ForUSE - Keynote Speech - Bill Buxton

These are raw notes...

Initial distraction about two handed design, designs systems for use with both hands. Inventory of skills a particular audience has acquired. EG musicians, painters.

Everyone is bad at making "new products" in the software industry (not n+1 or new releases). To allay this, most companies will buy a new product and make it an n+1. Why are other sectors that maintain high standards of new product able to do this but the software industry can't? Other industries have a design phase in front of engineering for making NEW products. This sits really comfortably with me – drawing a distinction between a NEW product and a n+1 product. N+1 products don't need as much design up front as totally new ones do to ensure it is as good as possible.

I wish people didn't ask questions that are designed to just needle a presenter. There's no point to the question and the answer doesn't help the understanding of the audience. Ken Schwaber said "no questions if they're longer than the answer". I like that rule.

Status quo is engineering to sales. We discussed last night that the design process needs to incorporate the time after launch and factor in post release maintenance, support and version updates.

Film industry example – the PACKAGE before work begins, or "green light" - Stars signed up. Nailed down with contracts FIRST. Script. Production Schedule. Release date and season. Title and campaign. How much will be spent on marketing. Location, Director of photography with photos of the location... etc. Then the decision is made. Lesson in Venture capital pitches - thorough scoping.

Automotive Design example – sketches, clay models, work closely with engineering. Based on a platform (chassis). Engines and Platforms are hugely expensive and are thus reused. The shell and styling are what make the car. End of the design phase has a concept car, that looks like you could drive away in it, but it's all clay! That happens before the 'green light' is given. Feedback from the audience. 'focus groups' are shown the concept car at shows etc to gauge potential success.

Design phase comes before 'green lighting' a project. This includes all projections of marketing potential, cost and acceptance. We need to have a marketing department who understands what we need from them.

Innovation can still come in the engineering phase, but through implementation, not design changes. It is also essential that there are engineers in the initial design team. A lot of car companies are trying to reduce the design phase through a lot more ethnographic research and contextual enquiry.

Essentially, Bill Buxton agrees with Alan Cooper in saying there's no way to design and engineer at the same time and end up with a well designed and successful product.

Criteria Weight increases as a function of time and investment. Criteria meaning the specifications for success. You need to have a 'dashboard' to continue to monitor whether the product is meeting the criteria you initially set. Think Ken Schwabers' geese finding solutions to a set of requirements when migrating. e.g. of Criterion "80% of the functionality to be understood and learned within 5 minutes" "must be able to be used on all platforms" etc.

Personality types: Designers - change in half an hour, innovate and come up with new features all the time. Engineers - totally organised and anal, not likely to try and introduce things at the end of a build...