Datasheet

(www.mrunix.net/webalizer/) is a similar package, but works only with Apache log formats.
ShortStat (
www.shauninman.com/mentary/past/shortstat_maintenance.php) is a PHP
application that can track various kinds of user data. These are only three such packages, and there are
dozens, if not hundreds, of alternatives available. Each has its own strengths and weaknesses, which
should be assessed according to your needs and technical requirements.
In addition to analyzing server logs, you should interview the site’s stakeholders. These are the decision
makers, the people who drive the direction of not only the site, but also the business behind it. These
people will have a strong bead on the site’s audience, and ideally have had close contact with them. As
such, they can provide valuable insight into your users’ needs, and into which areas of the site would be
most relevant to them.
Of course, the best method of creating effective personas is setting aside internal statistics and assumptions,
and actually meeting some of your users — after all, facts and figures can go only so far in identifying just
what it is that your users value. At some point, you need to set aside quantitative data for qualitative inter-
views; there is no substitute for sitting down with people who will (or currently do) use the pages you’re
designing. Talking with them about their needs and goals not only creates a vital feedback loop for you as
the site’s designer, but also helps you put real-life anecdotes and experiences behind the design decisions
you’ll be making. For example, you might poll your users on any of the following points:
❑ Technical information. What kind of browser do they use, and on what kind of computer? How
do they use the Web? Why do they use the Web?
❑ Customer information. How do they view the site of your company or client? Do they use it?
What do they think of the site’s competitors?
❑ Personal information. This might include such information as age, gender, and location (for
example, urban or rural).
❑ Design preferences. How would they define a “good” site design? While you might not ask
them to leap into an art school–esque dissection of a given site’s design, you might ask them to
tell you some of their favorite sites. Try to find out why those sites are their favorites.
Additionally, you could try to uncover what sites they like least, and why.
Of course, this isn’t a comprehensive list— the goal of any user sampling is to get as much data as possible
that will be helpful to you in your design effort. But rather than just seeing this as a data-mining initiative,
think of it as a series of conversations with real people. While demographic information and technical
statistics are vital to planning your site’s development, collecting anecdotes and quotes from actual people
will help to create a more effective persona. Remember that these are ostensibly the individuals that will be
visiting your site, and whose needs your design will need to address. When you think of them as users,
your design suffers; when you think of them as people, your site will be all the more successful.
Once you’ve collected as much data as possible, the analysis begins. By sifting through your notes,
trends should emerge. An overwhelming percentage of your users might use the Macintosh version of
Internet Explorer; a large minority of your users might be color blind, or suffer from poor vision; a high
number might be working mothers, or perhaps teachers looking to acquire professional development
credit. More than likely, your audience will be multifaceted, and could contain any or all of the above.
But no matter the spectrum that your users cover, it’s important to keep these seemingly disparate char-
acteristics in mind as you sit down to write your personas — because just like your “real” users, your
persona may not be able to be easily categorized.
10
Chapter 1
03_588338 ch01.qxd 6/22/05 11:18 AM Page 10