Shared services: Learning the lessons from Queensland and WA

Really, they should have seen it coming. All the signs were there. Queensland Health, a state government department responsible for paying its 78,000 staff some $210 million in salaries fortnightly, was in dire need for a replacement to its existing payroll system.

Support for its LATTICE system had expired in September 2008, leaving the department vulnerable and unable to make significant changes.

Implementing a replacement would be complex, covering 205 individual allowances across 13 different awards and five different industrial agreements. Things didn’t go well.

In fact, according to the Queensland auditor-general, the project was doomed from the start. Signs of project failure emerged early. Between February in 2008 — just two months after contracts for a new payroll system were signed — and March in 2010, project contractor IBM Australia submitted some 47 requests for changes to Queensland Government’s shared services supplier, CorpTech, a result, apparently, of poorly defined business requirements originating at the beginning of the project. Yet, CorpTech approved all requests.

“During October 2008, detailed planning revealed that the size, complexity and scope of this phase of the program had been severely underestimated, with the consequence that its revised implementation cost estimates significantly exceeded the original tender proposal,” the auditor’s autopsy report reads.

The Queensland Health payroll project would ultimately cost $102 million to implement but go on to take double the time and price to fix. The project’s failure, ultimately pinned on improper testing and lack of contingency reports, could happen to any organisation, but the repercussions of this collapse were huge. Queensland Premier, Anna Bligh, held IBM Australia responsible for the bungle, government agency CorpTech lost exclusivity as shared services provider to other departments and the Queensland Government Chief Information Office was submitted for independent review.

The payroll project and other cost blowouts and project failures among various governments have led to the viability of the shared services approach being questioned.

Most recently, the West Australian government accounted it will decommission its shared services body, following a report from the Economic Regulation Authority (ERA) which advised the state's shared service firm to cease rolling-in new agencies, and further, to disband entirely.

Despite the question marks hanging over shared services New South Wales has forged ahead with its plan, establishing and in some areas even accelerating a services framework to consolidate operational IT staff across major departments.

The Federal Government is also continuing with shared services initiatives as recommended by the 2008 review into ICT procurement by Sir Peter Gershon.

Even without the overarching guidance of lead agency, the Australian Government Information Management Office (AGIMO), individual government departments have begun discussing the potential of virtualising and sharing their infrastructure, creating an internal government Cloud that may ultimately become a precursor to aggregated data centre arrangements.

Page Break

Diving headfirst

AGIMO’s first assistant secretary of agency services, John Sheridan, is careful not to point fingers or comment on the failures of other shared services projects.

“These are, make no mistake, complex projects that require a great deal of work and support in governance terms, in business terms and in IT terms,” he says.

However, Sheridan is keenly aware of the potential pitfalls of shared service arrangements and the measures required to ensure the success of migrating resources and contracts between IT departments within government agencies.

As first assistant secretary, Sheridan is effectively second-in-command to Australian Government CIO, Ann Steward, and oversees much of the shared services work underway at the agency and across government.

Among those tasks, AGIMO, and the wider Department of Finance and Deregulation, has since 2008 been tasked with finding $1 billion in ICT savings as outlined by Sir Peter Gershon in his report to government, a task Sheridan argues was achieved in October 2009.

The report as a whole has become a bible for the federal and state governments; a source of inspiration Sheridan and his counterparts have drawn from to deliver successful ICT activities. And the bible’s decree on shared services is this: Tread lightly.

“[Peter Gershon] wasn’t recommending against shared services; he was commenting on the need to look at these things carefully,” Sheridan says. “Any major IT project is going to have risks to it, and these need to be properly considered and time allowed and resources allowed and the schedule to progress those things.”

A review into the progress of agencies enacting recommendations in the Gershon report, authored by government consultant Dr Ian Reinecke and released last year, issues a similar warning: Beware pitfalls.

The first of these, and perhaps most important AGIMO’s Sheridan argues, is a common law of ICT: Business must dictate technology, not the other way around.

“It’s not like you have to be a brain surgeon to work this out,” he says. “The challenge comes when you try to change business processes with an IT solution.”

The internal processes are the first step in determining how to improve the business, and ultimately, allow for innovation.

It’s a point Fujitsu’s UK head of public sector desktop services, James Mayo, agrees with. Any software, shared or otherwise, that attempts to mould the business process will ultimately fail. The Queensland Health debacle is a case in point.

“When do you bite the bullet on the multitude of terms and conditions in place across departments?” Mayo says. “As soon as you say, ‘We need a payroll that will accommodate the existing terms and conditions’, then automatically you’re not releasing the standardisation of shared services elements.”

Page Break

Co-ordinated not centralised

Like Cloud, shared services can easily become a semantic battleground; a mire of varying definitions, visions and confusion about the benefits shared services can deliver. While expected cost savings of between 20 and 30 per cent, and some reaching savings of up to 35 per cent, have been stated as goals across varied shared services projects, the business case only holds up when the project is clearly defined.

For Fujitsu’s Mayo, the important distinction is between ‘centralised’ and ‘genuine’ shared services. Under Mayo’s definition, a centralised model involves small agencies or departments tacking or migrating their services into the wider infrastructure of a larger hub. Similar setups are already in effect in Australia, with the federal Department of Human Services on-loading some of the infrastructure, IT staff and Web hosting services from smaller agencies such as the Department of Veterans’ Affairs. In such a case, there seems little benefit for Human Services itself; a goodwill gesture more than anything toward a fellow department.

In contrast, Mayo sees ‘genuine’ shared services as those requiring agencies of similar sizes or those unable to reach economies of scale themselves migrating their services together in order to reach the critical mass required for cost savings. The gulf between the two, he argues, is vast.

“The economies of scale are already there in the larger organisations, meaning a business case doesn’t necessarily stack up [for shared services],” he says. “If a small organisation does come along, then they are the small partner. The savings aren’t necessarily there in the big organisations to need to pass on the aggregated pricing, because they’re big enough themselves.”

For AGIMO, distinctions are made between ‘centralised’ and ‘co-ordinated’ procurement. “One of the things we don’t want to create, and we’ve gone out of our way not to create, is some sort of centralised procurement,” Sheridan says. “Our work is about co-ordinated procurement aggregating the government’s demand, getting the sorts of advantages that it gets but being conscious of the need to make sure agencies’ requirements are met.”

Mayo and Sheridan’s semantic arguments both indicate a need for those involved in shared services to decide on how the concept works in practice, whether anchored to a large department already capable of reaching economies of scale, or aggregating similar capacities into a larger pool of demand for resources and equipment. Mayo points to UK examples where co-ordinated approaches between similar councils have achieved greater benefits than the centralised approaches currently being trialled at higher levels of government there.

What’s common in both approaches, though, is tightly structured oversight of the shared services arrangement and an equal footing for all departments, regardless of size.

“True collegiate behaviour clearly comes to the fore,” Mayo says. “I think you can see that in the benefits that come — the more collegiate the organisations, the more those savings can be realised or indeed overstretched in terms of what innovation can be brought and the learnings from each of the implementations.”

At least at the Federal Government level, that’s already underway. As part of his recommendations, Dr Reinecke suggested AGIMO explore additional methodologies for risk assessment, benefits realisation and project planning and management in implementing shared services.

Page Break

On the lighter side

Those recommendations won’t weight too heavily for the agency. After all, Sheridan is quick to point to the success shared services — or ‘co-ordinated procurement’ — has already brought to the Federal Government in its eternal effort to cut costs.

Perhaps most obvious is AusTender, the Web-based procurement portal that serves as a frontend for publishing, communicating and subsequently awarding nearly all tenders at the federal level for services and equipment. According to the Department of Finance and Deregulation — AGIMO’s supervisory department and AusTender manager — the shared service provides a single help desk facility for the 109 agencies which currently use the portal. Some 84 of those use the portal to communicate with 86,000 registered suppliers and, across the 2009/2010 financial year, all agencies involved reported 80,969 contracts at a total value of $42.7 billion.

Some of the biggest departments continue to hedge their bets; the Department of Health and Ageing, for instance, continues to use its own procurement portal for contracting project-specific services including those used in the $467 million personally controlled e-health record. Government business enterprises, such as the National Broadband Network wholesaler and the government’s e-health lead agency, also tender for equipment and services by their own means. Despite a central help desk, contact details for each tender ultimately go directly back to a specified department lead.

Yet, apart from a few hiccups across several years of operation, AusTender has largely operated without a hitch and provides interesting insight into the complex world of government procurement. According to Sheridan, its success lies in its simplicity.

“It’s the same process across all agencies,” he says. “Same business process, common set of rules, single system, not customised.”

Lack of customisation is yet another semantic area Sheridan is careful to tread around. While individual departments can configure their software, customisation is discouraged. It’s a point well established in government policy too: While customisation of software is allowed to some degree, it is a fallback more than anything else and only approved by the agency executive after attempts to reengineer business processes. In all other areas, configuration is the limit of control by individual departments.

As a means of further standardising arrangements, AGIMO applied the same philosophy to a volume sourcing arrangement with Microsoft and reseller Data#3, which Sheridan says has helped the Federal Government move from 42 contracts governing 41 agencies to a single contract governing 80 agencies.

The new agreement has saved $51 million to date, and is now expected to save up to $80 million over four years. “That means that all those agencies — and remember there are over 50 FMA [Financial Management and Accountability Act] agencies that spend less than $2 million a year on ICT — those agencies are able to buy Microsoft software, their core desktop licences are provided centrally; essentially they just tell us the numbers they want and they’re able to get them.

“They’re able to buy software under contract arrangements at a guaranteed price and they can put their efforts into the things that make their agencies different, their competitive advantage, rather than into the things that are done commonly.”

Share the love

In any ICT project, shared or otherwise, failures are sometimes inevitable. It’s a fact of life that IT managers must ultimately become aware of in order to deal with them when they land. Though Western Australia and Queensland perhaps serve as examples of how shared services perhaps shouldn’t be done, the concept is far from dead. In fact, when implemented correctly, such arrangements have their merits, both in cutting costs and removing duplication.

However, keeping an eye on the potential pitfalls are vital to mitigating against failure. “There are always lessons to learn from things that haven’t gone to plan,” Mayo advises. “Part of the challenge here is to listen to those lessons but also implement them... At the end of the day it’s about how you take that information and change it.”

Related:

Copyright © 2011 IDG Communications, Inc.

  
Shop Tech Products at Amazon