Innovating through these styles generally takes longer than innovating through Cauldron, Spiral Staircase, and Fertile Field. But they can accomplish big tasks that the others by themselves could not.

The PacMan style works best when you don't know what the next big ideas in your industry will be. But it only works in a fragmented field, where there are numerous young companies with which a larger firm can do deals. The PacMan player invests in smaller firms. They must not only be working on projects significant for the long term but also on projects whose success the player can effectively judge in the medium term, so that the investor can recognize which of the struggling young companies should be purchased. When those conditions apply, a company with resources can acquire partially proven technologies and put them together into profound strategic capabilities - sometimes at modest cost. To a certain extent, Microsoft has pursued PacMan innovation as a long-term strategy, acquiring companies like WebTV and effectively betting on its innovation.

The PacMan strategy amounts to out-sourcing the early stages of the innovation cycle - coming up with ideas and testing them in the market - to entrepreneurs and venture capitalists. It makes most sense when there are just too many potential ideas for a company that might prove winners for any company, even Microsoft, to bet on them all at the first stage. Better to let the market screen these down to fewer promising innovations, absorb these winners, and begin the process of scaling them up to full-blown businesses. If one's company has the resources to pursue this strategy, the PacMan style becomes an efficient way to acquire successful experimenters.

Two key levers drive successful PacMan strategy innovation:

• Starting with strong internal R&D capability, so you know the fields you're investing in.

• Developing a well-defined process of integrating purchased companies into your existing businesses.