One of the key concepts of the component-based business or service-based business is that you can change the nature of the coupling or coordination between enterprise units. So we are seeing significant quantities of demerger, unbundling, outsourcing and so on.
Much of this unbundling is being pushed by regulators, especially in utilities (including telecom). See my previous note on Local Loop Unbundling. Think also of the structural changes that various regulators have threatened Microsoft with.
These are often formulated in terms of unbundling services, products or platforms. But they have obvious implications for organizational structure as well. (At one time it looked possible that Microsoft might be split into separate organizations, just as Standard Oil and ATT were previously broken up. And many years before that, IBM was also threatened.)
But the regulatory agenda should not make us think there is always a going to be a conflict of interest here. (Someone in a large telecom company once told me that the unbundling forced upon the company by the regulator had actually turned out to be beneficial for the company as well, because it had increased internal management visibility and innovation.) So sometimes unbundling is the right thing all round.
SAP has been making moves to support unbundling, and the SAP Roadmap includes guidance for unbundling (ECRM Guide, Feb 2005). There is also some standard IT support for inter-company collaboration and data exchange (SAP Press Release, Feb 2005).
John Hagel III and John Seely Brown have been writing about these issues for ages. John Hagel and Marc Singer had an article in Harvard Business Review called Unbundling in the Corporation (1999) (abstract only, purchase reprint). Hagel has been promoting a standard demerger pattern, which splits the enterprise into three standard chunks: infrastructure management, customer relationship or product innovation and commercialization. (This pattern has evolved in his writing, and his terminology has shifted somewhat.) In a recent blog posting (March 2005), he interprets recent developments at 7-Eleven as a successful implementation of this idea.
Hagel points out that these three activities operate with a different business logic on different timescales. Referring to Disney, HP and Morgan (April 2005), he comments: "Product innovation businesses and customer relationship businesses have completely different economics, skill set requirements and cultures. These are incompatible business types that cannot co-exist successfully within a single corporation. By trying to straddle both, these companies found that they could not be excellent in either."
Hagel has identified some of the relevant factors for determining the best way of unbundling the business - but this is not the whole story.
There are two contrasting approaches for unbundling. The approach favoured by Hagel assumes we can establish broad patterns and practices of demerger and subsequent collaboration. This has some advantages - among other things, it suggests that there will be multiple organizations offering broadly similar interfaces, yielding commercial flexibility. But we have some doubts about the generality of this approach. The alternative is to carry out specific analysis for each particular situation. More effort, but sometimes justified.