|
|
|
Rethinking the methodology of website creation
from an information architect's point of view
What is a website, really?
What is a website? Let's look at an analogy from old media.
What is a novel? One way to look at a novel is as a stack of
paper, with an assortment of words printed on each page.
While that may be a true, the physical appearance of a novel
is really only one of its accidental characteristics. After
all, that stack of paper could just as easily be a book on Javascript
coding as The Great Gatsby. What really makes a novel is not
the paper, or the words on the page, but the story. And the
book itself is only half of what is needed to make a story -
the other half is a reader. It is only once the words on the
page are read that the book stops being merely a pretty doorstop
and starts becoming a novel.
So what is a website? It seems that just as a novel could mistakenly
be viewed as nothing more than a stack of paper, those of us
who work in new media have fallen into the trap of thinking
that websites are nothing more than stacks of data. This is
the failing of many a modern information architect (those of
us who dictate the structure of websites) - the belief that
the job is essentially about arranging information in a certain
way. Of course, that's not true. A website is no more a pile
of information any more than a novel is simply paper - a website
is a complex interaction between data and user, and it is only
in that interaction that it has meaning and becomes real.
So an information architect shouldn't just be someone who "says
what goes where". In reality, an IA is the interface between
data and users, between coders and designers. That's why the
best IAs come from multidisciplinary backgrounds - understanding
all the pieces of the puzzle is essential to seeing the overall
picture. As this interface, an IA should know that a website
doesn't really exist in space (be it virtual, or in the two
dimensional space represented on the computer monitor) but in
time. It isn't a really a thing at all; it's an event, and a
specific type of event - an experience.
Where is conventional information architecture going wrong?
Though a relatively new discipline, the job of an IA is well
defined. The deliverables expected on any project are a sitemap
(i.e. the underlying structure of the site) and wireframes (i.e.
the conceptual framework for the interface). While having pre-existing
models and methodologies for site creation is undoubtedly a
good thing, the system as it stands is not flexible enough for
the future of the web. The problem is that rather than asking
of the current sitemap/wireframe system "Is this the right tool
for the job?" we ask ourselves "How can I make this website
experience fit into a conventional sitemap and wireframes?"
And the answer is that sometimes a traditional sitemap/wireframe
combination might not be the right tool or model needed, but
because of the usual deadline constraints we simply make do
with what we already know. We work arse-end first - an information
architect is defined as someone who makes sitemaps, rather than
a sitemap being defined as one of many things made by an information
architect.
As a rule, traditional sitemaps have two axes to represent the
structure of a site, with sections running horizontally, and
pages within sections running vertically. A sitemap like this
presupposes that the site being mapped is static and made up
of easily categorisable stacks of information. Which is generally
fine for representing content that conforms to a linear, hierarchical
pattern.
However, not all sites fit into that mould. And even if a site
does contain such logical and well ordered information (after
the IA has tackled the client's raw content - obviously no site
is comes pre-packaged in such a neat way, or else IAs like me
would be out of a job) the question remains as to whether it
should be modelled using the traditional methods. After all,
users' experiences on the web are rarely ordered and methodical,
but instead are far more cross-contextual and freeform. We should
challenge the underlying assumption that presenting the content
on a site in a hierarchical, linear way does in fact provide
the optimum user experience.
IAs are guilty of working to old ideas - that users are going
to treat websites as they would a shop or a library, with aisles
and a card catalogue. Even if that were once true, as the web
matures as a medium people use it in new and unexpected ways.
And so the emphasis should shift from information to experience.
What the optimum user experience is in any situation is always
up for debate - we'd be sensible to observe and learn from as
much user testing as possible.
Wireframes are also becoming outmoded, as they can fail to capture
the rules of interaction that complex dynamic systems require.
Often the key details of how a website is actually supposed
to work - what connects to what, and what every button does
- can become lost, especially on large projects, as the responsibility
for recording that information falls between the IA, the technical
architect and the business analyst (if there are any working
on the project), the designers and the coders. Ultimately the
IA should be the guardian of that knowledge, but that means
the IA has got to stop thinking in terms of structuring information
and start thinking about mapping an experience.
New challenges, new opportunities
The trick with experience is that it goes beyond the merely
creating an accessible site, and a usable site, to creating
a pleasurable, rewarding, satisfying and enriching site. The
optimum experience is not necessarily the one that simply gets
users from A to B the fastest (although that will always be
a consideration) - it's also the most pleasurable.
So where do we go from here? In the future there are going to
be an increasingly diverse range of experiences on the web,
and wherever we go entails finding new approaches for the mapping
and navigation of those experiences that move beyond the conventional,
flat and linear sitemap/wireframes approach. These new approaches
will, of course, co-exist with traditional IA deliverables -
after all, there's no need to reinvent the wheel, and for some
sites (those which require heavy logical structuring) a hierarchical
sitemap will remain the right tool for the job.
So IAs should take the well founded principles of information
architecture, and the best practice learnt through experience,
and add to that an exploration of interaction (especially in
non-linear, unordered, emotional experience). Out of this will
come new maps, models, metaphors and navigational devices that
are optimised for different types of experience (with some navigation
and mapping techniques no doubt being bespoke to unique projects).
The IA should continue in their critical role as the interface
between code, design and users, but should also take their place
at the heart of the creative team, and be unafraid to think
outside the box.
Some areas to explore
In practical terms, here are some areas that I think it would
be fruitful to explore, some ways of thinking about website
experiences that may give us new models and tools for us to
use, and some ways in which we could make website experiences
richer and more rewarding overall.
Non-hierarchical, non-linear
Not all sites can be made to conform to the conventional horizontal
sections and vertical pages sitemap structure. Think of social
networking sites such as Friendster for example, where any given
page is merely one point in a giant network, with no start or
end point. What is the most effective way for users to navigate
through such a network? How can we help them visualise the site
as a whole?
Fluid, malleable, personalised
Why should a website have the same structure for all its users?
Shouldn't there be ways that the site structure and navigation
can be optimised for each individual, and to the nth degree?
Could this work and still present the appearance of a unified
site for all users?
Narrative
Would some sites benefit from a narrative structure? Could a
site work more effectively if we think of a defined beginning,
a predesignated conclusion, and the highs and lows of a gripping
story in between? And this needn't be a linear narrative. After
all, that's the beauty of hypertext in its purest form - the
part it plays in the formation of non-linear branching narrative
structures.
Fruitful mistakes, unpredictability
Why should we tightly define what the 'correct' route is through
a site? There's more to a website experience than simply navigating
from A to B. Perhaps, if we loosened up our strict ideas about
error prevention, we could allow more for users to make fruitful
mistakes, to find things that they didn't know they were looking
for, but are happy to find. Think about how we often buy things
in shops we didn't go in for - how can we replicate that on
the web?
Chaotic, unordered, random, surprising
In relation to above, perhaps some sites need not be ordered
or logical at all. At the end of the day, while we consider
putting users in touch with mission critical information to
be one of our top priorities, finding such information is rarely
a key objective in an average web user's day. Instead they want
to be surprised, to find new things, and be entertained. That
is, in essence, the whole point of surfing the web.
Responsive, tactile, human
When we use a website, it's very much a case of us telling the
site what to do. It doesn't do much itself really, beyond respond
to our commands to bring forth certain information. Could we
make a richer experience if the website could be more responsive,
provide more feedback, and maybe even portray some sort of personality
in its responses? How could we develop more tactile and responsive
navigation devices?
Filmic, atmospheric, ambient
In terms of thinking about different ways to model a website
experience beyond mere stacks of information, perhaps some sites
could benefit from being visualised as an environment, something
into which the user is placed, and surrounded. This isn't to
mean that a site like this needs to be visualised as something
three dimensional, but it would be immersive and atmospheric.
Musical
As a variation on the theme of Narrative above, perhaps we could
take a musical approach to site structure, and view it as having
a rhythm of its own, and 'melodic' themes that run throughout.
Can a site's structure itself be a thing of balance, interplay,
and ultimately great aesthetic beauty?
Playful, teasing, flirting
Returning to the idea of a site's personality, perhaps it can go beyond being helpful
and start to become playful? Perhaps it teasingly returns deliberate
misinformation every so often, before it delivers the information
we really want, enriching the experience with the delicate interplay
of pleasure and pain, desire and satisfaction.
Self-intelligent, code-lead, learning, deriving structure
from within
Why should we enforce structure on a site at all? Perhaps the
code knows better than we do what form it should take.
Necessarily incomplete, growing & multiplying systems
Many sites are necessarily incomplete - how can we map and model
such systems to allow and encourage future growth without any
further intervention? And not just in terms of adding a page
here and there, but allowing for fundamental structural changes
to occur.
The principle of folksonomy applied universally
Folksonomy is an increasingly popular concept whereby web users
define their own taxonomy, usually for their own encyclopaedia
projects. Why not let users collaborate and define their own
taxonomy for a site? Why not go one step further and allow them
to collaboratively define and mutate site structure and navigation?
How could that work in practice?
CC Licensed 2005 by iivix.
Some rights reserved
You are free to copy and distribute this work under
the following conditions:
I, iivix, am attributed as the author.
You do not alter the text in any way.
You are not permitted to use this work in commercial situations. If you do want commercial use of this work, please email me:
|
|
|
|
|
