hidden complexities

well, they’re not very well hidden. I mean, you can’t hide things you haven’t defined yet.

I’m currently developing wireframes for UI specification based on user activity flows for applications for a client in financial services which is a sentence a bit more complex than I had anticipated when I started typing it even though I suspected there were elements to it that I hadn’t even realised might be necessary in order to provide context that might make it rather more full-featured than the initial scope I had in my head. which goes some way to providing a metaphor for the user experience design conundrum I’m facing, if, in fact, a conundrum can be faced. its not an unusual problem. I’m designing a solution for a problem that can’t quite be fully articulated. as I iterate on user interaction descriptions and map out user action flows and try to develop wireframes that support those interactions that can be translated into functional requirements for application development, its clear that the initial, narrow scope for must-have features for go-live of version-one don’t tell nearly enough of the story to enable me to successfully capture all the core interactions, let alone the extended feature set that adds the value to applications that might actually convince the business that they really should use it thereby making all this effort worth making.

and this is a complex environment. it is a number of degrees more complex than most web-based, rich internet applications. as a single application that is part of a broader application framework that supports a wide range of trading activities, it is actually a quite small, discrete deliverable, but even that small, discrete deliverable, in the operating environment in which is is embedded, has an interaction model that is as big as you feel the need to define. you could easily take a number of weeks or months refining the activity flows, building scenarios, iterating wireframes and doing all those things you do to try and ensure that the user experience is as engaging and intuitive as is humanly or machinely possible and still, on each iteration, and at every review meeting, you’ll find that there is another enormous chunk of user operations that you didn’t know about yesterday and that chunk begets another chunk and that other chunk requires an altogether different contextual description to enable you to understand its operation and how the user interaction model might need to be considered if, indeed, we are to include this chunk you’ve just told me about in the initial release, which we are, because we actually want it all now.

but this isn’t a bad problem. this is a good problem. as a user experience practitioner, the reason I do what I do is that I like to solve problems and provide elegant solutions. what I don’t like to do is take somebody else’s solution and just draw boxes and arrows around it. in this case, I get a complex problem and I’m asked to provide a simple solution – an excellent user experience – and how I solve it is up to me. perfect. I’m travelling more than five hours a do to do it, and they want it yesterday, but, really, minor irritations when to put them on the balance against job satisfaction.