‘Cloud first’ a circuit-breaker, says Ovum


news Taking a “cloud-first” policy has the potential to act as game changer to allow departments and agencies to break out of their current restrictive ICT procurement practices, technology analyst firm Ovum said this week, as discussion continues to swirl about how Australian governments are handling the new cloud computing paradigm.

Last month, Ovum Asia-Pacific research director Steve Hodgkinson published a provocative analysis arguing that the Federal Government had taken a “cloud-last” position on the adoption of the new generation of cloud computing technologies, with the analyst arguing that this approach lacked a clear vision of the benefits of the cloud computing model while being very clear about its risks.

Jurisdictions such as the United States and the United Kingdom have taken what they have described as “cloud first” positions, which means they will be default seek to procure technology as a service rather than implementing their own platforms. And New Zealand has also recently done the same.

Hodgkinson said because of Australia’s more minimalist approach, there had been much lesser adoption of cloud computing services by Federal Government agencies. “Agencies still regard cloud services as unproven and risky,” he wrote. “Consequently, agencies are deferring productivity and innovation benefits, and the ICT industry is frustrated that the government is slow to buy cloud services and is not, therefore, stimulating the development of robust onshore Australian cloud service capabilities. New Zealand’s stance, on the other hand, directly challenges the status quo, asking: “Why not cloud?” This is a much more direct catalyst for innovation.”

The analyst’s comments provoked substantial debate on the issue, with John Sheridan, the first assistant secretary of the Australian Government Information Management Office’s Agency Services Division, arguing that the Government actually took a “value first” approach, and others highlighting the fact that government regulations often require data to be stored in Australia – a fact which has the potential to bedevil cloud computing deployments, given that companies such as Microsoft, Salesforce.com, Amazon and Google do not operate datacentre facilities in Australia.

In a follow-up statement this week, Hodgkinson acknowledged these issues, and wrote that in truth, Australia’s stance was perhaps better characterised as “cloud when we, and it, are ready”, rather than “cloud last”. On the face of it, the analyst wrote, this might seem sensible enough, as concerns had been raised that a “cloud-first” stance might actually be counter-productive for procurement processes, leading to overly simplistic thinking and activity.

This seems sensible enough, from the point of view of incremental procurement process reform. However, it still begs the question: Does “cloud when we, and it, are ready” show sufficient leadership of ICT reform. “If cloud services are seen as a silver bullet, the result will be disappointment resulting from unrealistic expectations and a burst of poorly planned, insecure, non-integrated initiatives driven by vendor hype,” he wrote.

However, Hodgkinson noted, there was evidence that suggested that despite these factors, adopting a cloud-first strategy could actually work strongly towards the interests of departments and agencies. For example, Hodgkinson highlighted evidence provided by the US Government Accountability Office (GAO), in a report published in June 2012 which assessed the results of the 2010 “Cloud First” strategy in seven US government agencies.

The GAO report found that agencies had successfully incorporated cloud solutions into the IT and investment management policies and processes, and had implemented one or more services in a cloud environment by December 2011, with all seven agencies planning to achieve the target of three services in the cloud by the end of 2013.

“Overall, the GAO’s scorecard was quite positive about the implementation of its strategy,” wrote Hodgkinson. “It appeared that the strategy was effective in stimulating a focused burst of organisational learning that had accelerated innovation in ICT procurement and project management. Certainly, the US government, and also the US ICT industry, would not be as far ahead in its “cloud thinking” and cloud capabilities without the strategy’s stimulation of hands-on activity and demand for cloud solutions.”

Because of this kind of evidence, Hodgkinson wrote that Ovum’s view that a pro-cloud services policy was a positive step for many governments, even though it could seem a bit odd to bias procurement towards one particular sourcing model. Then-US Government chief information officer Vivek Kundra, Hodgkinson wrote, had used the ‘cloud first’ strategy to “kick the hornet’s nest” of government ICT procurement from 2010, shaking departments and agencies out of their existing established structures.

“In essence, the value of “cloud first” is that it authorizes, and even requires, agency executives to look beyond the constraints of the current approach to ICT procurement and to ask the question: “Is there a better, more agile way to use ICT to achieve our policy and service delivery outcomes?” wrote Hodgkinson. “The value of “cloud first,” however, is only apparent if one first opens one’s eyes to the reality of the challenges in the status quo, and the fact that we really do need to accelerate the introduction of a fresh approach to ICT in government.”

In Australia, there is substantial evidence that existing ICT procurement models are failing, especially at the State Government level.

In November last year, for example, Victoria’s Ombudsman handed down one of the most damning assessments of public sector IT project governance in Australia’s history, noting total cost over-runs of $1.44 billion, extensive delays and a general failure to actually deliver on stated aims in 10 major IT projects carried out by the state over the past half-decade. The report was undertaken over the preceding year by Ombudsman George Brouwer as a self-initiated action of his office, after a string of minor reports published by the Ombudsman and the state’s Auditor-General Des Pearson over the past few years had raised significant concerns with Victoria’s handling of IT project governance.

Similarly, state governments all around Australia currently appear to be experiencing a systemic inability to deliver major IT projects, and in some cases, basic IT services. Queensland, for example, has revealed that its botched payroll systems implementation at Queensland Health has already cost $417 million and will need some $837 million to fix over the next five years. Currently, the platform requires just over 1,000 staff performing some 200,000 manual processes on an average of 92,000 forms to make sure Queensland Health’s staff are paid the approximately $250 million they are owed each fortnight.

Queensland has also broadly abandoned its IT shared services model, as has Western Australia, and in New South Wales in mid-August, a landmark report into the management of the NSW Public Sector commissioned by the state’s new Coalition Government has described how dozens of overlapping and competing systems and services providers have created “chaos” when it comes to the state’s current IT shared services paradigm.

At least one state, NSW, has turned to the cloud in an effort to resolve some of its issues. In late July this year, German software giant SAP won a substantial deal with the NSW Government’s Trade & Investment agency which it described as its biggest deployment of its Business ByDesign software as a service suite globally, and its first cloud platform win in the local public sector. Queensland’s new IT Minister Ros Bates has also publicly discussed turning to the cloud as an alternative to help resolve the state’s problems with fundamental IT infrastructure such as providing email services to staff.

I agree with what Hodgkinson is saying here. Those who are concerned that a so-called ‘cloud-first’ policy could lead to silly procurement decisions where cloud computing is considered in situations where it wouldn’t be appropriate or even the best option are right that there is a risk that this could happen.

But the far greater issue in government procurement circles right now is that ICT procurement is just not functioning at all, at least when it comes to the majority of Australian state governments. We’ve seen failed project after failed project in this arena over the past five years – enough failed projects that the situation can be described as a systemic failure to deliver IT projects and to a certain extent, even basic IT services.

This situation represents the perfect storm for cloud computing. I’m not saying that all government ICT procurement should become focused around the cloud, and neither is Hodgkinson. However, what we’re both arguing is that the onset of this new paradigm in the IT industry is a golden opportunity to open up the box of government ICT procurement a little and start to look for solutions to the ongoing failures – solutions that are not the traditional ones. You can’t resolve failures by continuing to follow the same path you were on. In this sense, cloud computing represents a ‘circuit-breaker’ which could be used to create different conversations than the ones being had at the moment.


  1. Do existing agency CIOs (particularly State Governments) really want to open the box on IT procurement? I suspect not. So the driver for change will have to come from else where.

    With all due respect to John Sheridan and the arguement of a “value first” approach – such a comment is entirely predictable yet goes to the heart of the conversation – lots of language not so much reality – what is value, who defines it, and who receives it!!!!!!!

  2. It’s a bit of a confusing article.

    Cloud IT services aren’t much different to a shared services model. If your provider sucks, then you’ll be in trouble – although at least I guess you have the option to choose a provider who sucks least – as far as you know anyway.

    I can’t see how Cloud IT services fixes up or works around poor project management, sound knowledge of the business and technical implementation. You can still make a pigs arse of an implementation of pay roll for topical example “in the cloud” as you can with your own metal or managed services.

    I suspect a lot of cloud providers aren’t happy with the archaic and slow government procurement process – as they are typically burning money and will do so for at least 2-3 years.

    • Hey Thateus,

      My view is that enterprise-grade cloud services are, indeed, something new and different … in that they are shared services that actually do what they say on the label. They have been architected from the ground up to serve the needs of diverse customers globally. The fact that they (a) work as promised and (b) are too arms-length to be interfered with is a game changer for government agencies. The choices are simple. Does it do something that you need? Can you get your head around how to buy and use it? OK then … get going. The experiences of early adopters is that this proposition does actually work for selected applications and services … which makes it an important part of the solution mix when the experiences are also that in-house IT, in-house shared services and outsourcing models are proving in too many cases to be unsustainable.

      As budget pressures bite we will have increasing need to understand how to tap into shared services that actually work … and my bet is that enterprise-grade cloud services will be at the core of the way we will do this.

Comments are closed.