Decision-Making Structures and Innovation

You can’t design great (innovative) products as a group -- you need a core vision supplied by one person who calls the shots. Additionally you need to have decentralized power to get things done. Centralized decision-making will never result in anyone's ideals because employees need to be close to the people they are serving to develop ideas that make sense and will actually work. Local knowledge is critical if you want to understand how consumers feel about your products’ cost and quality. The ops VP wants a product that rides smooth, the engineers want speed.  They are trying to make a horse, but will end up with a camel. While a supportive C-suite is critical to your eventual success, you can have too many smart cooks in the design kitchen. 

The other problem with design by group is that things take far to long to bring to market. Consier Kawamoto’s comments on why Honda was lagging behind other Japanese manufacturers in the early 90s.

We’d get the people from research, sales, and production together and everyone would say “not this” or “not that.” We’d talk but there would be no agreement. Product planning would be on a tight schedule but we would have another discussion, another study and more preparation. Finally, the decision would come months later.

-  Nuuhiko Kawamoto (Former Honda CEO, 1991)

Now don’t get me wrong, team decision making can be great – it’s valuable in environments where the relevant specific knowledge for the decision is divided among employees and where the costs of group decision making is minimal. In general, when you’re trying to develop a new product, you need a strong core of individuals who provide their opinions, but one person who is ultimately responsible and can pull the trump card as necessary.

If you’re up for a super geeky treatise on why group decision making doesn’t always work, check out Kenneth Arrow’s monograph ‘Social Choice and Individual Values’. Arrow’s work describes the voting paradox – when majority voting fails to result in the best outcome because group values derived from groupings of voters will never result in a product that works for any one individual. Want to bring your product design decisions to a vote? Expect the results to disappoint when groupthink and majority/political campaigning destroy your vision. 

In this case, what’s good for the gander isn’t any good for the goose.