Q: What is the single most important factor in building a successful Web site
or Web based business?
A: Your users!
For years every successful business has built its success by providing users
with the products and information that are important to their customers and users.
The successful business teams have figured out what is important to their customers
by continually asking them for inputs and feedback. Most importantly, the successful
businesses listen to the feedback and act on it. GE does it; Microsoft, Sun, and
AOL do it. And so do many of the other successful companies across the world.
When it comes to Web sites, the team first needs to have a common focus on
who are the target users for the Web site and why they would visit. Once this
is written out and shared across the team, improvements and testing can be gauged
against the team's vision. Next there should be a regular program to gather and
track the users' experience while working with the site. Web sites evolve at a
rapid rate and the team needs to have a long-term memory to track what they did
in the past and how well it worked versus the improvements that have been made.
Also, whenever possible you should do competitive studies to ask users what they
think of your competitors' Web sites versus your Web site. The team can then pick
out the best ideas from the answers.
Users are one click away from leaving your Web site
There are a number of methods and techniques to improve your knowledge of what
users think about your site. Some are easy and inexpensive and some require a
moderate budget and some can take a large investment in time and money to accomplish.
Unsolicited User feedback - First, if you have a site, make sure you
invite users to send you feedback. While few users take advantage of this, people
do use the feedback mechanism to tell you if things are broken on your site. When
a link or feature is broken, most users will move on to something else on the
Web, but some will click on an easy link to send you feedback. Make sure your
team captures these comments and also sends back a thank you note. The cost is
just your time to monitor and respond to the feedback.
Surveys - These take a bit of work to ask helpful questions. At the
low end of the cost range there are Web-based applications that will allow you
to implement and run a Web based survey for several hundred dollars. More typically,
a well-run survey conducted by an independent firm will cost in the $10,000 +
range. With a survey, the users are taken out of the mode of using your Web site
and are asked their reactions and opinions based on previous experiences with
your site. Typically these surveys would involve interviewing a large number of
users (50 to several hundred) to document profile questions and usage areas. Then
you would receive a set of graphs and charts showing the users' responses to the
questions you created. The results greatly depend on the questions you ask, how
you ask them and how you track the results. For example there is a normal bias
on how users will respond to questions and you also have to understand how the
responses your survey received compare with a normal baseline. An example is if
65% respond that they like your site and the normal bias is 75% of the surveyed
people like a Web site, then you have a problem. The timeframe to build a good
survey project is several weeks and the budget typically can require $15K to $30K.
Online usage profiling - There are now new software tools that allow
users to provide detailed feedback to Web site teams as they use a Web site to
do typical tasks. These services works by having a panel of users doing real tasks
on your Web site, or your competitors' Web site. As the users work with the site,
the software tracks the pages they view, how long they are on the page, and their
written feedback and comments they provide. With this resource, the team now has
detailed information on what pages are actually used by users to accomplish tasks.
They can see if users scroll to look over the page or not, and what the user feedback
and reactions are as they use the pages across the site. This information is available
in online interactive reports available to the Web site team. The big advantages
here is the Web site team is collecting detailed usage information as people really
use the Web site from their home or office, where they are normally using the
Internet. As you get 8 to 15 users doing the tasks you will quickly see what areas
are frustrating for them and what areas are working well. This can quickly help
a team zoom in on the areas that are most important. What is of particular value
is that this approach allows the Web site team to gain insights on what users
like or do not like, as well as why they will or will not use the site again.
Also, this process has been developed specifically for the fast response and turnaround
needed for today's Web sites. Web sites can now include direct user feedback in
a continual process of improving the Web site effectiveness and ease of use. The
time frame for doing this type of usability testing is typically 2 to 3 days and
costs start at about $3,500. As the Web site is improved, tests can easily be
re-run to track improvements. Two companies that provide this new testing service
are SiteUsers.com and Vivadence.
Focus Group Sessions - This approach brings a small, selected people
into a site where you watch them interact with your Web site. Critical factor
here is having a skilled and knowledgeable interviewer to work with the users.
Typically the interview process will run for 1 to 1 ½ hours. Users will
need an incentive ($100 is typical) to come into your interview site. Typical
usability lab projects will require a budget starting at $5K for a straightforward
project interviewing 10 people. A more complex project will require a budget of
$30K to $50K and will include a larger number of interviews as well as more complex
questioning of the participants. These surveys will be done in a controlled environment
where everything is set up correctly. The sessions are video taped and transcripts
of the sessions are typically made available to the client. The timeframe for
this type of testing is usually several weeks to complete a project and generate
the reports for the team.
Observations on the art of making a site usable
Clutter - What causes people the most problem with a Web site? Clutter!
As we have done Web site testing for clients, we see time and again, that Web
sites that appear confusing and cluttered to the users, have problems holding
the user's interest and attention. When the Web site has a clear interface, users
can easily and quickly find the resource or information they came to your site
for. When the Web page is cluttered or confusing the users will try and find the
information they are looking for and if it is not obvious then they move on to
another page or another Web site. My personal rule of thumb is that key features
have to be obvious to a casual observer.
Organization - Your Web pages and Web site have to be clear and easy
to understand, even when you are not around to explain things. You cannot assume
users will read your information. This means that the pages have to be clear and
easy to use without spending more then 10 seconds to figure out what you have
gotten. From watching the time users spend on a page, the Web site will be losing
folks after 10 to 15 seconds. That may not be fair to the creative work, but it
is the reality of how users are behaving when we observe the real behavior with
Web sites from the home or office.
Web site colors - Now here is an area for great debate that has little
impact on the Web site usability for the most part. The color palette of the Web
site has to be one you are comfortable with. But, do not spend endless hours debating
the color pallet. First, users care about having information easy to find, and
as long as the colors do not confuse them then the colors are OK. Next, you really
do not know what your colors are going to look like on anyone else's system. Nearly
every monitor or laptop make "adjustments" to the perceived colors that
you will not be able to predict. There is one major area of color choice that
should be treated with care. Colorblind folks do not see the shading differences
between reds and greens. If your site depends on users seeing a color difference
to find information or buttons, make sure they are clear to everyone.
Forms - Time and time again users are confused on what is required and
what is optional. Web sites that make their forms clear and easy to work with
have positive user experiences. On the other hand, when users hit an error after
filling in a form you need to tell them everything they need to do to correct
the errors. It is very clear that if a user gets an error page two or three times
they will give up and go off to something more worthwhile. Many Web sites today
still process the form one field at a time. These sites reject the form and throw
up the error page on the first problem. If the user is going to have multiple
errors, this method will lose them! While some users may stick it out, most will
leave the site. We had one user going through a page and then hit an error page
some 20 times before giving up. That is one rare determined user. You cannot count
on all your users being that persistent.
An Ongoing Process
Whatever method you choose to use for managing the development of your site,
you need to have a regular program of gathering user inputs into your development
and release cycle. The recommended cycle is to identify the critical areas from
the user feedback, fix and improve them, release the improved site to your users
and then retest. What is great about Web sites today is this whole cycle can now
be implemented in a matter of weeks. Users can benefit from the results of your
work very quickly. Now your team has the potential for delivering a continuously
improving Web site focused on making it as effective as possible for your users.