During Midwest UX a few weeks ago there was a series of paintings that caught my eye. It was six interpretations of the same cow which becomes progressively more abstract. The first one was clearly a cow. By the final one it was simply a series of triangles, squares, and other geometric shapes. This evolution started me thinking about the way that designers do the same.
Start from the known
If you are anything like me you don't design in a vaccuum. When it comes time to begin sketching out ideas inspiration always comes from synthesizing pieces of other sites. The solutions that come out can be easily recognized as coming from ideas that already exist. Those first drafts clearly define the problem in the context of "this is a library catalog", "this is a shopping site", and so on.
Slow evolution
Once the iteration begins the designs begin to take on a life of their own. Most of the time the problem that you end up solving is not the one that you started out working on. After talking with users, showing your work to clients, and chewing over the feedback the design gains an identity. It may be small tweaks here or there or wholesale revisions. Either way after a while the current versions of the prototypes bear little resemblance to the original idea.
This process is like the one that the artist went through as they slowly took the painting of a cow in a field from a reflection of reality to an abstract set of shapes. I can't imagine that the final product was in their hand until they were well under way. Only after experimenting and exploring did they arrive at the six steps that hung on the walls of the canteen.
The final result
The final painting in the series looks like nothing more than a random series of geometric shapes. Bearing little resemblance to the first it was a completely different interpretation of the original idea. The same happens as iteration after iteration changes the way that we design a solution.
The idea that design is a linear process is flawed. Very rarely is it as simple as going from Point A (problem) to Point B (solution) without any detours or distractions. Instead design is often a messy process of constantly changing the goal and focus. Even when the end is in sight there may a temptation to do that "one more little tweak". Eventually we have to decide that the result is good enough (or that we've simply run out of time).
The most important part of any deliverable is knowing what problems were trying to be solved. Make the best effort you can to stay on task. Adapt to the problems that arise. Be creative and flexible. In the end be able to explain how the ideas that you've expressed through your design address the needs of the people you are working for.