Marty Neumeier on Wicked Problems

Marty Neumeier, author most recently ofThe Designful Company, gave a talk atAdaptive Path’s MX Conferenceon Tuesday. Nate Bolt posted some notes about it that included Neumeier talking about wicked problems, one of the areas I’ve been thinking about for a while also.

Bolt writes up the following from Neumeier’s talk, in response to a question about how to avoid bold dumb vision (as opposed to bold smart vision):

Ford had the bold vision of a ford pinto. They made the mistake of deciding their design vision instead of designing it.

That’s what happened with the Aeron chair. They wanted to make the best chair ever, and they threw out the style guide of all previous office chairs to do that. They made a prototype and tried it with potential customers, and they said “it’s sort of comfortable but it’s kind of weird. I don’t know if I would buy it.” Then they worked really hard on the comfort part of it, and then eventually people said “it’s really very weird but it’s super comfortable.” Then it takes off because it is different.

I worry about innovations that aren’t different enough, and the Pinto was maybe too different. Clairol “touch of yogurt” shampoo was going too far. You learn to spot a real innovation by it’s combination of being weird and good.That’s where the real art comes in, doesn’t it? Knowing the difference.

You protect against horrible innovation by prototyping.Test this out little by little. Either in the market or wherever. Stage-gate innovation is what we call it. Ventures do that by giving a little money, then a little more money. Businesses want to get into the market immediately, and they are impatient. So they take very small risks. That’s just me-too-ism. Anyone who can help prototype. Herman miller said “we’ve gone this far, let’s go one step further and keep trying it in the market place.

(Not to knock on Nate in any way as I know what it’s like to try and take notes while somebody’s presenting, but I just want to caveat what I’m about to write by saying that I don’t know for sure how accurately Nate’s notes capture what Neumeier said or intended, so I could be boxing ghosts here.)

I’m not quite sure what the first paragraph means without more context, but I take it to mean he’s drawing a distinction between a vision that is set from top-down and not tested out, versus one that is tested through a design process.

Undoubtedly Herman Miller did a lot of testing with the Aeron chair, and if Gladwell is to be believed (and I’m assuming Neumeier is familiar with Gladwell’s account - he gives a good talk here) it did indeed test poorly for aesthetics.

I have a hard time going along with the idea that the Pinto was more different from what was out there than the Aeron however. The Pinto looked like a small American car and was designed to compete with small Japanese cars during the fuel crisis. The AMC Pacer was way more different, and look how well that did, though it is something of a cult classic today. But that didn’t do AMC much good.

I agree with his statement about spotting innovation by its combination of weird and good, but then calling it an art rather defeats the next statement of testing out with prototypes — that isn’t really how art works.

Then things get a bit troubling. Yes, we want to reduce our risk by prototyping, and that does help. However, to hold up the Pinto and the Aeron as opposing examples seems a stretch. I’m sure the Pinto was focused-grouped too (though focus groups were very different back in those days).

Neumeier seems to be switching between saying you need to set an audacious vision and then stick with it, and that you need to prototype iteratively and get feedback incrementally that you then use to refine the design.

The Aeron was tested, but what did Herman Miller do? They didn’t change direction. They stuck with the plan and ploughed ahead. Perhaps partly becaused they’d invested several years of work into the chair (HM is famous for spending four or five years to design a task chair) and the architecture of the design was basically baked in. They could not take the weirdness out of the chair without drastically changing it.

So Herman Miller took a bet the farm risk, tested with prototypes, but they ignored the input, for whatever reason. Herman Miller had a “smart” hunch about how the chair would ultimately be received. They took a big risk that they knew what the fine line was between stupid and clever. They could easily have been wrong, and then case studies would be written about how they ignored focus group input and were too stubborn to change direction, so the chair was the biggest flop in their history.

The fact is, sometimes innovations succeed or fail through pure dumb luck. And we laud the ones that succeed in hindsight, but how many of us could have truly predicted them ahead of time, even if we prototyped the hell out of them?

Lessons in Innovation from Spore

Electronic Arts finally released Sims-creator Will Wright’s newest game, Spore this week. I’m not really a gamer, but it looks to be an intriguing and engaging game. I actually saw a demo of a prototype a couple of years ago (Wright demo’d it prolifically during its long gestation period), at an outstanding talk Wright gave along with composer/producer Brian Eno in San Francisco. The talk, put on by the Long Now Foundation, explored a common theme between the two men’s work of starting with simple elements and creating increasing complexity with relatively simple rules.

In the case of Wright it had to do with simple game elements working within simple rules about how they could be combined and how they interact. In Eno’s case it had to do with small numbers of notes, some simple rules about how they should be played, and then setting them off and running to see what melodies they create.

There are some profound lessons for innovation there. I wrote at the time:

The talk helped crystalize some things I’ve been thinking about with wicked problems, and issues of emergence and how one can use seeds (in Eno’s words) to generate and test ideas, rather than trying to build forests. This is a key aspect of dealing with wicked problems: rapid iterations of probing and testing to further understand the problem.

Plant seeds, not forests

When you’re trying to tackle a new problem where the problem itself is unclear (a characteristic of wicked problems), it is usually foolish to go whole hog and launch a full-fledged product (i.e. a “forest”). Chances are it is going to miss the mark and fail. So an incremental approach that acknowledges the emergent nature of the problem is far more effective. In other words, plant seeds and see which ones flourish.

In contrast to expensive, complex forests that take a lot of resource investment to make, seeds are cheap, disposable, and resource light. Spread them as liberally as you can, with some forethought as to which areas of soil look the most promising (that is, do some pre-screening of the opportunity areas).

Seeds in this case can mean prototypes, sketches, white papers, limited market tests, focus groups, even casual conversations with end customers. Really, any means of trying out a new idea, whether in early stage development (where cheap and fast are key) or actual product launch.

In the latter case, the mentality should be one of incremental launches that ideally build on one another and get one piece right at a time, rather than trying to build the whole thing right from the beginning. This is also a key principle behind the approach to software development known as agile programming, and can be seen in the endless stream of “beta” launches from many software companies that have real users bang away on the emergent offering.