Blog post

Cloud Computing May Become The Worst Enemy of Centralization

By Andrea Di Maio | August 05, 2009 | 1 Comment


I am at the Gartner Symposium in Capetown, which is always a very interesting experience. Of course cloud computing featured a lot during the event, starting with the analyst keynote where my colleague Daryl Plummer said, among other things, that cloud computing is the next step after consolidation and centralization.

While this is both sensible and rational, I do not believe it will necessarily work this way in government. As I wrote in a research note some time ago ( “Shared Services in Government: Obscured By the Cloud?” – subscription required), commoditized, cloud-based solutions to infrastructure or application needs may offer individual departments or agencies the way out from long and complex centralization initiatives.

I had a conversation with a client here in South Africa about the large IFMS (Integrated Financial Management System) initiative, which aims to integrate and migrate government finance, HR, asset management, logistics and other line of business solutions, into a single distributed transversal system. Vendors have been selected for two modules, and the State Information Technology Agency is responsible for integration and custom-developed modules. This is a remarkably ambitious project that will ultimately provide a single system to federal as well as provincial departments: I am not aware of anything of comparable scale addressing two tiers of government.
As far as I understand, departments are somewhat mandated to use this system, so apparently there is no opt-out possible. On the other hand, integration and full deployment are going to take almost a decade: over such a long period of time, it is not unlikely that emerging cloud-based alternatives become attractive to provinces or federal departments that need to replace their legacy systems but are still waiting for IFMS to be available. Therefore, even where there is no opt-out, centralized solutions will be challenged by lower-cost, more-flexible alternatives over time.

It is even more so in cases where departments and agencies join on a voluntary basis. I wonder whether the “storefront” approach taken by the U.S. GSA (and being considered in the UK too) is a better way to go, assuming that the best role of a central government agency is to become a broker for emerging cloud-based offerings. This is not free from challenges though, such as the definition and enforcement of enterprise standards and vendor selection criteria that allow such a storefront to evolve its offering over time, while maintaining consistency across different solutions.

Many paths will lead to cloud computing in government, but quite a few might not be what whole-of-government strategic planners and architects expect.

Comments are closed

1 Comment

  • Gijs Key says:

    interesting concept. allthough I’m no fan of government use of the commercial cloud, much can be said for a govt cloud. Isn’t the root of the internet based on the government desire to eliminate risk by distributing network traffic? One could imagine a similar approach for leveraging risk by distributing computing power…