, or even total automation, of standard commerce practices to more efficient electronic means.  There are many aspects involved in this process, and when considering a typical new E-Commerce mechanism, some aspects are focused on more than others.  Aspects commonly in the limelight include the potential for success of the ideas themselves, the actual implementation of the idea through technology, and even the race to be the first to actually succeed with the idea, among others.

At some point in all E-Commerce endeavors it is realized that the user experience is a very important aspect to consider.  Often, this realization comes too late, or even never at all, due to fact that all work is being put into the aspects already mentioned.  Consequently, focus on the user experience, or the usability, of these mechanisms is all too often neglected and new E-Commerce mechanisms fail as a result.

            This is why the study of Human Computer Interaction (HCI) plays a key role in E-Commerce.  This is plain to see, and will be shown through specific examples, but the important realization is that, to succeed, all E-Commerce endeavors should pull from the field of HCI to practically improve the usability of their mechanisms.

            The problem with this realization, though, is that it adds another layer of complexity to the work of E-Commerce developers.  They are left with yet another aspect of their already large task of innovating an E-Commerce mechanism.  And not only is this another aspect for them to consider, but usability is an area that very few developers have resource in.

            Faced with this lack of knowledge, HCI specialists can be turned to, but this route is expensive in both funds and time.  So, another solution needs to be found.  This solution is commonly called discount or cheap or easy usability testing or engineering (for the remainder of the writing it will be referred to as Discount Usability Engineering).

            But, yet another problem arises.  These “cheap and fast techniques” are still too expensive for developers to utilize.  Resources in this area are easy to find, but are flawed in usability themselves, either by being too in-depth or too broad to ever practically make a difference.  This is the last reason developers end up spending little to no time considering usability factors in their work; intimidation.

The solution to these problems should be a usable summary by a developer for developers.  This summary will be presented in this paper.  And, since this phenomenon is especially applicable in the arena of Business to Consumer (B2C) ventures, the summary will be focused toward usability-deficient B2C developers.

 

Procedure

 

The approach taken for this study is a review of the current state of information available pertaining to E-Commerce usability.  In that sense, the study is a summary of that information.  The results of that summary will be put into web form.

The summary will include a usable heuristic for evaluation of B2C E-Commerce web interfaces.  This heuristic will pull strongly from existing models, but will be focused through the cutting out of unneeded information.  Tools to aid in the automation of this heuristic will be found, discussed, and even implemented and made available along with the summary.

 

Why is usability important in E-Commerce?

 

What’s the situation with the information society and E-Commerce?

 

It’s incredible to consider how fast our society has become an information-centric one.  Simply put, society has moved from being focused on the production of physical goods to one who’s emphasis is on the production and exchange of information.  The Internet, and specifically, the World Wide Web, is cited as being the force driving this information-centric society [4].  Internet statistics source Nua.com estimates the Internet user base to be 513.4 million people as of August 2001, and they expect it to reach 1 billion by 2005.  It is astounding that 513.4 million is only 8.5 percent of the world’s total population and that even three years ago the estimate of Internet users was only at 150 million people. [5] 


This information-centric society gives rise to new financial trends.  We now have E-Commerce, or, the buying and selling of goods on the Internet [8, definition from whatis.com].  Figures vary, but estimated E-commerce spending for 2001 is at $52 billion for the United States alone.  This amount is also expect to show rapid growth in the near future.

 


But, There is a Problem

 

            These statistics point toward a new E-Economy, part of which we already experience, but one that will surely see even more change each year. Along with the proliferation of this new change and growth, of course, come many problems.  The simplest way to see these problems is by looking to the users themselves.

            The Graphics, Visualization & Usability (GVU) Center at Georgia Tech has been conducting Internet user surveys since 1994.  The data they have collected is useful in viewing the voice of Internet users.  In particular, the survey respondents consistently report that E-Commerce Internet sites are deficient in three key areas; 1. disorganized or confusing,  2. too slow to download, and 3. lacking basic information about products [9].

            These areas are all problems due to the lack of focus on the user experience of a mechanism in the design of that mechanism.  This focus on the user experience of an Internet mechanism is focus on the usability of a mechanism.

 

What is usability?

 

Usability is a measure of how easily users can interact with a given mechanism, or to what extent users can efficiently and effectively achieve their goals.  Usability fundamentally has to do with the way a user interacts with a system [1][2].

The study of how to improve usability is within the realm of Human Computer Interaction (HCI).  This is an interdisciplinary field in which computer scientists, engineers, psychologists, social scientists, and design professionals all come together to making computer-based systems easier to use and fundamentally more useful for people and organizations.  They do this through solving real problems in the design and use of technologies. [3]

In general, it seems that usability on the Internet will be a problem to focus on for years to come and that these HCI specialists have a lot of work in front of them.  This is true, but some would limit the impact that work on usability could have by citing Metcalfe’s Law.

This phenomenon, named Metcalfe’s Law after Bob Metcalfe, the inventor of Ethernet, says that an electronic network’s value grows as the square of the number of its users [11].  This effect is often cited to explain why new efforts on the Internet do not become ubiquitous.  The theory is that because only a fraction of the world’s population is on the Internet, the value of it as a network is limited.  It can be argued that the word “usability” fits in this context of network value.  If the number of users on a network is low, then by Metcalfe’s Law, the usability of that network is going to be limited to being low as well [10, pg 349]. 

            This does not render focus on usability useless, though.  Metcalfe’s Law applies most specifically to physical machines that are connected on networks [12].  The usability factors mentioned here apply greatly to the already valuable growing world of the Internet.  Though the Internet may not yet be ubiquitous, the network already has very high value.  Usability as defined and addressed here still has much significance, even in the face of Metcalfe’s Law, and any attempts to increase usability are not in vain

 

Why does E-Commerce need to focus on usability?

 

            Simply put, bad usability equals no customers [10 pg. 14].  On the Internet, the user controls everything.  They get to choose what site to visit, and then when they get to the site, where to click on that site.  With millions of sites available and even within a single site, many things to click on, users can at any point click away from where an E-Commerce site would like them to be; and that’s buying.  Clicks away from your site are bad, but users have the freedom, and are quick, to click away to another buying option.

            Users experience E-Commerce websites in a similar way to actual stores that they would visit.   Jakob Neilsen gives a great analogy for this:

“having bad usability is like having a store that is on the 17th floor of a building (so nobody can find it), is only open Wednesdays between 3 and 4 o’clock (so nobody can get in), and has nothing but grumpy salespeople who won’t talk to the customers (so people don’t buy too much).”

The customer’s needs must be put first.  These needs include how well the users can navigate through a site, find information, make purchases, find help if they need it, and the list goes on.  This is the fundamental problem that usability for E-Commerce tries to solve.  The idea is to solve problems before users encounter them.

            Again, the users’ voice speaks clearly.  More data from the GVU Internet user surveys [9] shows that, though Internet shopping is on the rise, users are not satisfied with their E-Commerce experiences in general.  Only 20% of Internet shoppers find what they are looking for all of the time.  Only 16% of online buyers could say that they haven’t had a problem with a site causing them to click away to another site.  The top problems they cite are sites being disorganized or confusing,  too slow to download, and lacking basic information about products.  These are all problems with site usability.

Not only are these reasons easy to see, but there are more numbers to support a usability focus. Zona Research, Inc. conducted a study on consumer behaviors when faced with slow download times, a key issue in site usability.  From their research, they concluded that $362 million is lost each month due to connection speeds alone [6].  That totals to $4.4 billion a year.  If the industry is still loosing sales at the same rate, then that figure could be well above $7 billion.

The Zona research focused on download time problems, while there are other factors that make sites less usable as well.  Jakob Nielsen estimates that all the usability problems of intranet design will cost $50-100 billion per year in lost employee productivity in 2001 (he also notes “$50B is the conservative estimate; $100B is the median estimate; you don't want to hear the worst-case estimate!”) [7].

           

Why is usability typically under-par for E-Commerce?

 

There are three major reasons for the sub-par usability of the majority of websites:

·         Not knowing or taking into account history,

·         Treating the web as a direct extension of what you already have,

·         And, that most people aren’t HCI specialists.

With a new medium, mistakes are made.  Fortunately, the Internet isn’t entirely new, so most mistakes have already been made.  New E-Commerce sites classically repeat the same mistakes that others have already committed and already fixed.  If development teams took into account good examples that are already on the web, as well as the wealth of information available from the field of HCI, they could avoid the sorry mistake of being the ones who do not know history and, therefore, are doomed to repeat it. [10, pg. 15]

Companies are big ships to steer.  They see the possibilities that the Internet offers and jump right to it.  A major result of this Internet rush is that they end up turning what they already have directly into a website.  The problem here is that when existing commerce processes go electronic, they generally don’t directly translate well.  The Internet is a new medium to be considered.  It is different from traditional brochures, paper catalogs, and storerooms.  Usable websites require a shift in thinking.

These two reasons of repeating historical failures and not being able to shift thinking seem like they could be fixed by hiring on an HCI professional, and although Jodi Bollaert at Compuware [2] claims that hiring a usability expert has a cost-benefit ratio of $1:$10-$100, companies rarely do hire a specialist. Two reasons are pointed out for this.

If 200 million people design sites without considering usability, and the losses due to this add up to $100 billion, then each designer is only causing $500 worth of loss.  That is not enough to justify the costs of hiring professional designers or paying for advanced usability work [7].

Also, the world has about 20,000 user-interface professionals.  If all websites employed a single HCI professional then, at the current rate of Internet growth, every one of those professionals would need to design one Web site every working hour to meet demand [13].

So, unless it becomes feasible to hire HCI specialists, the responsibility of making a usable website is usually going to be put into the hands of non-specialists.  One solution to this lack of expertise would be for developers to become specialists themselves through reading, gaining experience, and keeping up with the current issues of HCI.  The major problem with this, though, is that developers have no time for it.  Pressures due to lack of time are very common in creating new E-Commerce sites.  One of the many reasons for this general lack of time is the race to be first on the market.  Even if the development team were large and the skills of the members varied, few team members would have time to do appropriate usability research and apply it, let alone a developer working alone as a one-man team.

 

So what can be done?

 

Despite these roadblocks to making a site more usable, focusing on a site’s usability is still very important.  Sites will continue to go live without help from usability experts.  They will also continue to be produced by developers knowing little about increasing usability.  So, another solution needs to be found because usability still needs to be considered.  This solution is found in what is commonly called Discount Usability Engineering (DUE).

The idea is that doing something about usability is better than doing nothing.  It is not about knowing the whole field of HCI and applying the endless resource stored in that field, but it is about starting with something and actually taking action on it.  DUE is on a level of simplicity so that anyone can do their own usability work.  Simplicity also brings speed, so DUE is meant to be fast so that the precious time of development doesn’t get bogged down.  Along with simple and fast, DUE is, naturally, also cheap.

            Often, though, these “cheap and fast techniques” are still too expensive for developers to utilize.  It is increasingly easy to find resources in the area of usability, but when found they seem to be flawed in usability themselves, either by being too in-depth or too broad to ever practically make a difference.

For example, on the popular usability website and originator of the term Discount Usability Engineering, Useit.com, one can easily find links to specific problems such as DVD menu designs, but it takes reading a number of articles and following many links before anything truly simple and useful is presented.  Also, in a typical session trying to use Useit.com one runs into multiple links (2% of links [14]) that seem very interesting and useful, but lead to pages that offer reports for sale at prices often over $100.

Despite this, Useit.com is still one of the best resources for useful information about how to make sites more usable.  There are also many other sources of information available, but they often share the same problem; they themselves are too complicated to use.  Therefore, it can be seen that a usable summary is needed.

 

A Usable Summary of Discount Usability Engineering

 

            What follows is a five point Discount Usability Engineering approach developed through study of many other DUE approaches.  It is meant to be simple and concise in order to maintain maximum ease of use.  This summary should include the reason that usability itself is needed.  This reason has already been covered in the section “Why is usability important in E-Commerce?”  (Sources listed at [15])

 

1. Know your users.

 

Discount Usability Engineering always starts with one principle.  This is to simply know your user.  Begin by asking questions like,

 

·         Who uses your Web site?

·         Who buys your products?

·         What would they want to see?

 

If a developer knows the answers to these questions, and keeps them in mind, he immediately begins to create sites that are more user-oriented, and therefore more useable.

 

2. Watch your users use your website.

           

            Creators of tools naturally use those tools with ease.  Unfortunately, the creator is not the user.  Watching only five people use your site can reveal many specific problems that you would have never come up with on your own:

 

1.       Ask a user to use your website and speak aloud about the choices he/she is making

2.       Watch the user and take notes

3.       Apply what you learn

 

3. Navigation

 

            The top two reasons users click away from sites are confusion and the fact that they can’t find what they’re looking for.  You might not have what they want in the first place, but if you do, then you want them to find it easily.  These are simple ways to make your site more navigable:

 

·         2-3 clicks to any solution

·         Effective search engines

·         2 things always in view:

·         useful functions (i.e. shopping cart)

·         system status (where am I? - simple urls can help this)

·         Be consistent across the whole site

·         Don’t use frames

 

3. Latency (download time)

 

Second only to navigability, slow download time is listed by users as the reason they click away from a site.  To ensure users can access your pages:

 

·         All pages should be less than 30 Kilobytes

·         This will ensure even modem users download times of less than 8-seconds

·         Server Reliability with the ability to handle peak load times

·         A user should never receive a denial of service

 

5. Simplicity

 

            Surveys of users consistently reveal that they would rather have sites that download fast and are easy to navigate than sites that are loaded with eye candy.

·         Practice minimalist design

·         Every extra unit of information competes with relevant units of information and diminishes their relative visibility.

·         Also, to maintain simplicity:

·         Don’t use animation

·         Don’t require extra software (plug-ins) to use your site

           

 

 

And, B2C specific Points:

 

            These points are formulated in response to the top problems listed by users regarding B2C E-Commerce sites. 

 

  1. Put extra effort into giving product information.
  2. Never force disclosure of any user information before it is necessary.
  3. Make shipping information clear even before a user makes a move to purchase.
  4. Security information should be clear and concise.
  5. 1-800 number and physical location should be available.

 

 

Conclusion

 

Expected results:

 

I expected that the literature review would reveal that an adequate summary of E-Commerce usability issues did not exist in practical form.  To some extent, this was true, but in the end I was amazed with the amount of information available in this field.  Before starting the project I was basically ignorant of the whole field of Human-Computer Interaction.  I feel that I have a much better working knowledge of the field.

I also intended to produce a summary of usability issues, including usable heuristics for evaluation of E-Commerce web mechanisms.  I feel that I did this to the best of my ability, but this was hindered by the amount of information available.  I became stuck at points in depths of information that I did not need to be digging into.

Finally, I planned to present the summary along with implemented usability tools on the Internet.  This has been done at http://www.umich.edu/~blalond/usable

 

References

 

  1. ISO 9241 part 11 http://www.lboro.ac.uk/research/husat/inuse/f_usability_standards.html
  2. Is Usability Useful in E-commerce? http://www.compuware.com/intelligence/articles/e-usability.htm Written by Jodi Bollaert at compuware
  3. Course description for Class at Carnegie Mellon: 20-790 Human-Computer Interaction in eCommerce http://euro.ecom.cmu.edu/program/courses/tcr790/
  4. Stephanidis, C. et al. (1998). Toward an Information Society for All: An International Research and Development Agenda. User interfaces for all : concepts, methods, and tools. pp. 635
  5. Nua.com http://www.nua.net/surveys/analysis/yearinreview/archives/2000_yearly_review.html
  6. The Economic Impacts of Unacceptable Web-Site Download Speeds  Zona Research, Inc.  April 1999 http://www.xde.net/products/kb/slowsites_white_paper.htm
  7. Useit.com http://www.useit.com/papers/web_discount_usability.html
  8. Whatis.com: http://searchebusiness.techtarget.com/sDefinition/0,,sid19_gci212029,00.html
  9. GVU WWW user surveys: http://www.gvu.gatech.edu/user_surveys/

http://www.gvu.gatech.edu/user_surveys/survey-1998-10/graphs/general/q68.htm

http://www.gvu.gatech.edu/user_surveys/survey-1998-10/graphs/shopping/personal/q208.htm

  1. Nielsen, Jakob (2000) Designing Web usability
  2. The Siren Song of Internet Micropayments  iMP Magzine, http://www.cisp.org/imp/april_99/04_99crocker.htm
  3. Less than Metcalfe’s Law http://webword.com/moving/metcalfe.html
  4. Web Test Dummy, Webtechniques July 1999 http://www.webtechniques.com/archives/1999/07/carter/
  5. Data based on 2212 links: 81% external, 17% useit.com, and 2% nngroup.com
  6. References used in formulating the heuristic:

-          Super Easy Usability Testing at WebWord.com http://webword.com/moving/easytesting.html

-          Stupid Ecommerce Tricks: Five Real Ways Top Web Sites Drive Customers Away http://www.zdnet.com/anchordesk/story/story_2359.html

-          Web Usability at Monash http://www.its.monash.edu.au/web/slideshows/usability/all.htm

-          8 Quick Tips for a More Usable E-commerce Web Site at WebWord http://webword.com/moving/8quick.html

-          Don't hide key e-commerce information at ZDNet http://techupdate.zdnet.com/techupdate/stories/main/0,14179,2830115,00.html