analysis Several weeks ago the Victorian Ombudsman handed down one of the most damning assessments of public sector IT 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 government over the past half-decade.
Shortly after the report landed, Delimiter published an analysis of what the state should avoid doing next. Gershon-style reviews, super-chief information officers being flown in to take charge, focusing on shifting everything into IT shared services centres; analysts agreed that these sorts of quick fix solutions would just lead to more of the same problems that Victoria has already been experiencing.
With these obvious wrong turns out of the way, the focus today turns to some positive thinking. What are the steps which the Victorian State Government should be taking decisively forward at this point? Where can it look to for inspiration and a roadmap to better IT governance?
When we asked three leading Australian IT governance experts what Victoria should do next, one unanimous answer came back clearly and straight away: focus on accountability.
Steve Hodgkinson, a Melbourne-based research director with analyst firm Ovum and a former whole of government deputy CIO for the Victorian Government, issued a research note in the wake of the publication of the Ombudsman’s report noting that accountability in the “devolved structure” of government rests with department and agency senior executives.
“The solution should be approached from the perspective of ensuring that departmental secretaries, deputy secretaries and agency CEOs are paying adequate attention,” Hodgkinson wrote in the paper.
When such executives were paying attention, they would more carefully consider which projects were mobilised relative to their capacity to deliver, how projects were resourced and managed and how critical decisions were made, the analyst added.
Scott Stewart, a research director with Queensland-based firm Longhaus, agreed with Hodgkinson on the issue of accountability. He highlighted a project which then-US whole of government chief information officer Vivek Kundra had instigated in the first several years of the Obama administration which created an IT dashboard through which government staff could easily gain a birds’ eye view of all major projects and IT expenditure across all departments.
Part of the project was that each project had a picture of the accountable executive attached to it — usually a departmental chief information officer.
Although Kundra is no longer with the US Government, the IT dashboard lives on. And in fact, the technology behind it lives on and is publicly available so that other jurisdictions can adopt it. There is literally nothing stopping the Victorian Government from downloading the code to do so — a living example of the fact that public sector jurisdictions usually don’t compete with each other and have nothing to lose by sharing knowledge and technology.
Steve Bittinger, a government-focused research director with Gartner, also highlighted the IT Dashboard as a useful model for Victoria to adopt.
“The previous US federal government CIO, Vivek Kundra, focused substantial attention on this issue during his tenure. Notably, he established the IT Dashboard which shows status and spending information about all major US federal government projects. This is a level of transparency that we don’t yet see in Australia,” the analyst said.
Hodgkinson acknowledged that some would criticise this kind of accountability.
“At this point, some readers will react that senior executives in departments and agencies already face impossible burdens of accountability … so it is unrealistic to expect them to do more,” he wrote in his research note. “Our view, however, is that the sums of money involved in major ICT-enabled projects and their centrality to preparing departments and agencies to face future demands means that fixing this problem is now not optional.”
Re-pitching IT projects
Another initiative undertaken by the US Government under Kundra was an initiative he dubbed “TechStat”. The TechStat toolkit was also made publicly available alongside the IT Dashboard.
As detailed in a very useful article on the subject published by InformationWeek, TechStat sessions see an agency’s CIO and CFO, central government analysts and officials from related departments brought together with staff from the central whole of government office of the CIO to rapidly examine and evaluate the status and future prospects of a major IT project.
“It actually had an immediate effect,” says Longhaus’ Stewart, noting the sessions could be simple one hour sessions, but could take weeks for the departments concerned to prepare for. And if a project wasn’t re-approved through the process, it could have dramatic and decisive consequences for those running the project.
In many cases, after reviewing a project’s current status and likely future, the government decision-makers simply cancelled the projects or rolled them into other similar iniatives in different departments. In one example, Stewart says, three separate agencies came to Kundra’s office with smartcard projects. Two of the projects ended up being cancelled and a decision was made to have one single smartcard project across government.
On the website of the US Government of the CIO, the Government credits the two methodologies — the IT Dashboard and the TechStat system with over $3 billion in cost reductions. It’s likely the decisive decision-making style the systems brought in also had a wider ‘halo’ effect on the accountability and transparency of government IT projects in general.
After all, who would want to be hauled up before Kundra and his team for an intense one-hour TechStat grilling?
Ovum’s Hodgkinson, in his note, promotes a similar system for the Victorian Government. The Ombudsman recommended that a new sub-committee of Cabinet be created to approve and review the progress of ICT-enabled projects within a state. A similar committee existed briefly in 2004/2005. Hodgkinson noted that the challenge in making this sort of system work in practice was that it was difficult to sustain the interest of Cabinet Ministers in “such low-level operational matters”.
An alternative, he said, would be for the sub-committee to only meet twice a year, and for it to focus on accountability — instead of quarterly reporting detail.
“The question that the sub-committee should ask of each secretary, deputy secretary and CEO in charge of a major ICT-enabled project is simply this: ‘Will the project be completed as per the business case and approved project plan, benefit realisation plan and funding budget?'”
If the answer was no, Hodgkinson said, then proposals should be discussed and the project must be revised and re-approved (Kundra-style) or possibly cancelled. “Disappointing the sub-committee should be very damaging for executive bonus proposals,” he wrote.
“If department secretaries and agency CEOs were expected to provide unequivocal assurance that projects were on track to a sub-committee of Cabinet, then they would actually need to know that the projects were, in fact, on track,” he added. “The knowledge of this accountability would flow through into the way decisions were made to propose projects for funding, to set up projects for success and to ensure that the right decisions are made during implementation. The result would be for focus and accountability to devolve downwards from the top.”
Furthermore, Hodgkinson added, if there was any significant change in the chain of command for an ICT project, then the project planning should have to be re-endorsed afterwards by that department’s secretary.
“If accountability is a firm expectation,” he wrote, “then executives will think twice before accepting responsibility for a poorly planned, badly scoped, under-funded, under-resourced project expected to be delivered in unrealistic timeframes. It will no longer be an excuse that the project is just stuck with a set of bad decisions made by people who have already left the building.”
Gartner’s Bittinger points out that the Federal Government already uses the P3M3 project maturity assessment model developed by the UK Government to assess projects and agency capabilities in Canberra. Perhaps this is the sort of framework which could tie in well with the accountability and rapid evaluation suggestions discussed by the other analysts. Hodgkinson also suggested looking at the areas of investment management frameworks, so-called ‘Gateway’ reviews and best practice guidelines, although he noted that the problem is not the availability of good ideas and tools … “it is the appetite of department and agency senior executives to put them into action.”
Master architect groups and skills
One idea which Longhaus managing director Peter Carr is particularly keen on is the establishment of a central architecture group to guide cross-government IT architecture decisions. This would not be a ‘super-CIO’ group, but rather one focused purely on architecture. Major Federal Government departments such as Defence are known to have a great deal of focus on key architecture decisions.
Carr draws an analogy between town planning and IT architecture development, noting that for every major city council, there is a central town planner.
“They are the master architect,” he says. “That role has key clout in terms of the master plan of the city and all town planners must flow through that. The central IT groups; they’ve been unable to achieve that same level of central governance and architectural standard for a master plan for any of this stuff. There’s no true master plan.”
Carr says there is no escaping the fact that there is “tremendous, tremendous complexity” inside the systems of a major government department, but that he doesn’t believe that kind of ‘town planning’ capacity even exists at the agency level — let along across government.
The result is that different agencies procure different solutions for IT projects which can actually be quite similar — such as the situation where multiple smartcard projects exist within the Queensland Government, just as they did in the US before Kundra stepped in.
Carr points out that if each of Australia’s states wanted to replace their driver’s licence systems, it would end up costing something like over a billion dollars in total. However, he said, if the project was architected centrally a solution could likely be built for several hundred million.
This lack of central IT architecture planning results in situations where eight agency CIOs are looking to buy a case management system individually, instead of one coordinating body saying we want a scalable case management solution across government. And that, Stewart says, plays into the hands of vendors, who would rather sell the same solution eight times than once.
The Humane Genome Project had been able to eventually succeed in the gigantic task of mapping the human genome, Carr said. “Most organisations can’t even map their IT systems or services. We’re fundamentally failing. The portfolio view, a central architecture group: We don’t have that in IT. We should have it.”
There’s also another area which Carr believes the Victorian Government can make substantial headway; on skills.
In Delimiter’s last article on this subject, Carr put forth his view that there simply isn’t enough skilled IT labor resources in Australia to go around; especially at the price which Governments are willing to pay for it — and even if you count the resources of IT services companies in the mix. Carr says one positive step forward Victoria could take is to stop ignoring the opportunity to leverage offshored resources to get its jobs done.
“Governments have largely ignored the whole opportunity to leverage offshoring,” said Carr. “They’re the first to complain when they get charged $4,000 a day by IBM or Accenture to have project managers sitting around a meeting table three times a week. But the fact, is, IBM and Accenture will continue to charge $4,000 a day while there’s a model for it.”
“Governments need to fundamentally change the underlying mechanism for it.”
Virtually every major Australian organisation uses offshored resources of some kind, especially when it comes to technology. Every major bank has partnerships with offshoring companies like Infosys, Tata, Wipro or even traditional IT outsources like IBM or HP, which also operate their own facilities in low-cost countries such as India or China. Some, such as ANZ Bank, even operate their own internal facilities in such countries, while offshored call centres are extremely standard practice in 2011.
Up until now, Australian governments have been very reluctant to use offshore resources, due to issues such as data sovereignty, the need to protect the privacy of citizens’ data and the political unsightliness of pushing jobs offshore. However, the financial services sector is also subject to harsh regulatory controls in such areas has been able to deal with many of the same issues by using technology like virtualised desktops which lets data and systems remain in Australia but work be conducted on them internationally.
It may be time Australian Governments started looking at similar options; the times are certainly desperate enough.
The cloud and innovation
One significant aspect of Hodgkinson’s analysis of the way forward for Victoria relates to the need to change the way government technologists think about technology in general. In his research note, the analyst noted that the past half-decade has seen the IT industry experience “revolutionary changes” driven by new technology in areas such as mobility, social networking, cloud computing and advanced analytics.
“The ‘art of the possible’ has significantly developed and there are now many more options for how a policy or service delivery innovation might be addressed in ways that are faster, better, less costly and less risky than the traditional major ICT-enabled project approach,” writes Hodgkinson.
“The paradox is that governments typically tend to regard new solution approaches such as cloud computing as risky because they are unfamiliar … while the reality is that the familiar solution approaches of major ICT-enabled projects too often turn out to be the biggest risk of all when they waste time and money and totally fail to meet business needs.” Hodgkinson points out that some Victorian departments already have hands-on experience with public cloud computing platforms — which have proven “highly effective” as a better alternative to more traditional ICT project approaches. However, new thinking will be required to tap into these new paradigms.
“As the proverb goes, when you have a hammer, everything looks like a nail,” concluded Hodgkinson. “The Ombudsman’s report has highlighted that the hammer of major ICT-enabled projects is failing. While the faulty hammer needs to be mended, we also need to accept that there are many alternative ways to meet business needs which don’t necessarily require the hammer at all.”
There’s no doubt Victoria’s in a difficult spot. However, there is also no doubt that a way forward exists for the state when it comes to remediating its technology governance. Some positive thinking comes from international jurisdictions such as the US and UK, where governments are grappling with extremely similar issues to Australia’s states. And some thinking comes from within Australia, where the private sector has already conquered many of the same problems which Victoria is facing. Major ICT projects are increasingly running on schedule in the banking, retail and telecommunications sectors — industries the public sector could learn from.
Things may seem dire for Victoria’s public sector technologists at this juncture. But the illness can be addressed through a combination of emergency surgery and the long and patient application of the right techniques. Stay tuned for our next report shortly profiling the key players who will be responsible for administering the medicine.