The more systems are integrated, the higher the potential that a mistake (involuntary or purposefully introduced) can propagate trough the entire system and pollute its overall accuracy. Hence, integrated systems are intrinsically fragile.
Promoters of integrated systems fight this issue by adding a number of filters to the entry of information in the system, de facto decreasing significantly the system flexibility and the possibility to adapt to changing needs.
John Hagel mentions that it is the same with organizations: the bigger and more monolithic the organization, the more fragile it can be. And very large, integrated organizations and systems are an impediment to change and adaptation.
The solution is hence modularity and creation of small, understandable, and interconnected entities or systems. The key lies in devising standard interfaces to deal with the interfaces. This obviously has an additional cost compared to integrated systems and organizations (it is less efficient), but on the medium term, the adaptation capability should produce better and more resilient results (thus, a higher effectiveness)
When will we stop thinking that efficiency, integration, standardization is better? Do we need some more meltdowns to prove that effectiveness, modularity and diversification is much more desirable?
Hat tip to ParisTech review, papers on the future of the organization (in French).
