Re: question about estimating relative importance ofucd
6 Jul 2005 - 1:19am
I really did mean the relative _importance_ of UCD.
If a particular system has three users and a bad user interface means they
spend twice as much time as they otherwise would doing fairly regular tasks
and they had to spend half a day reading documentation each to figure out a
few basic features and never quite get the rest, then obviously there is a
cost associated with this bad design. On the other hand, there is also a
cost associated with doing user research and designing a good user interface
and possibly with reengineering the entire system to make it possible to put
this good user interface on it (which now takes these three users hardly any
time to learn and half the time to complete their tasks). Very cool. But the
effort to create this good design _might_ cost the company more than they
would save from the increased productivity of the three users. In which
case, I would rate the relative importance of the UCD effort pretty low.
On the other hand, if we're talking an interface that's directly
revenue-producing and core to my business, that potentially has millions of
users who pay a small fee every time they use it and who have lots of other
choices among the competition and low switching costs, and who need to learn
a whole new way of doing things to use my thing (whatever it is), then I
would say the importance of UCD is pretty high.
This example uses $$, because many projects in a capitalist society are
about profit. But plenty often, I might need to measure it in increased
productivity (if it were relevant and I knew how to quantify it) or in
behavioral change or educational value (if, let's say, I'm a green
revolutionary who wants people to use my "thing" to understand how their
lifestyle impacts their environment and to change their spending
accordingly) or in the ability to do something that could save/cost a life
(as in advanced medical applications). Undoubtedly there are other metrics.
That's really not what concerns me.
If we start with the assumption that resources are scarce and that the
potential work these resources do is practically endless, then it would be
extremely helpful for me to be able to figure out which projects to apply
them to. And it probably isn't going to be an interface that only I would
ever use--I'm not that important, even in my own eyes.
[Please voluntarily trim replies to include only relevant quoted material.]
I am not certain what you mean by "importance" in the context of your post.
However, my ignorance has seldom kept me silent...
If the measure of importance is cost savings, or revenue increase, or
whatever else, then the ordinal relative importance of some degree of UCD
could probably be estimated with reasonable certainty for a specific pair of
My hunch is that you might be saying the extent of ethnographic research or
contextual inquiry required to understand the target users and their
context(s) of use could be substantially less in the case of 3 domain
experts. If so, UCD might be easier in that situation, but not necessarily
Even in the simplest case: designing a tool for just yourself, UCD is very
valuable. It just happens to be easier because you ARE your user, your goals
and preferred methods are well understood if not yet stated explicitly.
In your example of 3 experts, optimization of the design is very important
because they have to interact with the design continually to achieve most of
their work-related goals. In your example of domain novices who only
occasionally use the artifact, optimization is also very important - but to
different ends for different reasons. They need a more discoverable
interaction in which time-efficiency is less important.
UCD is a means by which to optimize either design.
I don't think any broad generalization about relative importance of UCD at
either end of such a continuum can be very useful. (Nor is this post, in all
Opinions expressed are necessarily mine and not necessarily those of my
I'm looking for resources that address the relative importance of UCD to
different projects. All the materials that I can find appear to assume that
UCD is equally important in all cases, whereas I'm pretty sure that it's
less important for a project that will have 3 end users who are domain
experts and who would spend 95% of their day using the system than for a
project that would have 4 million end users who don't really understand the
domain and would use the system once a year.
It would be great to find some discussion of this general notion and some
approaches for estimating beforehand how important the user experience will
be to the success of the project.