Yesterday my colleague and friend John Kost and I challenged each other on stage at a public-sector-only session at Gartner Symposium about cloud and consolidation in government, and whether they are conflicting or complementary.
He talked about his experience in Michigan in the nineties, as well as other infrastructure, application and process consolidation initiatives around the world, about their benefits, constraints and the critical role of governance to make them succeed.
I stressed how cloud computing is – and will increasingly be – a choice made by the business and not by IT, as the former gets increasingly challenged with change and frustrated by the inability of the latter to respond timely and efficiently enough.
Of course we both knew that none of the positions is right in all cases, and one size does not fit all. There is room for both, and in both cases the governance process around them is the most critical factor.
What was more interesting were the reactions from members of the audience after the session. Most people, I believe, enjoyed the debate, but some felt that there was not enough time to drill down on some of the most important issues that play a critical role in deciding either way.
A few attendees highlighted that the problem is not consolidating infrastructure in one jurisdiction (e.g. state), but looking at opportunities to consolidate across jurisdictions. My reaction is that the same dynamics apply. Whereas consolidation across jurisdictions can be a way to make costs more affordable, that is not necessarily the most sustainable option, since service providers mature their cloud offerings and give choice to individual jurisdictions as well to agencies therein.
One attendee observed that in a politically-driven environment there are few incentives for sustainability. The CIO and the IT organization aim at meeting the political objectives of the incumbent administration, and additional investments required to make solutions sustainable beyond the political timeframe of that administration are unlikely to be given a high enough priority, especially if the costs of those investment stay with the current administration and the reward stays with the next one. My response is that this is one area where being a smart government is particularly important: rather than embarking in large-scale, long-term initiatives that are susceptible to the above risks, governments should pursue a more evolutionary approach, aimed at delivering short-term gains while putting the administration on a sustainable path to transformation. This would call for fewer ambitious big-bang consolidation initiatives than some that we commented about during the session.
Another attendee said that in many areas cloud is not ready for prime time, especially from a security perspective. While this may be true today – at least for certain types of workloads or data – the progression toward increasing trust and security is inevitable as the market evolves, and one needs to be smart to embed the ability to take emerging cloud opportunities into account when they become available.
The bottom line, as I suggested on stage, is that consolidation and shared services initiatives might fail because of their inability to incorporate change. As soon as the business case for a major consolidation initiative is fleshed out and approved, and the required cross-agency governance is in place, shared service organizations inevitably focus head down on achieving their intended objectives. While this is understandable, given their challenging task, they may end up losing sight of how market conditions evolve, and may come too late (or never) to the realization that some of the services in their original catalogue should be discontinued, because they can be more efficiently and effectively procured from the market by individual agencies.
Therefore the ability to pull the plug on what they do at the right moment in time is as important as providing the best possible services.