[Sigia-l] The fuzzy line btwn IA and Design

[Christopher Fahey [askrom]]

Ziya wrote:
> You can essentially bring any type of non-brain-dead vector 
> format into
> Illustrator and 'roughen' it with a preset style or even have slightly
> different roughen settings for an 'irregular' look throughout 
> the page.


Deliberately making wireframes look more low-fidelity than they actually
are seems to me to be a bit backwards. I have an inherent reaction
against dumbing-down anything - it's usually a sign that there is
probably a more elegant and efficient solution. 

First, here are several reasons why wireframes are often constructed in
low-fidelity, in order of what I think to be their legitimacy:

(1) FOCUS: It's a waste of the IA's time for them to bother with graphic
design details. The IA needs to be able to modify their wireframes
quickly and frequently. Making changes and on-the-fly modifications to
the information architecture is a lot easier when the IA is able to
focus on just the IA. 
(2) TURF: The wireframe should not 'steer' the graphic designer towards
a particular design approach, particularly while the graphic designer is
still working on the high-level design specifications. The designer
should have creative breathing room. Oftentimes, the IA may have
less-than-adequate visual design skills.
(3) CONFUSION: A high-fidelity wireframe might confuse stakeholders
(clients, managers) into thinking that the wireframe is the final
design. They may sign off on a visual design layout

FOCUS is by far the most compelling reason to do low-fidelity
wireframes. We IAs are designing interactivity and planning relative
content positioning, we're not (usually) doing typography, layout, or
illustration. In a tight schedule, the IA should be able to nimbly
juggle page elements, modify forms, change nomenclature, etc, with
complete ease and fluidity. The FOCUS rationale is why we commonly use
Visio or Powerpoint and not Photoshop or HTML.

TURF is sometimes a good reason for low-fi wireframes, but only
sometimes. While a site is still in the conceptual stages, it can be
helpful for the IA to provide low-fi wireframes of key pages to
designers. But it is incumbant on the IA to work with the designer as a
50/50 equal partner in laying out page elements, even before the first
wireframe is created. The TURF rationale is only valid insofar as it can
help the creative/collaborative process (as a kind of brainstorming
tool), but the better your relationship is with the designer, the less
validity the TURF rationale has. If the layouts are produced in
collaboration with the designer early on, then really there's no ongoing
TURF reason for the Wireframes to be low-fidelity. Super low-fidelity
wireframes are quite useful for a few days while the full experience
design team discusses all of the options, but once brainstorming is done
the IA can ratchet up the fidelity of their deliverables for the
remainder of the project.

The CONFUSION rationale, where clients confuse wireframes for finished
pages, is one that IMHO I've managed to defeat. Before I show a
wireframe to somebody, I give a long, substantial speech about what it
is and what it is not. I listen to the client, look into their eyes, get
a feel for what they think they're going to see. I set VERY low visual
design expectations, I profoundly insult my own design skills, and
basically try to give the client the impression that I am one of them: a
non-designer. In essence, it is the IAs job (as always) to understand
their users. In the case where the 'user' is a client who cant tell a
wireframe from a design comp, then it's your job to explain the
difference and make sure the explanation sticks, not to spend energy
making wireframes that contain less information and detail than what is
available. In my explanation, I even mention specific things that might
change along the way: "This navigation may move to the bottom of the
page, this menu might be collapsed in a DHTML tree, who knows?".

All of this is to say: If you have the time, if you have the skill, and
if you have an effective professional relationship with your visual
designer (if you don't, shame on you!), there's no compelling reason to
deliberately dumb down your wireframes. Your schedule should be the only
thing that determines how high-fidelity your wireframes will be. 

Like I said in an earlier post, if I am sure the final product will use
Times New Roman on a light blue background, then my wireframes will
certainly have Times New Roman on a light blue background. Why the heck
not?

-Cf

[christopher eli fahey]
art: http://www.graphpaper.com
sci: http://www.askrom.com
biz: http://www.behaviordesign.com