Innovation Incubators: Why Separate is Better

Some conventional wisdom that actually makes sense in this week’s Business Week, from consultant Jeneanne Rae. She suggests four principles for successful innovation:

A clear challenge statement that expresses aspirations to a worthy goal without prescribing the means. This should be expressed in terms of a customer need, not a business need.
A well-designed, well-facilitated process that includes multidisciplinary participation and sources of cutting-edge ideas.
Emphasis on developing concepts that combine multiple elements of innovation (e.g., business model, IT platform, and channel) to increase impact and distinctiveness.
Techniques and structures that counterbalance the forces of risk aversion.

She suggests a number of ways of addressing the fourth and most difficult of these principles, including this one:

Form a special petri-dish environment where new concepts can grow. Pitney Bowes (PBI) has a concept studio designed to explore opportunities far afield from its existing lines of business. IBM (IBM) has a similar unit, called “EBO” for Emerging Business Opportunities. This approach minimizes distraction to the ongoing business and permits concentration of special innovation skills. Successful projects are then ‘sold’ back to the business units.

Almost every innovation guru out there has hammered this point home: When it comes to innovation incubators, separate, and autonomous, is better. Clay Christensen puts it this way:

IBM, when the mini-computer disrupted the mainframe, were very late. But they then set up a separate business unit in Rochester, Minnesota, and when the personal computer ó as it disrupted the mini ó IBM set up a separate business unit in Florida.And they’re the only one of the major computer companies of the 1960s and ’70s that did that, and they are the only ones that survived….If you look at any company that you would say has transformed itself over the last 30 years or so, it is GE. In every case, they achieved the transformation by setting up or acquiring new disruptive business units and selling off or shutting down ones that had reached the end of their lives. In no case did they transform the business model of an existing business unit to cause it to catch the disruptive wave.

There are two reasons for doing this, he explains: First, nurturing innovation usually requires different skills, different resources, a different benchmark of success, a different management style, less aversion to risk, and a different focus from the mainstream business. And second, innovation can be a distraction to the mainstream business, threatening the processes and attention to traditional customers that have made the mainstream business successful. Australian design and engineering company Invetech offers these suggestions for encouraging innovation:

Establish a separate business unit
Set up an ìincubation management armî
Establish technology ìcentres of excellenceî around areas selected for pioneering plays
Establish strategic relationships ñ it is not within your walls but it is within your reach.
ìCherry pickî a SWOT-team reporting to the CEO/Exec team

Even innovative companies (like Sony when it developed its PlayStation) use separate incubators (not to be confused with R&D ‘labs’) for new disruptive innovations. Meanwhile, British Telecom and Shell use a slightly different model, operating separate incubators to fund new innovative ventures within their divisions. And companies like Herman Miller select innovation advisory teams that consist almost entirely of outsiders with a creative and entrepreneurial bent.

There are, of course, some obvious drawbacks to ‘isolating’ innovation efforts in a separate unit. If resources are scarce, management may prefer to invest in established, profitable units even if this means starving the innovation units. The established units may see themselves as ‘competing’ with the upstart incubator, which can strain communication and collaboration. And if the incubator has to fight for the organization’s best minds with established units, resentment and protectionism can arise. One solution to the fight for talent may be to publicize the new incubator internally and allow people to volunteer (self-select) for the new unit.

Christensen’s research suggests that separation and autonomy (but not isolation) of an organization’s innovation efforts significantly increases the likelihood that they will be successful. In Meeting the Challenge of Disruptive Change (requires fee for online access) he says:

When innovation requires a different cost structure to be profitable and competitive, or when the current size of the opportunity is insignificant relative to the growth needs of the organization, then and only then is a spinout [separate] organization required. It was not until HP decided to transfer the struggling ink-jet unit to a separate division in Vancouver, that this business finally became successful.

The primary requirement is not [separate physical space] but that the innovative project not be forced to compete with projects in the mainstream of the organization [because they] will naturally be accorded lowest priority. Managers may need to run two businesses in tandem — one whose procedures are tuned to the existing business model and another that is geared to the new business model. Only the CEO, however, can ensure that the new organization gets the required resources and is free to create processes and values that are appropriate to the new challenge.

Christensen notes that this separation is usually unnecessary when launching sustaining innovations (new, higher-margin, significantly more valuable products and services brought to an existing market, a known group of customers), but usually essential when launching disruptive innovations (new products and services that extend the market to a whole new class of customers).