Graphics Reference
In-Depth Information
do. Having found a way to make the product function at a minimum level of quality, he
declares victory, convincing himself that his discovery has led to the best outcome. Though
our design could function much better if he would only learn to use it properly, he never
does. Instead, the user gives up and decides that the product is working just fine. A good
designer sees this coming. This is when it's time for the designer to make choices, stripping
down menus and options to their optimum number.
Branding is at the core of most marketing efforts, and has a vital role in design. A com-
petent digital designer must be aware of his client's brand. His products must fit the iden-
tity the brand has established, adhering to the company's overall vision of itself. He must
ask himself: How does this item fit with this company's other products and services? Can
my design make that fit better? If a company is just starting out, anything he designs will
be aimed at establishing their new brand. Whether the brand is old and established, or a
work-in-progress, the designer must learn as much as possible about a company's identity,
and that knowledge must inform his designs.
Before we can benefit from UX, we have to communicate with users. We can do this
with rough sketches and prototypes, and while a design is still offline, we can test our
designs in a usability lab. This is where we bring in users, set them up with a product,
and watch how they use it. In this way, we can observe individuals' interactions with the
product, noting where our assumptions are met, and where they fail. This is a kind of “re-
tail” user testing. Another method is to take the product live online. This can be limited to
a set of users who've been invited to the website, or we can do it in a full product iteration.
We have tracking tools for both methods.
When we take a product online in a complete iteration, we test it on a much larger
sample. Unlike our usability lab with its individual subjects, here we begin examining the
behavior of groups. This is where user testing becomes more like traditional polling. Like a
good pollster, we slice our sample different ways to learn different things. How many users
remained with us, and how many dropped off? What percentage bought something? How
many stayed awhile, then left? How many came back? What was the average time for per-
forming each function?
We get this data from tracking tools like Google Analytics, Ominture SiteCatalyst and
Mixpanel. These help us slice up a large audience into its smaller, more identifiable com-
ponents. We can see which groups use it, how much they use, when they use it, and how
they use it. This helps us better understand a product's appeal. Is it attracting the audience
we expect? Do we have measureable numbers of surprise visitors? We gather the numbers,
do our calculations, and then use that data to better adapt the product to its potential audi-
