April 11, 2001: I went to a great BayCHI talk last night called "Experience Modeling" (although it was really about concept modeling in general). Among the panelists were Sandy Speicher and Harry Sadler of MetaDesign, who were terrific, providing more useful ideas and perspectives than you usually get from a whole conference. Among the ideas:

They prefer thinking of User Experience as the context in which we work rather than something we can explicitly design. (I take some issue with this. I agree we can't fully design people's experience, but we can greatly influence it through our design and it's our responsibility to do so).

There are various uses of conceptual models, including:
- visual models can make complex concepts (relatively) transparent (on a similar theme, check out the xplane elephant page)
- models can "hold state" of a design, by aggregating a bunch of decisions
- modeling helps you focus on a subset of the myriad aspects of a problem
- models are a medium for thinking, a support for conversation, by making design ideas tangible
- models force out inconsistencies, gaps
- models can serve as specs

Some simple models teach (here's how it is), other more complex models are used to "support a conversation that's been going on a long time." (I love that idea!)

Things we can model include:
- brand (color study, image collage)
- relationships (between customer and company)
- interaction (scenarios of use)
- interface (prototypes, layout grid models)
- an entire project (picture a single map made up of all the separate individual models you used throughout a project)

Someone suggested showing a client multiple models of the same concept and having them choose the one they like and want to design around.

What has input into formation of a model (e.g. of a web site):
- client (business goals, design preferences)
- user (needs, goals, expectations)
- context (competitors, technology, location of use, etc.)
- designer (past experience, aesthetic preferences, preferred tools)