• The leader shared the rough picture first with few trusted associates, who modified and refined it and begin to experience passion for it.

• They then shared it with larger and larger groups, which refined it further, added details, and started finding ways to remove the barriers that kept the vision from happening.

• As large numbers of people removed barriers and began working on changes, innovation after innovation occurred. Effective Cauldron-style strategy innovation processes often get Spiral Staircase-style and Fertile Field-style innovation processes going at lower levels in the firm.

Houston-based Enron, which has grown from a gas pipeline company into a diversified $30 billion-a-year global energy firm, exemplifies the Cauldron style. The company's key approaches to innovation originated in the late 1980s when Jeff Skilling, now Enron's president, created what evolved into the Enron Capital and Trade Resources business unit. Skilling recruited and promoted creative dealmakers. He showed them that they could make big money by developing creative new energy deals and working with their peers in the finance, legal, and other departments to get the resources to make their deals real. Ken Rice, who now heads the division that Skilling created, describes the resulting work environment as a "nuclear fusion reaction." Enron eventually developed ways of evaluating new ideas that were remarkably well understood within the organization. A large corps of Enron people became repeat innovators.

Enron's experience shows that the vision that drives successful Cauldron innovation doesn't need to be precise, and in fact it can't be. The senior leader can't possibly understand the achievements that will emerge with precision.

Another example is the launch of Lucent Technologies as a spinoff of AT&T divisions that had been growing at only 2% a year. Senior managers' "roughly right" picture of the future consisted essentially of three elements:

• A strong sense that Lucent had to turn itself into a "high performance growth company,"

• A sense of how the organization's values had to change to make that possible, and

• Some numerical goals developed through benchmarking successful high tech companies.

Lucent did not develop routines for evaluating and approving new ideas as well as Enron. It has recently experienced some real problems. But the work of Henry Schacht, Lucent's CEO during its spin-off, remains a model of how to launch Cauldron innovation. He demonstrated an honesty who showed he knew a little bit about the future but admitted how much remained unknown. This inspired hope and excitement among people who could make big contributions to innovation. With widely-shared passion and broad, albeit vague, vision, the Cauldron can keep boiling for a long time.

But success also requires structures that make continual change possible. Groups practicing Cauldron innovation adopt an "organic" style of work, in which key people are often unsure of exactly what their job is. They're expected to find work that's important and do it. And yet structure isn't less important just because it is looser. The Cauldron innovator needs institutions and habits that allow ideas and resources to move effectively from one part to another. It may require a fast-moving "internal market" for ideas and people, for instance. If innovators' ideas fail to win support in one unit, they need to be able to take them to another.