news Queensland’s new LNP State Government late last week revealed plans to dump the troubled colossal whole of government email project begun under the previous Bligh Labor administration, with IT Minister Ros Bates highlighting the possibility to shift to a “cloud-based solution” instead.
The project, dubbed the Identity, Directory and Email Services (IDES) program, was begun in 2007 with a budget of $252 million and was slated to consolidate some 80,000 state government email accounts into one overarching platform based on Microsoft Exchange 2007, by December 2009. It was being undertaken by state-owned ICT services group CITEC, and was also to see some 350,000 identities managed, likely through Microsoft’s Active Directory software.
In the process, numerous installations of IBM’s Lotus Notes/Domino and Novell’s Groupwise platforms were to be rationalised. The key vendors at the heart of the project were Data#3, which was hired to conduct content filtering and email archiving operations, using Symantec software, along with Dimension Data, which was to conduct the migration.
However, in December last year, the Courier Mail newspaper revealed the state had spent spent $46 million on its whole of government email platform, but had only actually migrated some 2,000 accounts. At the time, Simon Finn, then-Information and Communication Technology Minister reportedly said that the project ws within budget, and that the number of users, which were rising every week, was slated to reach 53,000 by the end of 2013. By June this year, only 3100 users were using the platform, according to the new State Government.
This report came after Queensland’s Auditor-General had already heavily critised IDES in June 2010, noting that delays had been experienced, although the project as a whole was under budget. The key problem with IDES appears to have been that there was “no program board with adequate stakeholder representation to drive the program forward and to deliver the outcomes and benefits,” according to the report. CITEC was the only body involved in decisions about the program’s schedule.
Last week, new LNP IT Minister Bates said in a media release that the project was to be scrapped and never should have been approved in the first place. “The central problem with IDES was that it never represented good value for money and Departments could source alternative products,” Bates said. “The Auditor General termed IDES a financial and functional disaster, in large part because there was no effective oversight of its implementation.”
A Departmental review of IDES recently concluded that a further investment of $25 million over the next three years would be required to make it viable, Bates’ statement said. The program will now be closed by 30 June 2012. “I have recommended that we stop IDES and explore other options including a potential cloud-based whole-of-Government email solution,” Bates added, noting that a cloud-based solution “could provide benefits for the Government’s short-term cash flow position and could also provide a sustainable long-term option from a technological perspective”.
“A cloud-based solution doesn’t require a large capital investment and provides an effective way to manage a commodity-based information technology service,” she said. Bates said existing IDES users would be supported until a replacement was established.
opinion/analysis
Queensland has done the right thing here, in my opinion. Mainstream enterprise IT strategy currently usually holds that the provision of email to users is a service that is highly commoditised in 2012 and can usually be outsourced to a more capable external organisation and delivered as a service. It doesn’t make sense for large organisations to be running their own email platforms, when external companies can do it more efficiently.
If I was advising the Queensland Government, I would recommend that it investigate getting a major IT outsourcer such as Fujitsu or CSC to build a large version of the latest version of Exchange for it, and host that implementation in the outsourcer’s own datacentre, with the entire platform to be provided as a service. This kind of stuff isn’t rocket science, and there’s a number of major companies operating locally who are very well versed in these kind of implementations now.
The identity management portion of this is a little bit more tricky, however, and I note that Bates hasn’t addressed it in her media release. I would envisage that identity management is currently too difficult a project for the Queensland State Government to tackle on a whole of government basis, and that this aspect is best managed by each individual department for now. Queensland has bigger fish to fry when it comes to technology projects right now.
Bates’ mention of ‘cloud-based email’ may have some jumping up and down and stating that this opens up the possibility for Queensland to shift onto Google’s Apps platform, or even to Microsoft’s Office 365. However, I consider a migration to these platforms to be a very remote possibility for Queensland, due to their nature as offshore-hosted environments. Australian Governments usually want to keep their data onshore.
“and was also to see some 350,000 identities managed, likely through Microsoft’s Active Directory software”: it’s the IBM Tivoli platform, with the licenses purchased for whole-of-Government as part of the quoted IDES costs.
Very interesting; I hadn’t been aware of that. Thanks for the info :)
“It doesn’t make sense for large organisations to be running their own email platforms”
It does when they have privacy concerns, statutory requirements, or general disdain for outsourcing of business-critical infrastructure. I am a mail admin at a regional university, and although we outsourced student email, there is no plan or even discussion around moving staff email offsite.
I agree with you that there are many cases where large organisations should run their own email platforms. However, staff email at a regional university is not one of those; and many Australian universities have already outsourced the staff email function.
My management doesn’t agree with you; just because it has been done elsewhere, does not make it a good idea. While it has been pretty smooth for the students, their email is not a large problem domain, and on a few occasions I, the mail admin, have had to depend on Google support. Outsourced mail is just a black box – good when it works, totally opaque when it doesn’t.
I had to deal with the WoG mess many times.
It ws poorly managed, and poorly implemented IMO.
I don’t know if an outsourced solution would be best. There is a lot of sensistive data that is handled in these government organisations. Sensitive both to the government and to we the taxpayers personally.
I understand that these companies have a wide variety of agreements etc , but the reality is it is still going offshore.
That said, I can’t see any internal government department managing a WoG type concept well. They just don’t normally have the flexibility and responsibility to manage the needs of the varied requirements of so many organisations.
Frankly other than saving costs, I don’t see a major advantage in the WoG concept anyway. At least when the email comes from a Department of Education/Transport/Main Roads/Premier and Cabinet/State Development address, you know which part of the amorphous beast you have reached.
“They just don’t normally have the flexibility and responsibility to manage the needs of the varied requirements of so many organisations.”
This.
Cloud solutions are really about dynamic workloads, with reduced maintenance and hardware costs as some of the big advantages. Effectively, replacing CAPEX with OPEX, and works really well with a reasonable agile business; where operational funding is the bread-and-butter model used for IT, or where funding is tied more to scale (more funding as a business grows, for example, rather than static allocations).
There is a whole world of hurt in this case simply because of the likely very diverse environments, legacy platforms, dozens of requirements (legal, policy) around data retention, security – it’s really one of the most demanding types of environments there is.
This isn’t a case of saying it’s “bad” per-se. Far from it. Just complex. Real. Complex. And that always attracts a higher cost. You can only manage, plan and budget for so many variables, before hitting diminishing returns – as is often the case in so many areas.
Indeed, but I guess the coalition needs to weigh up the cost of going “cloud” vs the cost of having so many independant Exchange/Notes/whatever Mail systems, and more importantly the admins that support them.
It definately would be a complex undertaking, and long term it would probably be worth it. But short term, they would probably be better off sticking with what they have, and building standards across all the “departments”. So all exchange environments need to work towards this set of standards, all Lotus to these standards etc. Thus moving your independants to a point where the support can be taken over by a single remote location. Reducing the requirements for expensive Sys Admins, and allowing consistency of support. (Yes I know its not that simple, I didn’t say it was simple, but it can be done)
If it was simple, everyone would be leaping into the cloud. Or transitioning to a flat once-size-fits-all type model. :)
Complexity, versus politicking, versus data policies, versus governmental requirements, versus delays = expensive outcomes.
It is my understanding that if you have enough accounts Google will host your data physically where you want it to. (Based on informal discussions with another Google client.)
No, they generally won’t do that, as much as the local salespeople would like to, Australia is just not a big enough fish for them to do it. And regardless, even if they did host it in Australia, as the company managing it is in the USA, the USA PATRIOT act would require them to hand the data over in response to a legal request for it… i.e. the companies offering the cloud services can’t sign up to the T&Cs that Australian Governments are looking for hosting this data in the cloud.
When it comes to google, remember they’re primary business is internet advertising and page ranking. That’s it. They got a big bag of money when they floated, and bough lots of ancillary technologies and companies that embellish this core business.
If they were a genuine IT services company, they would have been organising data centres and written contracts to woo government to them. I think IT managers have ignored the question, “who am I handing my infrastructure to ?”
QUESTION: name a Queensland Government ICT Project that has actually really delivered during say the last 5-8 years.
AMUSING IRONY: Who has lead the QGCIO and what methodologies were used by QGCIO for ensuring appropriate governance of ICT in the Qld public sector. Who is currently leading the audit of ICT and what methodology is being used.
PREDICTION: THIS PROJECT WILL BE THE FIRST OF MANY.
STATEMENT OF FACT:QLD GOV IT IS A MESS.
TAX PAYER JUSTICE: LETS SEE IF THEY REMOVE THOSE WHO ARE RESPONSIBLE.
This decision is typical of Government decision making. Ministers not understanding IT are advised by those whose skill is in Policy Making, not in having an understanding the area on which they advise action plans. The policy advisors in turn are given a Snow Job by the Director Generals of the same skill set, who are more interested in protecting their empires than participating in a Whole-of-Government solution to a Whole-of-Government problem. Even a Generalist can learn IT over time, but IT graduates [ not Peter Principle technicians! ] make better Generalists because of their training.
I left the old Education & Training last year when a perfectly viable, scalable, economical and effective system comprising Exchange-based email technology, accounts created using Identity Management which allowed security-controlled access to a secure stable data structure and a bandwidth-limited comms network, served over 2000 schools and well over hundred Administrative sites with all the user interface equipment comprising desktop, portable, smart phone, printer, scanner and fax machines and the servers – real and virtual. The non-Microsoft servers in TAFE Colleges had an interface. The TAFE environment non-Microsoft contract was renewed due to a Non-Technical consideration instead of being included when the administrative area of Training was integrated with the Education system being rolled out. This system worked and needed only a small coordinated three-level support team for tens of thousands of users.
To integrate every administrative Public Servant in other QLD Govt departments into this big, scalable working system would be a matter of using the same processes – it would need resourcing but the Corporate Memory is reasonably fresh and the staff would still have their skills.
CITEC has been a failure for years but it had powerful protectors. It was too big to be allowed to be admitted a failure, and a viable solution to e real problem was ignored . The Big Bang Theory of Project Development was too seductive for Ministers to adopt the Incremental Approach. Upscaling a working solution from the wrong Department was not glamorous. So contracts were let to develop a totally new system – reinventing the wheel..
Anyway I am out of it now and the new Government still does not have a clue on what Treasures can be found among the Dross. But I still resent my taxes being wasted by incompetents advising well-intentioned new Ministers to implement poor decisions.
More cleanouts to come. CITEC has been a problem for years. It has little credibility left particularly now that the spot light is being flashed around. I doubt it is saleable except if it carries existing ongoing contracts value. Will be interesting what happens now with this department.
QLD government runs a closed environment to protect their data (to a degree). So outsourcing Email / collaborative projects wouldn’t fit that model. News has just come in on the federal eHealth project being hacked before it even got started so that is the world they are operating within regarding confidentiality.
Wow… 2,000 accounts at a cost of $46 million, that’s $23,000 per account!!! Geesus! $23,000 per user, taking it out linearly for 80,000 users that’s $1.8 billion!
So, will Microsoft be promoting the TCO benefits on this one in their marketing ?
It will be very interesting to see how this story unfolds. The three options are already canvased above:
(1) Pick the biggest/best internal solution environment and migrate the other departments into it and scrap the rest. This seems low risk/low cost … but doesn’t drain the dysfunctional governance swamp and would likely prove to be high risk/high cost in practice because they would still need to create a service-oriented WoG shared services organisation in order for all agencies to be happy vegemites … good luck with that.
(2) Outsource to an onshore managed service arrangement. This would put it in the hands of a supposedly competent third paty and keep the data onshore …but the government would still end up having to fund all future infrastructure and software development and maintance costs at exhorbetant consulting rates and historically has not demonstrated competence at managing big outsourcing “partnerships”. Government’s propensity to continually churn its requirements would make this a “nice little earner” for the managed service provider …
(3) Go public cloud with Google or Microsoft. This has huge benefits in terms of cost reduction and accelerated innovation arising from the iterative evolution of the services’ functionality, support for mobility etc. … but the executives need to get their heads around a whole new logic of procurement and risk management, which is challenging. The irony is that a public cloud service is actually the only way in practice that they could ever actually achieve the promised WoG shared service … because it already exists … cloudy is as cloudy does. They just have to get their heads sorted to actually use a shared service rather than wasting time, effort and money trying to reinvent the wheel.
Its all about tradeoffs. My view is the by far the best option is public cloud. It can and has been done at scale – talk to Monash University – and the experience of early adopters is without question better, faster, less expensive and less risky when all things are considered. All of the supposed “showstoppers” are simply business requirements that need to be thought about carefully and appropriate governance, process and techical solutions put in place. Patriot Act FUD is simply that … and would likely exist with any US-based managed service provider to the same degree as it would in a public cloud solution.
Prudent and professionally managed progress towards public cloud services will deliver the best ‘bang for the buck’ for Queensland taxpayers because it will develop the skills and knowledge needed to undertake a step change in the quality and performance of agency ICT. Perpetuation of the status quo of fearful underinvested, poorly managed, poorly skilled, in-house, on-premise ICT will deliver … what exactly? More of the same …does that sound like a good plan to anybody other than the contractors and other folks with a vested interest in the agency ICT gravy train?
The biggest benefit of public cloud services is that it radically externalises ICT capabilities beyond the agency, beyond the state, beyond the country … to a scale of shared services that actually works and can’t be screwed by by the dysfunctional governance of ICT in government. The practical experience of early adopters is that radical externalisation is a salvation more than it is a risk precisely because it means that all agencies can do is consume a service rather than endlessly stuff around with trying to get even the most basic things to work … IMHO.
Just imagine what could be achieved in government if we stop wasting everyone’s time and money reinventing and poorly implementing and delivering these commodity ICT services and instead focused on using ICT to drive policy and service delivery innovation?
Comments are closed.