Thursday, October 27, 2005

Skunkworks teams for innovation

Continuing on the theme of innovation, there is another common technique used by companies to produce innovation. Often it used to develop somebody's innovative idea and is sometimes used to generate innovative ideas as well. This is the skunkworks model - or to give it a less jazzy title: separate your imaginative team.

In this model, the team that is to produce the innovative product is separated from the main organisation. The people involved a ring fenced, they may work at a separate location, they are removed from the day-to-day life of the company, and in particular, the politics and blocks on innovation that exist normally. Sometimes these teams are kept secret.

This technique has been documented in countless stories, indeed, Lockheed Martin have trademarked the term skunkworks. If you want to know more about this method you could read Coplien and Harrison's Skunkworks pattern and in their book, Organizational Patterns of Agile Software Development, 2005.

(I have also discussed the technique in pattern form, Separate Imaginative Teams.)

Hamel and Prahalad also noticed this technique in their Harvard Business Review article, Corporate Imagination and Expeditionary Marketing - May-June 1991.

There is something macho about this technique: the image of a bunch of brave souls going off to design and create a new product, cut off from the Corporation, free from the politics and infighting. And sure it does work, companies do create new products this way, however, this technique also has a downside.

This technique may create a new product, it may bring about an innovation, it may get you out of a hole right now, but it does little to make the overall organisation more innovative. In fact, it may detract from the overall company's innovative ability.

To start with the innovative people are separated from the rest of company so none of their expertise or experience is directly accessible by the rest of the company. Neither do they form role-models for other people in the company. Many times, the new product development is invisible to rest of the company - they just get on with their regular work.

And when the product is produced in must somehow be folded back into the company. The rest of the company may not understand the new product. Indeed it might be quite different from the company's main products. Therefore there is a learning curve, while the new product becomes part of the stable.

The people who have created the new product also need to be folded back into the company. But while they have been outside the mainstream, they may have got used to a different way of working, a freer environment, a lack of politics or structure. To these people re-entering the corporate fall might be difficult. Indeed it might be easier to leave the company altogether.

Meanwhile, the people who remained with the company working on the old products are not working on the shiny new thing, they may become resentful of those working on the new product - especially if the skunkworks team are seen to be given privileges or better resources.

And there's not forget one the points made by Arie de Geus which I noted a few days ago: it is not just the taking of decisions that takes time but the acting on them too. The people outside of the company have made all sorts of decisions, and when they returned to the company they expect those others to act on them - there will inevitably be a delay. Indeed, there may be a repeat of the learning curve.

Perhaps, one of the most famous examples of this approach was Xerox's Palo Alto Research Centre - or PARC for short. Xerox set up a research centre on the other side of the country, stuffed it brilliant people and gave him plenty of money. The project succeeded, it invented most of the features we find in the modern PC.

But the project also failed, the team was so far removed from the main Xerox Corporation and the company could not usefully exploit their innovations. Still the researchers found a way, and many of them left Xeroxto found successful start-up companies in Silicon Valley, for example Adobe and 3Com.

So, on balance, I am not a fan of the skunkworks approach to innovation. If you want your company to be innovative you have two embed, within the company, and within the values.