plan to throw away good designs
Last week at SxSW (which I have got to get to sometime!), Michael Lopp, one of the senior engineering managers at Apple, was part of a panel about design. During the panel, he was asked about the design process at Apple. The Tech Beat blog at Business week has the best description of his complete answer.
One of the points that he brought up was the need to plan to throw out good designs. He put it as "10 to 3 to 1". Start off with ten entirely different good designs. Good is the key word there -- they can't just be a couple of good designs and then a bunch of other junk to make the good ones look even better. You have to make ten really good designs, which is to say that you have to come up with ten different ways to reach the same goal. In doing so, you stretch yourself in thinking about the problem and your design. Then you whittle from ten designs down to three, iterate on those three designs, and finally end up with a single strong design.
Lopp isn't the only one singing this tune. Bill Buxton (whose latest book is currently top of my to-read pile) and Alan Cooper (his latest is also in my to-read pile, but a bit further down the queue) have both been talking about this in their own way. Buxton talked to Channel 9 about design and user experience last year, and gave a keynote at Interaction 08 earlier this year about the design ecosystem. Both are fantastic talks, I can't recommend them highly enough.
This is how you get design right. You have to really invest in it. Investing in it isn't about money, it's about time. Your designers need time to think about solving the problem, and they need to tinker with lots of ideas that get towards solving the problem. They need to invest time in coming up with ten good designs. This means that you have to be okay with throwing out good designs. You're throwing out good designs because good isn't good enough — you're looking for greatness. Greatness rarely springs fully-formed from your forehead (no matter how much you might wish it would). Greatness comes from a lot of work, and a commitment to the work that is necessary to achieve that greatness.
It's a huge commitment, but the payoff is also huge. In last month's interview with Fortune, Steve Jobs talked about 'push[ing] the reset button' on the design for the iPhone because he couldn't 'convince myself to fall in love with this'. Everyone talks about setting the bar. Setting the bar to something that you need to fall in love with? What a design goal! But look how many iPhones Apple has sold. Would they have sold so many if the design weren't something you fell in love with?
Comments
Anonymous
March 17, 2008
Nadyne, what do you think about the reference to hi-fi, pixel perfect designs?Anonymous
March 17, 2008
The comment has been removedAnonymous
March 21, 2008
When I read about that presentation on Apple's design process, I was so extremely jealous. In my experience, it's almost impossible to convince the project manager of a software team that it is worth their while to invest time in a design. For them, if there's one good design, that's good enough. We're finished with designing, on to coding. Sofware developers can understand the benifits of iterative programming strategies, but not iterative design. It can be beyond frustrating. I read Bill Buxton's book and what I love about it is that he is trying to educate the "management" on the design process rather than the designers. But I wonder, how can I get my manager to read this book? I don't think he ever will because it looks like a design book to him. That's not what he's interested in. I'm lucky right now because my boss does let me redesign, but only if I can convince him beyond a doubt that the reason we're stuck in our development is because of a design flaw. It can be like pulling teeth sometimes and it really shouldn't be. In my ideal world, it would be part of the project plan that every few months we go back to the design and pull it apart. Unfortunately I don't make the project plan ;)