Qld Govt releases detailed, comprehensive cloud-first roadmap


news The Queensland State Government has published an extremely detailed cloud computing implementation model which it will use to formally push its many departments and agencies into a cloud computing-first procurement model, as the state attempts to address its substantial issues with fundamental ICT project and service delivery.

In June 2013, the state published its first comprehensive ICT Audit. The report found that ninety percent of the Queensland Government’s ICT systems were outdated and would require replacement within five years at a total cost of $7.4 billion, as Queensland continues to grapple with the catastrophic outcome of years of “chronic underfunding” into its dilapidated ICT infrastructure.

The systemic problems are particularly troubling for the Queensland Government, coming as it does after the disastrous failure of a major ICT project at Queensland Health with respect to the replacement of its payroll systems. The project ballooned out in cost to the point where it is currently estimated that it will cost some $1.2 billion in total by the time the project is complete. The failure of the Queensland Health payroll systems upgrade left many of the state’s medical workers without pay for lengthy periods, and contributed to the public’s poor opinion of the previous Bligh Labor administration.

In the wake of the failure and the ICT Audit, Queensland ICT Minister Ian Walker announced the state would follow the US Government with several key ICT reforms, including initiating a cloud computing-first policy.

Today, Walker published an extensive, 78 page Cloud Computing Implementation Model (PDF available online), with which the state aims to aid departments and agencies in implementing cloud computing projects. The state plans to deliver significant savings, better frontline services and more opportunities for small business through the new strategy.

In a statement, Walker said the strategy gave the Queensland public service the tools to transition to a cloud-based environment, which would reduce costs by government moving away from owning and operating expensive ICT assets.

“Cloud computing dovetails with our push to create the best public service in Australia, and is part of our strong plan for a brighter future in Queensland,” Walker said. “The former Labor government was stuck in the past and it cost Queensland dearly. This is one of the most robust suite of tools to be delivered by any Australian government, helping to revitalise frontline services by giving agencies the agility to respond to changing business needs and priorities.”

“The Strategy and Implementation Model is an update to the Queensland Government’s ICT Strategy and is critical to progress our ICT as a service policy, allowing us to only pay for what we use. We’ve seen this in action through initiatives such as the recent ‘cloud ready’ Microsoft deal that will save the government $13.7 million.”

“Cloud computing is a key component of our ICT renewal agenda, which also includes the One-Stop Shop program to improve public access to government services and the highly innovative Open Data initiative.”

Walker said the cloud approach would cut the cost of owning and maintaining ICT assets, keep government systems at the forefront of technological advances, allow a higher level of responsiveness and greater mobility, keep information secure and foster new approaches to government business.

“A rigorous governance framework around this strategy gives Queenslanders confidence that tax payers’ money is being spent prudently while ensuring the effectiveness of cloud services,” he said.

Queensland Government chief information officer Andrew Mills, said this was an excellent first step in leveraging the benefits of cloud computing.

“Our focus will now be to support agencies in the implementation of cloud solutions to meet their business needs,” Mills said. “The strategy update and implementation model were completed after extensive industry and government consultation and was a key recommendation of the Independent Commission of Audit and Queensland Government ICT Audit.”

The actual implementation model represents one of the most detailed and complete examination of the cloud computing paradigm which has thus far been published in Australia, especially by a government entity. It appears to contain a comprehensive cloud computing implementation framework, with recommendations for agencies in areas such as getting their organisations ready for cloud adoption, setting the foundations of cloud adoption, setting cloud engagement principles, accelerating adoption of the cloud, and cloud governance.

It also contains a recommended timeline for agencies and specific details around a large host of other factors which should be considered.

I have to admit, I was all prepared to criticise Minister Walker when I first looked at the Cloud Computing Strategy document the Queensland Government issued today. It’s a document of only half a dozen pages (PDF) which doesn’t contain anything, really, beyond motherhood statements.

However, the Cloud Computing Implementation Model document which I found after poking around on the Queensland Government’s website is really something else. Detailed, comprehensive and considered, this is really a landmark document which I suspect will be very widely read within the Queensland Government and in other jurisdictions, as well as in the private sector. This is precisely the kind of roadmap agency CIOs desperately need to start justifying cloud adoption to their secretaries, and I congratulate Walker and Mills on its production.

At times I haven’t had a lot of faith in the Queensland Government over its IT project and service delivery capabilities (I can’t think why … can you think of any examples?!). However, the publication of this landmark document has done a lot to re-establish that faith. It will be interesting to see what impact it has over the next 2-3 years. In the meantime, I commend it to readers (PDF).


  1. I’ve been a strong advocate for CloudFirst ICT strategies in Australian governments for the past few years – so it is pleasing to see Queensland following this path. Both NSW and Victoria have similar positions, and the Australian government is heading in this direction as well, so it is not unprecedented. The reason why I believe that a cloud first stance is important is because it creates a bias towards sustainable innovation that has been absent from strategies that were more oriented towards cutting costs through consolidation, rationalisation, standardisation and shared services.

    The challenge that CloudFirst creates, however, is that agencies need to become intelligent consumers of shared, cloud, services. This means that they need to adopt new ways of thinking more pragmatically about benefit/cost/ risk trade-offs and new more agile ways of framing business requirements. Better to pragmatically adapt requirements to the relatively standardised functionalty of a cloud service than to perpetuate the current high cost, high risk, customised path that most agencies currently follow. This is a major culture change for both agency executives and IT professionals – new mindsets and skills are required.

    Developing the new skills will take time, and is best done via a process of compounding organisational learning. Many small projects that create learning over time. CloudFirst is a good stance because it obliges agencies to at least consider a cloud service for any new or renewed workload or application … so it is creating a stimulus for innovation and fresh thinking. This is also important as a signal to the market to invest in onshore cloud services capabilities in anticipation of demand from government. Many small cloud services projects that result from agencies simply choosing better, faster, less costly and less risky solutions to meet their own ICT needs will add up over time to a major change in the culture – and performance – of government IT.

    I’m not so confident, however, about Queensland’s implementation model. The document is an excellent blueprint, but the danger inherent in its approach is the risk of perpetuating failed “Daddy know’s best” central agency leadership. This need not happen, but unfortunately I suspect that it will.

    The implementation model is very much based on two flawed assumptions in my view: (1) that a strong central strategy/architecture/brokering role is necessary to manage the “100 flower” risks of agency-by-agency ad hoc cloud services procurement. (2) That the model can be successfully executed without creating an unintended, and costly, central bottleneck for cloud services adoption in agencies.

    Clearly there are downsides to an entirely fragmented and decentralised approach. The issue is that such a rigorous and prescriptive implementation model also creates potential risks and costs. The risk is that the centralised whole-of-government activity creates bottlenecks, inefficiencies, overhead costs and unintended errors etc. that actually slow down cloud services adoption by agencies. Queensland has made choices that a strong central lead is required for cloud services adoption … but what evidence is there that they have the capacity to execute on this implementation model?

    My sense is that the balance is better struck towards accelerating cloud services adoption in agencies within ‘light touch’ leadership facilitation and collaboration arrangements that encourage peer-to-peer sharing of lessons learned, solutions that work well, solutions that don’t etc.

    Cloud stores, for example, seem like a good idea … but aren’t necessarily so vs. just relying on the open competitive market and encouraging the sharing of information between buyers. Think of using Choice magazine or social media shopping recommendations vs. buying only from a particular trusted shop. The problem with cloud stores is that they imply that the central agency will create an effective assurance framework around the stock on the shelves. If assurance is high/in-depth, then it takes time and effort … and is costly and creates a bottleneck. If assurance is low/cursory, then what value does it add? Cloud stores are a two edged sword. My understanding is that the US government’s GSA abandoned its cloud store for this reason – it became a costly non-value-adding bottleneck … agencies were better off just buying from the market within the FEDRAMP assurance framework.

    I think we need to use the new digital technologies and social media capabilities to accelerate cloud services adoption – so the approach should be primarily founded on creating a minimum set of online enablers for decentralised agency-by-agency activity. This would comprise:

    1. Proactive Thought Leadership – engaging with line-of-business agencies about the new ‘art of the possible’ created by as-a-service models and the need for more agile thinking about business requirements and benefit/cost/risk tradeoffs.

    2. Creating an enabling cloud first policy that is refined regularly based on cloud services adoption behaviors.

    3. Providing clear and pragmatic guidance on key compliance requirements for contracting with cloud service providers and for managing security, privacy, record keeping and audit obligations etc. This guidance should be focused on enabling achievement of line-of-business outcomes, not just managing a narrow concept of ‘risk’.

    4. A transparent cloud services catalogue – a central shared peer-to-peer repository of information about cloud services already in use within agencies. This serves the purpose of establishing a central ‘clearing house’ of information about ‘who is using what services’ – help to reduce unwitting duplication of categories of services that would be better reused. If another agency is already using a service that you are thinking of buying why not talk to them about it, see if it is any good and reuse some of their procurement artifacts if possible … maybe even buy off their contract? The catalogue can create a virtuous cycle of Buy -> Share -> Collaborate -> Reuse behaviors that will give better assurance of quality than the rather theoretical assurance that is possible in a Cloud Store. Ex ante assurance in a cloud store is based on a theoretical purchase, where ex post assurance in a cloud services catalogue by agencies is based on actual evidence of procurement and use. Think of the catalogue as like eBay. Who decides what goods are sold in eBay? Market forces … Who provides the feedback that determines the reputation of sellers? Customers.

    5. Removal of blockages – such as reform of procurement rules to allow more agile, iterative, adoption of as-a-service offerings and reform of agency funding regimes to transition from capex to opex funding models.

    6. Showcasing successes – proactively documenting and showcasing successful cloud services adoption by agencies to celebrate success, build confidence and share lessons learned.

    My view is that central agencies need razor sharp and highly critical thinking on the minimum set of activities that they can perform to create catalysts and enablers for cloud services adoption in agencies … otherwise they risk becoming part of the problem (again) rather than part of the solution.

    There is no question in my mind that cloud services procurement by agencies can and will produce better results than previous IT procurement models. Whether whole-of-government ICT strategies will speed up or slow down cloud services adoption however, is a matter of ongoing debate. First do no harm.

    The Cloud Computing Implementation Model is great in theory … but what confidence do folks have that it will be useful in practice?

  2. you aren’t wrong about not having faith… but i have to say if this gets done right and comes to fruition as hoped i will actually be able to say Queensland has done something right in the IT space – at long last. it will feel weird to say it too; the QH payroll debacle looms rather large in ones memory….

    it certainly is timely given this http://delimiter.com.au/2013/06/07/systemic-business-risk-90-of-qld-govts-ict-needs-to-be-replaced-total-cost-7-4-billion/ is just under a year old; and i certainly doubt the total spend figure identified there has come downwards any.

    • Well … hopefully they are turning the corner to a brave new world.

      I’m in the Qantas lounge in Perth waiting for a flight, and the more I read the docs the more I like them … you are right Renai in some regards … but I just hope they can pull this off in terms of resourcing and executing on the Cloud Computing Implementation Model. It just seems too ambitious to me given the track record of ‘unsuccess’ for whole-of-government approaches in Queensland (and in other state governments).

      Anyway, here is my quick summary of the key gist of the strategy for other folks interested in public sector cloud services:

      The Cloud Computing Strategy (really an ICT-as-a-service strategy) is part of a suite of documents also comprising ICT as-a-service policies, an ICT-as-a-service Decision Framework and a Cloud Computing Implementation Model.

      Key implementation areas include:
      – Cloud ready (educating and informing agencies on best practice procurement and management of cloud services);
      – Cloud computing policies and decision frameworks.
      – Cloud foundations (building blocks necessary for a holistic approach to acquisition, secure consumption and management of a multi-provider cloud ecosystem).
      – Cloud engagement: (trusted advisers to assist agencies with cloud transition and migration plans).
      – Cloud accelerate (pre-qualified procurement arrangements for common and commodity ICT cloud services); Whole-of-government panel arrangements (for cloud email and IaaS).
      – Cloud governance (to ensure alignment with a ‘cloud-first enterprise’ vision).

      (all good stuff actually)

      The Cloud Computing Implementation Model describes a comprehensive top-down enterprise architecture oriented approach comprising:
      – Establishment of a whole-of-government cloud marketplace and storefront.
      – Establishment of commercial arrangements for external cloud brokers of cloud services to support agencies.
      – Establishment of a private community cloud to support common government requirements and workloads unsuitable for public cloud.
      – Establishment of a whole-of-government identity federation platform to support the secure sharing of common ICT services across government.
      – Transformation of agency ICT divisions from a service provider role responsible for building and managing ICT assets and systems to acting as a trusted broker of ICT services from external suppliers.

      This part though is a tough gig … I wish them well. I also (anyway) hope that agencies also just get on with buying and implementing cloud services as and when they make sense to them individually.

      Don’t wait for Dad to tell you what to do kids, just use your common sense and get on with it … make sure you chat to your siblings and friends and support eachother. Stay safe!

  3. I would really hope that this plan doesn’t automatically suggest moving workloads to public clouds first. Such as AWS … Seriously do the math IT’S NOT CHEAPER !

    Hopefully the plan is to consume from a central private cloud (still within the government’s control) with well documented SLAs and performance chariteristcs matching those Service Levels.

    Still within the fallout of the IBM QH project I still shake my head when the Premier said they would not invest in IBM and then TMR spend $4m replacing an IBM Mainframe ! Hardly seems a good use of tax payers money ..

    • Hey nomore,

      The cloud services model (even IaaS) is not necessarily (and not only) about reducing costs is it? We need to (a) look at this in terms of long run sustainable TCO and (b) look at this in terms of providing a catalyst for innovation in areas where cloud services are better, faster, less costly and less risky that other ICT sourcing models.

      Can you share more about your “do the maths” comment? Have you done a detailed analysis and concluded that AWS is actually more expensive than a fully loaded TCO of a comparable on-premise service? This would be interesting to share …

      Most analysis comparing the fully loaded TCO of traditional on-premise or shared services hosting to an enterprise-grade IaaS offering that I’m aware of shows the opposite … by a substantial margin … even in full-on public tender situations.


  4. ‘Cloud service brokerage platforms…
    ‘Agency ICT divisions to transition … to a service broker role…’
    ‘Cloud brokers appointed for SaaS, PaaS and IaaS services (sic) and engaged to orchestrate other providers…’

    In all, ‘broker’/brokerage’ is mentioned 191 times in the implementation model, but nowhere is the word ‘broker’ defined.

    Looks like a lot of people will be employed for a long time ‘brokering’ something. At least all this brokering of stuff will keep some Queenslanders employed for a while.

    Methinks we would be better off putting all these brokers onto our own version of the B Ark, ala’ the Hitchhiker’s Guide to the Galaxy. (http://www.bbc.co.uk/programmes/b03v379k/profiles/golgafrinchans )

    • Hey BK … hmmm … brokers and rubber ducks are similar in that one never need be alone when they are around (I’m with the Captain on this point) but ducks are a much better solution for the bath use case as brokers are a less convenient size, tend to become waterlogged and then start to smell if they are not dried properly afterwards …


  5. Renai

    If you are going to continue label the 5000+ good people who work in Queensland Government ICT with the stain of one ‘so called’ failed ICT project I think you should find something a little more timely than the Health Payroll.

    The Health Payroll was started in 2006 and implemented (very poorly) in 2009. It’s a 5 year old example at best. And one where the in-house (Queensland public service) delivered technology significantly out performed the outsourced (IBM) program management.


    • Hi Peter, Not sure your continued defending of those “good people” in ICT positions that remain in QLD government is helping (as many might simply see such comments as to do with defending one’s own legacy – rightly or wrongly).

      Furthermore,such advocacy also actually neglects the issue that those most afflicted by the stigma of association (of failed government projects) are in fact those that often have the least support; those that no longer work in QLD Government ICT and have had to leave QLD to find work or no longer work in the ICT sector at all – Indeed, there are many that now find themselves un(der)employed on the back of a long career in the QLD public sector ICT environment.

      The QH Payroll does rate a special mention because it’s been given special mention – it was used by the incoming Newman Government as a contributing reason for its cuts to the public sector.

      Furthermore, QLD government project failures have been more wide spread than simply QH Payroll – and that’s been ongoing and systemic. I could well imagine that it will take some considerable time and a series of so called “successes” to change such perceptions. I wish those who remain in the QLD ICT environment well and I also wish you well in your new endeavours.

      • Hello 14000

        That a thoughtful and considered response of yours. And I appreciate your feedback and your good wishes. Let me sharpen my comments in line with your points.

        (1) I know I run the risk of self promotion. Although that is never my motivation. I’ve had a long career. People who know me are not influenced one way or the other by what they might read here. That said, I do feel some ongoing responsibility to support people who I know are doing a good job in more than trying circumstances. It’s hard not to jump in when I see lazy and unreasonable negative comments made.

        (2) Yes the Health Payroll was political. A lot of these so-called failed projects in QLD have been magnified for blatant political purposes. The $1.2 Billion figure on the Health payroll for example incorporates the development costs, the running costs and the replacement costs. It’s just simply wrong to say it was a $1.2 Billion project using metrics like this. We don’t measure ANYTHING else that way. But it was a good whipping boy and everyone just jumped in and believed the figures. The Press never asked how they were calculated. A very disappointing outcome.

        (3) The ICT Audit conducted in QLD in 2012 showed the government runs 1100 concurrent projects with varying amounts of ICT components (business process change, infrastructure etc). The vast majority of these projects were delivered within acceptable tolerances. (One’s that were fairly tight). If you were a project manager or a member of one of these project teams you would get pretty tired of people using a 5 years old example (a project you weren’t even on) to sell the idea you were no good!

        (4) I worked with other jurisdictional CIOs around the country. All of them have ageing ICT assets – just like Queensland. And all of them are facing alternatives for replacement that balance as-a-servcie delivery with in-house replacement. The industry as a whole is in for a big shake up. If we have all done one thing wrong it’s this: We haven’t done a good job preparing our workforce for the step change in ICT business models that is now on the door step. We should all be doing more.

        (5) For the people who lost their jobs in ICT in QLD (and I’m probably one of them) we face an industry wide credibility problem. Governments of every side do not see information systems are a core competency. Here’s an example: The Rudd government implemented 15% cuts to ICT across government at the same time it ran the Pick Batts program. In their eyes Pink Batts were a more important to the economy than investments in information systems. QLD was no different. People were let go because their discipline wasn’t valued. There is both a human cost and a social cost if we don’t value the ICT discipline. In my view our discipline is critical to the economic success of the country. But if people keep saying we are somehow incompetent then we will keep being undervalued. No one wins then!

        14000 I also wish you well. If you worked in ICT and have left the government you will bring with you some precious skills: the ability to manage complexity, the ability to manage scale and the ability to work with layers of abstraction. Very few people outside ICT have these capabilities honed into their being. In many many industries you are way more valuable that you know.

        In closing, you know who I am so come up to me one day and say 14000. I’ll know what you mean. We can continue the chat – over a beer maybe.


  6. Peter,

    Sorry for the delayed reply – have been away in regional Australia (without mobile or internet coverage) – I am a Telstra customer! :( and its 2014!!!!) Anyway,thank you for your reply – I am grateful my comments engendered a further reply and detailed response. I wrote and rewrote my comment several times trying to find the appropriate tone. I apologise if I didn’t succeed and was not seeking to be personal at all. Yes. I agree with many of your subsequent points. 1 of the 14,000

Comments are closed.