It was the kind of assessment that no government organization would have hoped for. After reviewing IT spending and control structures at the Commonwealth Department of Veterans' Affairs (DVA), in 2001 the Australian National Audit Office (ANAO) identified $75 million of IT overspending during the department's five-year outsourcing contract with IBM Global Services. This included $47 million worth of work that was not initially included in the scope of the contract, but was performed with other suppliers. Such experiences are hardly isolated to DVA, but the DVA report was particularly poignant because it provided clear evidence that the government's then-heated push into broad IT outsourcing had not necessarily reduced costs as was expected. DVA was among the first departments to outsource, and its cost blow-outs confirmed that the business plan supporting the government's push into whole-of-government IT outsourcing was fundamentally flawed.
A follow-up ANAO report, released in July this year, found that DVA had introduced several change management methodologies to better align business expectations with IT capabilities. However, there is still work to be done: DVA's Balanced Scorecard, for example, was flagged as failing to incorporate quantitative measures of client service that are essential for testing performance against business case expectations.
DVA's ongoing development of change and performance management infrastructures reflects the complexity of large-scale IT deployment, and highlights just how tricky it can be to build accurate business cases within government organizations. While most government IT executives would like to think they have a good handle on the process of building compelling business plans, ongoing analysis suggests that IT project approval and follow-up processes are still far less mature than they should be.
The Problem with Government
In the private sector, business case requirements are clear: proposed IT projects are evaluated and approved according to strict criteria that reflect the project's potential contribution to the bottom line. If a project cannot reduce operating costs or improve profits, it is pushed back in the queue behind other projects that can.
Many organizations compare business cases against hurdle rates - arbitrary targets for profit improvements - that must be met for a project to be approved. Although this approach reflects the corporate expectation that IT investments generate profit, it can also pressure project sponsors to justify IT investments that are seen to be the right thing to do - but may very well not be.
This, in turn, leads to questionable business case development - particularly when project sponsors rely on outside estimates of future returns. Vendors have ready archives of business cases justifying their respective technologies, but do not believe them. Predictions from vendors and analysts - educated guesses at best and deluded optimism at worst - all too often become treated as fact, replacing natural scepticism in the quest to deliver a project plan that will cross the hurdle rate.
"In lots of cases, people put together fictitious business cases," says Dr Kevin McIsaac, research director with META Group and a veteran of many business case assessments. "I look through many vendor business cases and they're wildly optimistic; there's a lot of fudge factor thrown into them. To get their pet projects through, many people start with the ROI they need to have and work backwards. We've got to be a lot more critical about these things."
Government business cases must reflect the many differences between government and private-sector organizations. For example, many private-sector business cases assume a certain level of cost cutting - particularly as IT systems make many employees redundant - in projecting their benefits. In government, however, mass layoffs would be political suicide and are all but unthinkable, so business cases simply cannot be built around savings from salary elimination.
Private-sector business cases are often built around intangibles such as brand strength and market position. In government organizations with legislated monopolies over service delivery, however, such concepts have no meaning and therefore cannot be counted upon to justify a business case. Intangibles of importance to government organizations are far less concrete issues such as customer satisfaction, service efficiency and operational efficiency.
Whereas private-sector organizations benefit from the follow-through that comes with relatively long executive tenures, government organizations must expect major policy shifts after elections every three or four years. This makes long-term business cases difficult, particularly when political restructuring merges two departments with unique IT plans into a combined entity requiring painstaking technical and procedural integration.
In such cases, business plan horizons are substantially extended and cost overruns can easily ensue. Needed systems upgrades may be delayed, business cases must be revised, and change management processes will be tested. Yet throughout all this, political overseers may still be counting on the IT organizations to deliver on policy goals - thereby setting the stage for the cataclysmic failure of major initiatives.
Guidance from Above
Within government organizations, increased scrutiny on expenditure - including control processes such as dual review layers and the potential for intense public scrutiny - would seemingly foster a more realistic business case environment. And while failure of a major IT project benefits nobody, it remains all too common. As illustrated by the Commonwealth's disastrous outsourcing push during the late 1990s, approving IT investments based on presumed benefits can become a nightmare if the presumptions used are even mildly over-optimistic.
Cost overruns, lack of formal change management, and fragmented reporting and performance monitoring frameworks have all driven dramatic shifts in government IT policy. Recognizing that government organizations would benefit from clearer guidance in IT business case creation, state and Commonwealth governments have formalized guidelines that seek to eliminate spurious business cases by ensuring IT investments reflect broader policy objectives.
The NSW Office of Information and Communications Technology (OICT), for one, recently updated its Business Case Development Guideline (www.oit.nsw.gov.au/pdf/4.4.4.BCD.pdf). And Commonwealth departments benefit from work by the Australian Government Information Management Office (AGIMO), which recently published its Demand and Value Assessment Methodologies (DAM & VAM, at www.agimo.gov.au/government/damvam).
Designed to help government agencies assess the potential value of e-government initiatives, this resource includes the Demand & Value Assessment Methodology Manual, 19 activities related to business case development, a CD-ROM and a variety of other materials. Training courses are being held in Canberra to help project managers and other Commonwealth government IT decision makers apply DAM & VAM principles to their planned e-government projects.
Western Australia has taken a different approach to IT business case assistance, building on the August launch of the state government's e-Government Strategy for the Western Australian Public Sector. Offering broad, strategic online service delivery goals for all state agencies, the strategy's introduction has been backed by the publication of numerous resources to guide departmental IT decision makers.
The WA Office of e-Government (www.egov.dpc.wa.gov.au) has published a guide, Guidelines for Aligning Projects to the e-Government Strategy, that spells out how individual projects should be planned to coincide with the broader goals of the e-Government Strategy.
"There's a lot of work being done on building business cases for e-government," says Jo Bryson, executive director of the Office of e-Government, within the WA Department of the Premier and Cabinet. "What isn't happening is looking strategically at them and asking whether they're in line with where the government wants to go generally. We're looking quite strategically at those issues that you don't necessarily have in your run-of-the-mill business cases."
WA government organizations preparing proposals for major e-government initiatives must now be assessed against eight core enablers spelled out in the e-Government Strategy: leadership, the change to a "corporate WA" culture, governance mechanisms, a citizen-centric approach, collaborative relationships between departments, policy and legislative framework, technology architecture and interoperability, and information management.
The requirement that e-government projects reflect state-wide IT objectives is a significant attempt to unify business case development strategies across the government. Inter-agency collaboration and resource sharing are particularly encouraged, as are "citizen-centric" projects with a focus on improving access to government services.
Body of Knowledge
Promotion of collaboration between government departments may ultimately prove beneficial in other ways: open information sharing between departments will fuel a cyclical feedback process that should improve overall business case development practices.
Creation of a business case body of knowledge within government organizations will be a natural outcome of two other initiatives now gaining currency at various levels of government. The first is shared services, through which several state governments have centralized core IT functions into a single body that delivers services to various departments on a fee-for-service basis. This approach will add predictability and a clear understanding of IT benefits, which will in turn foster the creation of more accurate business cases.
Government is in a unique position to do this, since the lack of a competitive element gives departments unparalleled incentive to share knowledge. Business imperatives may be handed down from above, but failures must be shared as well as successes, so that departments can learn from experiences good and bad. Knowledge sharing can also help smaller organizations overcome resource limitations that consistently present obstacles to business performance monitoring.
"Commercial organizations are well ahead of government organizations when it comes to measuring results and lining them against what was approved in the business case," says Richard Harris, Asia-Pacific vice-president with research firm Gartner. "A lot of government organizations would be keen to follow best practices if they knew what those best practices were. Many agencies I see are trying desperately to do this, but they're often hard pressed to have the people and skills to do it."
For this reason, government agencies will also benefit from the filtering provided by panel contracts in specific areas of IT. By streamlining the selection process for particular types of technologies, panel contracts will remove much of the subjectivity from procurement processes, eliminating the need to assess proposals from untested organizations whose ability to match promises with results remains untested.
This approach will be particularly important with relation to new technology areas, where everybody is starting with a level playing field and there is a significant chance for poor business cases to survive simply because they cannot be judged against previous experiences.
Panel contracts will bring needed predictability for government organizations taking their first steps into assessing business cases for new technologies. The NSW OICT's recent launch of its panel for open source is an example: by adding some measure of predictability to the open source business case, this structure should give departments confidence in the deliverability of open source solutions.
Building a Better Business Case
In the absence of promises of relative profit improvements, comparing business cases for various projects can be difficult. Business case assessment usually results in a relative and largely subjective score for different projects. Multimedia Victoria's Service Based Business Case Guidelines framework, for example, triages projects into "potential quick wins", "unsure" or "high priority". Any business case assessment will require a similar scale, with projects assessed against standardized criteria.
Changing market dynamics, backed by the overall maturity of IT capability within government, have rewritten the pecking order when it comes to prioritizing investments. Whereas the early days of online government saw a focus on "high priority" projects regardless of the cost, financial prudence has shifted the focus towards the "potential quick wins" side of the scale. Considering this issue in formulating business cases can improve adoption rates, simply by breaking down larger projects into short-term efforts with clear deliverables.
Cost overruns, poor change management, and fragmented reporting and performance monitoring frameworks have all put outsourcing contracts on the front line in the effort to shrink and shorten project expectations. Yet the government's rocky relationship with outsourcing is an important lesson for government organizations at all levels - particularly as departments now work to develop IT policy within an environment marked by politically charged issues including IT offshoring and increased use of open source software.
Although government bodies are still outsourcing, the scope of those projects has narrowed considerably, creating more opportunities for suppliers to specialize in particular areas. Service provider Volante, for one, is bringing in more than $50 million annually from four recently secured managed service contracts for different Commonwealth government agencies.
Volante's success came after the Commonwealth government's Cluster 3 outsourcing group was fragmented into smaller, more manageable component parts. State government organizations have taken a similar approach: the South Australian government has broken up its once-notable whole-of-government contract with EDS, while Victoria's VicRoads recently reallocated $101 million of outsourcing business from incumbents IBM and Fujitsu. EDS, Telstra and Mincom won the business in smaller chunks, reflecting an overall trend towards building business cases in smaller, discrete components.
Small is not enough on its own, however: when individual projects reflect small pieces of a whole rather than the whole itself, it is critical for business plans to link the smaller efforts with an overall IT strategy - as happened in Western Australia.
"The goal of what you're trying to do is not just a bunch of isolated changes, but a whole bunch of purposeful changes," says META Group's McIsaac. "Purposeful changes require an overarching strategy, but only about one-quarter of the organizations I deal with have reached that maturity of having an overall architecture."
In other words, government business cases should be purposeful, deliberate and aligned with strategic objectives. Their creation should also involve a range of accepted strategies, such as repeated surveys of and close interactions with business stakeholders, splitting of business cases into smaller entities, separation of responsibilities so one person is responsible for measuring benefits and someone else assesses project costs, and careful attention to good governance so that project presentation and approval reflects best practice.
All the good planning in the world is not going to ensure the success of any IT project, however, without a clear mechanism for tracking a project's real performance. Lack of performance monitoring strategies was a major focus of ANAO's initial audit of DVA, and was also a focus of ANAO's follow-up report this year. That report cited problems with DVA's Balanced Scorecard methodology, which did not reflect customer service performance; DVA says these metrics will finally be incorporated into the system by year's end.
Balanced Scorecard or no Balanced Scorecard, most government organizations (McIsaac estimates the proportion is around 90 percent) still lack the robust results measurement system necessary to determine whether or not their projects have met the goals outlined in the business cases that spawned them.
Whether through technical means - such as introduction of network, service and resource utilization monitoring tools - or through non-technical approaches such as user surveys, departments must find acceptable and executable ways of reporting on their performance. Without a good grasp on potential problems or shortfalls in deliverables, departments risk unexpected surprises, trial by media and investigations into the procedural issues that led to budget blow-outs and unused capacity.
For all the planning in the world, there is still no silver bullet to building a sure-fire business case. But with careful guidance and an effort to ensure both buy-in and compliance with overarching strategic frameworks, IT executives can at least make sure their projects make sense when they are finally started.
Well begun, as the saying goes, is half done.
Autor: David Braue
Quelle: CIO Magazine, 03.12.2004
