By Paul Boag

A good starting point for planning the future of your website is to
analyze what you already have. To some extent we are doing this all the
time. That is how new projects happen. However, a more formal approach
helps to better inform your decision-making throughout the web project.

There are two ways to better understand your current website: qualitative and quantitative.

Qualitative feedback is received by requesting comment from stakeholders and users.

This is traditionally gathered using the following techniques:

Techniques Description
Stakeholder interviews The role of stakeholder interviews is to provide those being
interviewed the opportunity to comment on the existing website. These
comments should be collated for later analysis.
Feedback mechanisms Allow visitors the opportunity to comment on your site using contact us forms, polls, and surveys.
User testing Watching visitors use your existing site can be very enlightening.
It is a powerful way to identify problems in the usability of your site.

Quantitative analysis on the other hand draws upon various automated
analytical tools that provide information on different aspects of your
site’s performance. These include:

Tools Description
Web logs analyzers Every time a user interacts with your website information about
that interaction is stored. Analysis of these logs can help identify
areas of improvement.
Automated performance checkers Automated checkers assess things like accessibility, download times, and browser support. These help maximize your audience.
Online visibility trackers Having a great site is important, but if nobody knows it exists
then it has failed. There are a number of ways to gain an understanding
of how visible your site is online.

Let’s look in more detail at these two approaches and better understand the role of each.

Qualitative feedback

We’ll now take a moment to focus on other feedback mechanisms.

Most websites provide some method by which users can submit
feedback. This is normally a contact page. However, a contact form is a
passive approach and something more proactive is needed if you want
user feedback. The majority of users will not think to send in comments
unless they are frustrated with your site. The problem is that in such
situations they tend to simply leave rather than complain.

If you want feedback on your site then specifically ask for it. This
can be done with a simple feedback form or a more comprehensive survey.
However, a word of warning if you are considering a full-blown survey.
Few users take the time to complete a long survey, so keep your
questions to a minimum. Also avoid making your requests for feedback
too intrusive. They should not hinder a user from completing his or her
goals.

If you are considering making changes to an
existing site, it is well worth implementing a basic feedback mechanism
to canvas opinion before you begin.
Whether you are getting
feedback from your site through user testing or via stakeholder
interviews it is necessary to assess the value of the comments made.
When analyzing negative comments about your site, use these four
criteria to judge how seriously those comments need to be
taken:

  • How often the comment is being made?
  • Who makes the comment?
  • What effect the problem has on the user and your objectives for the site?
  • How easy the problem is to fix?

Adobe support provides a simple and yet unobtrusive feedback
mechanism. The side column asks users if they found the support
document useful.

The more often you are hearing the same negative comment, the more
likely it is that the comment is justified and needs addressing.
However, you cannot rely on numbers alone. If your biggest customer has
a problem then you had better address that concern fast!

There is also a need to ascertain the seriousness of a problem. Does
it stop the user from completing a task or is it simply a mild
inconvenience? Does it in some way hamper a business objective? If it
does then it will need addressing.

Finally, establish how difficult the problem is to fix. Even a minor
problem is worth fixing if it is easy to do. Conversely, fixing a major
problem might be unjustifiable if the expense is prohibitive. In such
situations, look for a workaround that lessens the seriousness of the
issue.

Ultimately these decisions are about return on investment. Does the seriousness of the problem justify the cost of fixing it?

Although nothing is better than feedback from your users, it can be
a battle. Stakeholder interviews and user testing are time consuming,
while site feedback mechanisms are often ignored. Fortunately,
quantitative analysis is much less work. However, it should only be
used to support qualitative feedback—not as a replacement.

Free survey service

Providing a method that allows user feedback does not have to be
expensive or complicated. There are a number of services, such as questionform.com that allow you to create free surveys in minutes.

Quantitative Analysis

I am bad at assembling flat-pack furniture. Part of my problem is
that I never have the right tools. Fortunately when it comes to
analyzing the strengths and weaknesses of your site, there is no
shortage of tools.

Let’s look at the three types of analytical tools I mentioned earlier, starting with web logs.

Web Logs Analyzers

The most well known form of analysis is carried out on a
site’s log files. Log files track where a user has come from,
what pages they have visited, how long they have spent on each page,
and other data on users’ interaction with your site.

The problem is that log files are hard to understand. There are many
tools available to help with this, from free open-source software to
expensive enterprise-level products.

It is probably best to start with something cheap and cheerful. In
my experience the majority of website owners won’t use the
advanced features offered by high-end tools. You can always upgrade
later.

Another option is to use a statistics collector that doesn’t rely on log files. One such tool is Google Analytics.
This collects considerably more data than web logs, and has an easy to
use interface for analyzing the results. It is free of charge and only
requires a small piece of code on each page to work.

So, how do you judge if an existing site is performing? There are three basic things you can look at:

  • The number of unique users and where they are coming from.
    If many sites are linking to you, it is a good indication that you are
    doing something right. Traffic levels also indicate the performance of
    existing marketing campaigns.
  • The percentage of repeat visitors compared to first time users. If users are returning to your site regularly it is normally a sign of satisfaction.
  • How users are moving around your site. How long are they spending on individual pages? How many pages are they viewing? Which page are they leaving the site from?

The final test is trickier to interpret. For example a user might
visit many pages, which could appear to be a sign of interest in the
site. Alternatively, it could mean they cannot find the information
they require. Instead, compare the time on site to the number of pages
viewed. If they are looking at a good number of pages for a reasonable
time then you know things are going well.

By looking at where a user leaves, you can sometimes get an
indication of potential problems. Are users just looking at your
homepage and then leaving? If they are leaving without viewing other
pages then you have a problem with your homepage. Are users getting to
checkout on your ecommerce website and then giving up? Perhaps
it’s time to user-test your checkout process.

There is a lot more you can do with web stats, but that should be enough for you to analyze you existing site.

Let’s now turn our attention to automated checkers.

Automated Performance Checkers

Sometimes when analyzing your web stats you will notice a
significant number of users who leave your site after only viewing a
single page. This can happen for a variety of reasons. For example,
they may have simply come to the wrong site. However, it could also be
because they have met technical difficulties accessing your site.

There are three tests you can easily perform to identify any potential problems.

  • Check your site on as many different browsers as possible.
    I recommend you look at your site in at least the last two versions of
    Internet Explorer, Firefox, Opera, and Safari. If you do not have
    access to all of these browsers then consider using an online service
    such as browsershots.org.
  • Check your site’s accessibility using an online accessibility checker.
    These tools provide a report outlining the various accessibility
    problems with your site. However, a word of warning; these automated
    accessibility reports can be both misleading and confusing.
    Nevertheless, they can help you identify possible accessibility problem
    with your site.
  • Carry out a speed test on your site. You should be
    looking for download times of less than 10 seconds on a 56k modem.
    However, up to 20 seconds is acceptable. Automated checkers have a
    broader role than monitoring site performance. They can also be used to
    track the online visibility of your site.

Online Visibility Trackers

Web stats and performance checkers provide information on site
usability and accessibility, but they don’t tell you how easy
your site is to find. Fortunately there are tools that do exactly that.
Start with a site like popuri.us.

Popuri.us is a free web application that allows you to check your
site’s ranking on search engines, blog listings, and even social
networking applications. The site checks various sources to ascertain
your online visibility.

If you want information about your site’s ranking for specific search terms, then a tool like GoogleRankings.com
will help. Despite the name, this free application checks all major
search engines, reporting your rankings for whatever terms you specify.

There are also a number of desktop tools that bring all of this
functionality (and more) together. However, for the purposes of
assessing an existing site, the free online tools will be adequate. You
need to monitor your site’s visibility on an ongoing basis,
especially when tracking marketing campaigns. In this situation a
desktop application may be more convenient.

Of course, knowing that your site ranks 4653 on Alexa or that 364 people link to it from del.icio.us,
isn’t in itself that useful. The real power of online visibility
trackers is that you are not limited to checking your own website. You
can check on your competition as well.

This article is based on Website Owners Manual, to be published March 2009. It is being reproduced here by permission from Manning Publications. Manning early access books and ebooks are sold exclusively through Manning. Visit the book’s page for more information.