Brian PrenticeThomas Otter has been posting a number of thoughtful blogs on the topic of user interface design. Design is a topic near and dear to my heart. It is my firmly held view that one of the ramifications of the consumerization of IT is that enterprise IT will either become “design-focused” or it will become marginalized, if not irrelevant.
But I think it’s a trap to see user interfaces as the focal point of design. As I’ve been looking into the topic of design I came across an interesting distinction between conceptual and physical design:
- Conceptual design is a basic foundation that defines the structure of the solution, including the functional elements of the product, their relationships and the system behaviour.
- Physical design is a more refined level that defines the aesthetics of the solution. In contrast with conceptual design, physical design defines the success or failure of the product appeal.
The challenge facing this industry is to spend more time on conceptual rather than physical design issues. Only then will we be able to tackle the single biggest impediment to good design in IT. That is our habitual tendency to over-engineer.
IT people, in general, do a pretty good job with product requirements. We dutifully ask “users” what they want. We try to understand the benefits those requests will deliver and then determine which needs get addressed first. But at the end of the day let’s be honest – when it comes to a feature request it’s generally a matter of when, not if, it shows up in a product or project. To paraphrase Will Rogers, most IT folk haven’t meet a feature they didn’t like.
But at some point all user interfaces will break under a relentless onslaught of new features. Products become bloated. Complexity creeps in. Users feel disempowered if not openly hostile. It doesn’t matter whether we’re talking about command line consoles, graphical WIMP systems or web site design. The cycle keeps repeating itself. How long will it be before we find it difficult to navigate the new crop of gesture-driven interfaces?
While user interface design is important it’s essentially retrospective. It’s a way to sort out and present the things a product already does. As a result a user interface can’t act as a brake to the over-engineering which ultimately erodes its value. Therefore user interface design must also be coupled with conceptual design processes that addresses prospective functional requirements in a different way.
These types of conceptual design processes are things I’m planning on exploring through my 2009 research efforts and this blog. But I am prepared to put a stake in the ground on a few principles that I believe need to underpin any such effort:
- It is infinitely more powerful to understand how a technology will probably, rather than possibly, be used.
- A feature is only of value when it is required by a majority of of its current user base.
- When complexity is exposed, design has either failed or was never considered
Category: Feature-itis & The Design Imperative Tags:
1 response so far ↓
1 Thomas Otter January 8, 2009 at 2:53 pm
Brian,
Indeed. Often the best design is no UI at all. Processes that are so well designed they don’t need human intervention.
The tyranny of the UI is something I blogged about a fair bit before I joined Gartner. Business Software needs to be user.centric yes, but the real user is not always the person using the system. The general ledger is not all about the accounts entry clerk, it has a much broader and more fundamental role.
Software needs both design and engineering. I’m fascinated by the Bauhaus movement, for both its successes and failings. Software could learn a lot from them.
I’d really like to cooperate with on the Design Thinking Research. It is very close to my heart, to the point that most of my Christmas presents were design books.