CIO

Vets Gets Selective

There is evidence to suggest that selective sourcing does better for sourcing clients in terms of critical success factors such as achieving targeted cost savings, maintaining or improving service levels, and renewal of contracts.

One of the earliest - and most successful - of Australia's government outsourcers, the Department of Veterans' Affairs is now counting on a policy of "selective sourcing" to deliver new benefits to the agency.

Given its function and title IT seems somehow appropriate that the Department of Veterans' Affairs (DVA) should have been one of the first Commonwealth agencies - one of the first enterprises in Australia in fact - to outsource its information technology. The agency is as proud of its status as a veteran outsourcer as it is of its service to Australia's veteran community.

For a trailblazing novice at the outsourcing game, DVA did not do too badly out of its original outsourcing deal, signed way back in 1992. Indeed in the overall scheme of things, that original sourcing deal was considered quite a success - especially in its contrast to the abysmal whole-of-government contracts that came later.

But time and humanity's knowledge of the state of things inevitably moves on, and now DVA is looking to capitalize on its early experiences, and the lessons it has studiously been learning ever since, to achieve even greater benefits from its outsourcing arrangements. CIO Bob Hay says by unbundling some of its services from its renegotiated contract with IBM, DVA is confident that it can continue to benefit from outsourcing while achieving vastly better cost transparency.

"Bundling was seen as being a strategic move at the time, but the other side of bundling is that you lose transparency about the cost drivers of the various services that you obtain," Hay says. "And that's where I think the direction is heading these days - that you try and unbundle them all [so that] even though you may get the services provided by the same provider, you get more transparency into the nature of the costs of those services. It gives you a better opportunity to test the market for those services and whether you are still getting value for money on an ongoing basis, especially where the technology changes and your environment needs to reflect that so the services you get from your service provider need to change as well, to support that new environment."

Selective sourcing is enjoying new currency in Canberra, and expectations of outsourcers have changed dramatically. Essentially, a selective sourcing approach involves choosing a stable of "best of breed" vendors to administer carefully selected business services or business functions, in either a total or limited point solution. Growing numbers of organizations are starting to recognize the selective sourcing model as a viable and attractive alternative to the single-vendor approach.

Australian Customs Service CIO Murray Harrison told CIO recently the government's CIO Committee is leading the push towards selective sourcing.

There is evidence to suggest that selective sourcing does better for sourcing clients in terms of critical success factors such as achieving targeted cost savings, maintaining or improving service levels, and renewal of contracts. Now DVA is determined to put that theory to the test.

Hay says one year into a three plus one year extension to its 1997 contract with IBM (renegotiated from the initial five plus two plus two arrangement), DVA has a highly effective cost projection model that has given it an excellent handle on its costs and likely future cost changes, giving it vastly improved control over the arrangements it has put in place.

Page Break

Early Savings

DVA exists to serve members of Australia's veteran and defence force communities, war widows, widowers and their dependants through programs of care, compensation, commemoration and defence support services. The agency has some 500,000 veteran and dependent clients, and around 60,000 clients through the Military Compensation and Rehabilitation Service. And its obligations just keep growing. For 2001-02, the parliament appropriated $8.75 billion for budget measures for veterans' affairs, compared to $6.4 billion for the year in which the agency first outsourced its IT services. DVA is heavily dependent on high quality and responsive computing infrastructure to deliver that support.

In 1992 the agency contracted out its mainframe computing services to Ferntree, and was able to claim savings of some $10 million by the time of a 1996 Australian National Audit Office (ANAO) report into that original deal. The same report credited the arrangement with having achieved communication and service delivery efficiencies, a 24-hour operation with an uninterrupted hardware supply and generally a more professional attitude among staff.

Five years later DVA was even able to confound some of outsourcing's critics by successfully retendering the mainframe operations, throwing in networking and desktop support, and - to Ferntree's disappointment - giving the whole lot to IBM Global Services Australia (IBM GSA). DVA retained application development for itself.

But the same ANAO report into the original sourcing deal also found the contract between DVA and the supplier deficient in a number of areas, and the agency has been working to address these deficiencies ever since. Happily, some contracts expire as agency needs evolve so now this veteran of federal government IT outsourcing is moving into a new era of sourcing arrangements - and this time, the key words are "unbundling" and "selective".

In February 1997, with the original contract expired, DVA became the first federal government department to outsource its desktop-to-mainframe computing infrastructure. The contract was granted to IBM company Integrated Systems Solutions Corporation (ISSC) Australia Ltd, at an announced contract value of $65 million. The deal was to extend five years from its signing in February 1997, with potential for two extensions each of two years duration. Prices were to be reviewed twice during the contract while other terms and conditions, including service levels, were expected to remain unchanged. The contract provided that it should operate in accordance with the concept of "partnering" as agreed by the parties.

Under the contract, DVA sold its IT assets to ISSC for $5.5 million, with the understanding that in future it would buy its IT and telecommunications goods from IBM.

In September 1997 that contract was replaced by an expanded contract which - amidst a storm of controversy over dubious tendering processes - drew the Department of Finance and Administration (Finance) and ComSuper into the deal with the idea of achieving economies of scale.

Again, the contract has delivered some good value over the years, but an ANAO report prepared in the lead-up to the contract's initial expiration last November found significant room for improvement. As well as cost blowouts, which ANAO found had pushed total DVA IT expenditure over the period to $140 million, the report criticized DVA's failure to adequately manage the performance of outsourcer IBM GSA.

"There was insufficient alignment between DVA's IT strategic plan and the contract, particularly given that the department had expanded its use of client/server technology [including implementation of new systems and attendant network upgrades] after the contract was signed," the report's authors concluded. "This made it difficult for the department to obtain assurance that it was achieving best value from the contract in the context of its strategic direction . . . There was limited information by which to assess whether the department received value for money from its IT services . . . and limited application of risk management."

In short, DVA had taken for granted that outsourcing would deliver it real value, and had failed to establish the hard metrics needed to truly test that proposition. Now DVA has determined that all that will change.

Page Break

Bringing the work back in-house on expiry of the contract was clearly no answer - Hay acknowledges that far too much corporate knowledge has been exported for that. Instead, DVA looked at available market offerings and the arrangements other Commonwealth agencies had recently negotiated in the marketplace, entered into intense negotiations with IBM and was "eventually satisfied with what the costs were for the services that we wanted to get".

It started by renegotiating that contract at the end of the fifth year (in 2002) based on a careful examination of the nature of the services it wanted, and drawing on advice from industry about comparable arrangements in the market.

"At the end of the day we made our own market assessment as to what the new renegotiated arrangements would be, and came to the conclusion that it was still value for money to extend that agreement," Hay says. "So we extended that agreement for a three plus one timescale, which still essentially kept it within the original nine-year time frame. At the moment we're just over the first year of that three-year extension."

Before doing so, DVA made a concerted effort to get a feel for its present and future cost drivers and the nature of its future business and future infrastructure, and examined ways to control those costs. It changed some fundamental principles of its operations, which would, over its forward year time frame, provide the agency with a controlled environment in terms of understanding its costs and cost drivers, and give it the ability to map its changing needs on a consumption basis to the services provided. Hay says the outcome was "quite an effective cost projection model" allowing it to achieve a good handle on its costs.

Naturally enough, the major cost drivers lay with the mainframe (which provides a core environment for most major applications), followed by disk storage and support costs to DVA's 60-odd sites around Australia. Changing the infrastructure in order to reduce those costs thus became a priority for both DVA and IBM. The answer ultimately proved to be to centralize the major service, and move to a different, thin-client technology. That transition - a joint project between DVA and IBM - has just been completed.

"It's a joint project between ourselves and IBM because infrastructure provides the mechanism to operate, but clearly, applications and the operating environment on that need to be jointly managed," says Hay. "So the project has essentially operated under a joint management arrangement, and has been very tightly controlled in terms of what needs to be done. It's an interesting example of where we, as a government organization, have worked very closely with a non-government organization in managing resources which are available to both parties."

Hay says the exercise - which required significant culture change within both organizations - has been a highly effective exemplar of how two organizations can work cooperatively and achieve common objectives. It also illustrates a mature understanding about the roles both sides have to play and the nature of the cooperative relationship required for outsourcing to work effectively.

"We see IBM as a strategic partner, and while there's still obviously a commercial relationship between us, we've found that the working relationship has been very effective in achieving both our own objectives," Hay says. "It's a matter of us understanding what their drivers are and them understanding what our drivers are, and recognizing where we need to work together to achieve that. So I think that's been a very good example of where we have been able to work effectively with a private sector organization."

The next step was to look at the full range of services to be outsourced and to identify which as a business IBM was really interested in supporting, to identify which other service providers specialized in particular fields, and whether at the end of the day it was better value to go down a different path. The outcome is a long list of activities that are now unbundled and sourced through other arrangements, including:

  • capacity management
  • disaster management / recovery / disaster recovery plans
  • building environmentals (previously IBM was responsible for UPS, air-conditioning, and so on)
  • software management (DVA is now responsible for all software management, including upgrades)
  • certain movements, additions and changes applying to phones, desktop devices
  • management of PABX operators
  • various procurement activities
  • hardware and software refresh strategies other than those specifically identified in the contract
  • voice facility management
  • printing services.

"The new agreement enabled us to get much greater clarity about all the other elements of the services that we're providing, from mainframe CPUs to server costs, to disk storage charges et cetera, et cetera, which under a bundled arrangement you don't get to see," Hay says.

"And that's one interesting aspect in terms of business operations. Where you've got a corporate resource, it's often taken for granted that these resources are always available to other areas of the organization without actually understanding what the true costs are. It's therefore a particular emphasis that we are now focusing on: How do we provide much greater transparency to the business areas, about the nature of the resources and therefore the IT costs that they are consuming?"

Page Break

Selective Benefits

Although it is early days, and Hay admits the organization is yet to meet any of the potential downsides of selective sourcing, he says the benefits on offer look promising. Above all else is better transparency and more certainty over costs, he says, but that is just the start. DVA also gets the benefit of an infrastructure refresh and a new way of supporting its business operations, of which the thin-client environment is the key element.

"[Thin client] provides us with a great deal of additional benefits in terms of responsiveness, particularly in rural and remote areas. It provides us with a greater flexibility in terms of having a mobile workforce, and gives us plenty of opportunities for being able to leverage off that capability for the way we do business in the future," Hay says. "Apart from giving us an updated platform, it gives us an opportunity to look at how we do business better in the future."

Employees in the regions are also happier because they get much improved access, to the extent that they now say they prefer thin clients, Hay says.

Meanwhile, the new arrangements have only served to improve the relationship with IBM.

"My observation is that IBM understand very well what our business is and they have a very pragmatic approach to dealing with government; they understand and are ready to work with us in meeting our needs and providing ways and options to do that," Hay says. "Now, they've got concerns, sure - it's an element of business that they no longer have. But on the other hand they've still got a larger business that they're supporting. And there are always business opportunities that emerge for them during the life of a contract anyway, which still provides them with an opportunity to continue meeting their revenue stream, if you like."

Hay says in three or four years time DVA will again make a strategic decision about the nature of the services it wants, and will ask itself more seriously whether there is any value in bringing services in-house (probably unlikely) or whether to go back to the market.

"Having now built up our experience in terms of being able to ask the market about a range of services, I think we're in a much better position to be able to go back at that time frame to look at what we get," Hay says.

"From a government organization's perspective, there's a much greater degree of collaboration now between government organizations, and one of the issues we will need to consider is the extent that we collaborate with other organizations who receive our services. Piggybacking arrangements are also becoming much more frequent. I think it gives us a new dimension to the nature of the range of services that we can get and the sources of those services."

DVA is also much more focused on taking an architectured approach to the way it wants to do business internally, Hay says, and this will shape some of its infrastructure requirements.

"Fundamentally though, like most organizations, we have a lot of IP in applications that have been built up over a long period of time, and will continue to need to support our business operations," he says. "We will have to recognize that those legacy systems are systems that we will have to continue to support, so we will have to make sure that the infrastructure is able to support them. At the same time, we have to take advantage of new opportunities that come with industry in terms of applications development, or the way that technology architecture works to ensure that the services that we get, the application services that we need to provide to our users, are also able to be effectively delivered."

Hay says although it is far too early to assess the results of the new arrangements, he is optimistic. He says the lesson for CIOs is that collaboration is important. There is leverage simply from a purchasing perspective in being able to recognize the common grounds between an outsourcer and client, and in figuring out your real costs before entering into any outsourcing arrangement.

For the rest, only time will tell.

SIDEBAR: Contract Management

Be Prepared By John Kopeck

Successful sourcing requires pre-contract planning

Successful sourcing relationships are based on careful and detailed preparation before the contract is signed. Both parties need a clear understanding of the type of relationship desired. Sourcing contracts can consist of anything from a brief document with a price list attached, to hundreds of pages of legal documentation, terms, schedules, and flow charts.

While important, written documentation should be preceded by a thorough understanding of the type of relationship the client wishes to achieve and maintain over the life of the contract. The amount of time and effort required to structure a deal depends on the type of relationship being built. Key considerations include the following:

  • Pricing: is the client willing to pay a premium for specialized expertise and business knowledge, or is low cost the primary objective?
  • Vendor involvement: does the client expect a high or low degree of input and advice from the outsourcer?
  • Length of relationship: does the client seek a long-term relationship, or are frequent vendor changes to be expected?
  • Number of qualified vendors: are the services sought by the client highly specialized or widely available?

Based on combinations of these various factors, four different types of outsourcing relationships are feasible.

VALUE-ADDED PARTNERSHIP. In one scenario, clients seek a high level of vendor involvement over the long term, and are willing to pay a premium. Few vendors are qualified to provide this type of service. This type of relationship is the true value-added strategic partnership. The client expects significant ongoing input from the vendor to identify opportunities for performance improvement. This may come in the form of access to the vendor's technology, experience, or thought leadership. The vendor puts world-class technicians and relationship managers on the client-facing team, for which the client is willing to pay premium rates and negotiate a long-term contract.

Clients in a value-added relationship tend to be large multinationals operating in complex, fast-changing markets. Vendors capable of providing this level of service tend to be large themselves, with global operations, highly competent and experienced practitioners, and deep technological resources. The value-added relationship will often undergo significant change over the life of the contract, requiring a higher degree of flexibility in its construction and undertaking.

CRITICAL SERVICES. Another type of relationship involves a high level of vendor involvement, at premium prices, over a long period of time. The difference is that markets are served by numerous vendors of approximately equal capability.

Audit and tax firms, many outside legal advisers, and global recruiters fall into this category. The client seeks specialized, on-going advice, but is willing to substitute vendors to get comparable service at a lower cost. That said, longer-term contracts are desirable, as the vendor gains knowledge of the client - thus minimizing the learning curve of each incremental engagement.

SPECIALIST PROVIDERS. Highly specialized service providers fill a specific niche that are not strategically critical to business operations. However, clients recognize that only a few vendors perform these functions well - and are thus willing to pay a premium for them. Global payroll agencies and international travel firms are examples of such providers.

TRUE COMMODITIES. Finally, in true "commodity" relationships, clients outsource non-strategic services to one of many competent providers on a short-term, price-oriented basis. Long-term contracts are not desirable as a new player might enter the marketplace at any time, offering significant rate discounts as they build their business. The client does not seek significant input on how the job will be done; they just want it done, on time and to specifications clearly delineated in the contract.

Page Break

Switching Objectives

So long as both client and vendor understand where their objectives and capabilities reside, the relationship will likely succeed. Difficulties arise when either party is unsure of, or changes, their position or stated objectives. Such shifts naturally lead to organizational and contractual tensions - sometimes ending in claims of contractual default.

Consider, for example, the client who emphasizes access to world-class capabilities in interviews with potential vendors, but who then negotiates solely on price. In essence, the client is asking for "added value" - but wants it at a discount. Similarly, vendors who represent themselves as providers of world-class capabilities, but who then staff an engagement with junior or inexperienced practitioners to save costs, are just asking to be treated as commodity providers.

Clients who over-spec contracts are frequently frustrated by the lack of flexibility demonstrated by their vendors, while those same vendors are frustrated that they can't demonstrate their creative capabilities to the client under the constraints imposed by the contract.

Occasionally, clients declare functions to be commodity-type services that really aren't. A number of insurance and financial services firms learned this lesson the hard way, after customers voiced dissatisfaction with service levels provided by outsourced contact centres. As a general rule, direct customer contact in service or retail businesses is best considered to be a strategic function and not to be outsourced to commodity-type vendors.

Freeing resources for other, more strategic purposes might seem like a viable goal, but only if the outsourced operations don't require significant two-way communication between client and vendor, or an in-depth understanding of the business that the vendor can gain only over time. A number of high-profile sourcing contracts have unravelled after outsourcers, particularly offshore outsourcers, failed to demonstrate adequate knowledge of the client's business and customer needs.

Picking and Choosing

For clients who take the time and effort to truly understand the costs and strategic implications of their non-core operations, "selective" sourcing is often the best approach. In other words, a client that needs access to world-class capabilities for some operations outsources those operations to a world-class provider under a long-term, relatively flexible contract, and then listens to that provider's advice. For this level of service, the client should expect to pay more than "best market price" and in return regularly receive value-adding input.

This same client can then outsource operations not requiring world-class capabilities to a commodity provider under a tightly-constructed, cost-focused contract which includes provisions for termination should the provider fail to meet expectations, or should a lower-cost provider come along. Under the best-case scenario, the client will optimize the cost and performance of these operations before outsourcing them to the commodity provider.

A well-managed approach to selective sourcing offers the client the potential for the best of both worlds: strategic help and access to world-class capabilities to add value, and low-cost provision of non-core services to maximize efficiency. As a side benefit, the client can free resources for other purposes, hopefully purposes that improve the company's focus.

John Kopeck is president of Compass North America