In the past few years most Western Governments have put in place a “Cloud First Policy.” A piece of paper that exhorts government agencies to look at Cloud services before traditional purchasing methods. But when you look at New Zealand, the United States, Australia, and the United Kingdom, the uptake is very low and slow. Where the rest of the world is moving carefully to Cloud, government is being left behind, and it will cost us.
In 2011 the then US Chief Information Officer Vivek Kundra instituted the Federal Cloud Computing Strategy, which established a Cloud First policy for government. Three years on the progress to Cloud is extremely slow and problematic, according to an Accenture report that surveyed 286 government technology leaders. The results were interesting:
- Of the twenty cloud migration plans submitted to the Government Accountability Office for approval in 2012, only one has been completed. Eleven of the projects failed to report performance metrics and seven did not include thorough enough cost estimates.
- Less than half the manager’s were familiar with their agency’s Cloud strategy, which is not unusual, given that only 30 percent of respondents claim to be implementing a cloud strategy.
- Only four percent of agencies are actually building Cloud, and the majority of that is private.
- Over two-thirds of respondents don’t believe they have the skills to deploy Cloud.
- A third of agency’s believe they needed at least one more FTE to assist in Cloud.
- A majority didn’t have enough budget to get on with Cloud.
In the United Kingdom, the much publicised G-Cloud programme has made some interesting in-roads if you believe the reports, in particular, the number of small ICT businesses being able to get access to government contracts has increased. But here’s the rub. Of the (up to) twenty one billion pounds government spends on ICT each year, less than half a percent is bought through G-Cloud. The G-Cloud is generally considered to be the most advanced model of its kind in Government.
Australia came very late to the party not declaring a “Cloud First” policy until 2013 after a report slamming agency’s for low uptake. Now, they are caught in a political mess of epic proportions with all future moves held up until audits and trials are completed and its so bad, they can’t even agree on basic terminology. They are going nowhere in a hurry.
In New Zealand, we see that government agencies are taking up Cloud, but there is no measurement as to volume, success, cost savings, benefits, types of Cloud, or any other useful data, and when we are talking Cloud here we are talking more than the Common Capability services, we are talking actual Cloud services.
Anecdotally, I see that Software as a Service (SaaS) use in Government departments in New Zealand is quite high. Certainly every agency that I have talked to is using some kind of SaaS service, from small departmental applications to much larger implementations.
As an interesting aside, Malaysia, China, India, and other large Asian nations are far more closer to Cloud with heavily mandated and subsidised Cloud based services. India is in the process of rationalising all of its datacentres into regional sites before moving them to Cloud. China is spending hundreds of millions on government sponsored Cloud facilities.
So why when the rest of the world is slowly making its way too Cloud, is government failing to keep up?
A lack of money
Globally, the financial crisis has had a massive impact on the amount of cash that is available. In particular, governments are required to be seen to be controlling the purse strings (austerity) and have done so without thinking too much about the consequences. Moving to Cloud, whether state controlled or private, requires a lot of investment. It is a fallacy to think that transition to Cloud costs nothing and that your operating costs will drop. Your capital costs may drop, but your operating costs will increase. In a period when government still has downward pressure on its agencies, this presents a major problem.
A lack of resource
There is a definite lack of resource. Cloud professionals are hard to come by. It is no surprise that in almost all ICT employment surveys they hot skills for 2014 are Cloud, Business Analysis, and Mobile App Developers. All three go hand in hand and all three are in very short supply.
The messages are confusing. It is not clear what rules, if any, there are around data sovereignty. There is no central risk management in general. Some governments mandate and some do not. Mandate does not always mean you must. Agencies worldwide struggle to find out what the rules are. Privacy laws are not clear.
There are other issues
There are bigger issues that Cloud that are taking up agencies time. A lack of money, reducing headcount, and ageing legacy applications are far greater problems than trying to figure out how to move a chunk of disc to a Cloud provider. In fact, until those problems are solved, services can’t be moved to Cloud.
Cloud is not the whole answer
Cloud is only part of the puzzle. It needs to be considered as part of a strategy not as the whole strategy. Agencies have to ensure that all ICT services are designed to service what the business requires and fit in with existing processes and systems architecture. That is an incredibly complex dance and putting Cloud at the head of it makes it more complex.
Cloud readiness is overlooked
Cloud readiness is overlooked. A Cloud strategy is required, as part of an agency’s wider strategy and infrastructure must be in place, service management must be appropriate, changes to ICT organisations are likely required, and a host of other areas need to be accounted for, before you can transition.
Older business models
Government is necessarily conservative, but that does not mean they need to be slow. The lifecycle of a product inside of a government agency can be years, where as we see Cloud’s product lifecycle measured in months if not weeks. This means that by the time an agency has analysed a Cloud service, that service can have substantively changed. By the time all the approvals have gone through, what was originally approved, is no longer available.
Government also has a love with contracts. Armies of lawyers and commercial experts ensure that all services have some kind of contract. True Cloud doesn’t do contracts. A lot of agencies get stuck between wanting the security of a contract and the fact that the future doesn’t have those contracts available.
Government must recognise some key truths before adoption of Cloud increases:
- It costs money. Not only to transition, but too operate, and those costs will increase over time as agencies ICT needs grow.
- It takes transition planning. It is not able to be switched on instantly, Cloud readiness is required. That means more money.
- It needs resource. Whether agencies train their own staff or contract in specialists, it takes skill. That means more money.
- Central government messaging can be unclear. What are the rules around data sovereignty and risk for example?
- There are other issues that may be more pressing. Investment in applications, as an example, has been very low of the last decade. They require investment before they can be moved to Cloud. More money.
- It changes business models. Service Management must be introduced and the product life cycle time reduced to months and weeks.