One of the most potent arguments posed by critics of the federal government’s controversial whole of government outsourcing strategy back in 1997 was that IT, once outsourced, might never be able to be brought back in-house.
With corporate memory lost to the outsourcers, opponents argued, how could departments and agencies ever hope to regain control should circumstances change and outsourcers no longer prove capable of meeting their needs?
Now the Department of Education, Science and Training (DEST) has defied the wisdom of those critics and current-day key IT advisory groups alike to bring its IT back home. A carefully implemented strategy has helped it transition from what was effectively an outsourced arrangement for the provision of IT infrastructure services to a model where the bulk of these services is now provided in-house.
In the process it is achieving significant cost savings and improved business value as well as noticeably better levels of service. It has also claimed for itself the chance to start with a clean slate, to assess what had worked well previously and what had not, to adopt industry best practice with new procedures and state-of-the-art systems, and to promote a service-oriented culture among the new staff.
An Interim ArrangementWith an annual budget of around $13 billion and 1600 employees Australia-wide, DEST has responsibility for policy development and program administration in the education, science and training sectors.
Created by a machinery of government change in October 1998, DEST was an offspring of the former Department of Employment, Education, Training and Youth Affairs (DEETYA). DEETYA’s employment function was taken over by the now Department of Employment and Workplace Relations (DEWR).
At the time of the split, department stakeholders believed that the IT infrastructure of both departments would soon be outsourced to an external provider under the government’s IT outsourcing initiative as part of the Group 11 cluster of agencies. Until that time, they recognised, there would be need for an interim arrangement.
With approximately 2500 employees, DEWR was significantly larger than DEST so the decision was made to locate the entire IT infrastructure group within DEWR. The two departments then signed a Memorandum of Understanding for DEWR to provide IT infrastructure services to DEST.
Through the MoU, DEWR would provide all of DEST’s IT services excluding:
—application development and maintenance (ADM) (to be managed in-house in DEST through mix of permanent staff and contractors)
—voice services (to be provided by Telstra)
—hardware procurement related to desktop and some mid-range equipment related to ADM (to be managed in-house in DEST)
The essence of the agreement was that the two departments would share a common infrastructure that would be managed by DEWR, with minimal amounts of effort put into separating the departments at the infrastructure level. Both departments would continue to use the same data communications network, messaging, network security and disaster recovery systems. Both departments would retain ownership of their respective IT assets.
The two departments continued to work jointly on a tender for the Group 11 process with the expectation of signing a contract with a service provider by mid-2001 until, acting on the recommendations contained in the Humphries report, the government formally terminated the Group 11 process in March of that year.
Left in the LurchThe decision did not unduly disadvantage DEWR, which got to keep control of the infrastructure group. DEST, on the other hand, was left in an extremely uncertain position with a wide range of options under consideration, including:
—continuing with the existing MoU arrangement
—pursuing the outsourcing initiative with an alternate set of service providers
—bringing some or all of the services back in-house
On the face of it, continuing the status quo would require the least effort by either department. Results from independent IT user satisfaction surveys indicated that DEST users were generally satisfied with the level of service delivered under the MoU. The results and recommendations of an independent cost benchmark study and sourcing strategy for IT infrastructure services also indicated benefits for both organisations in continuing the existing arrangement, subject to a number of issues being addressed.
However, two years of operation under this arrangement left it clear some of these issues would require careful consideration.
For one thing, the MoU was to be a temporary measure only, designed to minimise disruption to each agency’s normal business. As a result a number of key areas including pricing, service levels and division of responsibilities had been formulated with considerably less rigour than would normally have been applied to a relationship of a more permanent nature.
The service levels in the MoU had been deliberately specified at a high level. Yet while these arrangements were good enough for an interim arrangement they did not provide DEST with sufficient transparency, certainty or autonomy to optimally operate IT services in the longer term.
Further, the corporate level had significant concerns over the level of flexibility and responsiveness, particularly in relation to the deployment and support of IT applications. There was also a range of boundary issues between applications software and development and operating system software and infrastructure support.
Time for a ChangeIn January 2002, DEWR was formally advised DEST was considering the withdrawal from all or part of the services currently provided by DEWR.
For planning purposes, the IT infrastructure services that were being provided at that time by DEWR, and that would continue to be required by DEST under any future arrangements, were partitioned into the following broad groups:
—help desk (first level support and user administration)
—desktop support (second level support and SOE maintenance)
—data communications, including an Internet gateway environment
DEST undertook to develop an overall sourcing strategy based on the best-of-breed approach, including an independent benchmark of the cost to deliver the department’s IT infrastructure services.
In considering the viability of the outsourcing options (either a continuation of the existing MoU or a contract with a new service provider), the department was able to review its own experiences and the experiences of other agencies that were already operating under such arrangements.
Of particular importance was the level of responsiveness that could be guaranteed to any future changes in strategic direction. DEST’s executives could clearly see that, given the dynamic nature of the modern IT environment, there would be significant challenges in specifying contracts that allowed for sufficient flexibility over the longer term.
For comparison purposes the project team developed a detailed proposal for the option of bringing certain services back in-house. The proposal identified initial and ongoing costs and resource requirements, and outlined a comprehensive transition plan with an appropriate risk assessment.
But while the ongoing costs compared favourably with industry benchmarks and showed significant savings over the current arrangement, clearly the case for bringing services back in-house would not be argued on cost alone but would have to be based on the option that provided the best business value to the department. Could the same level of service be provided at a cheaper rate, or could a better level of service be provided at a rate that the department was prepared to pay more for? The in-house proposal argued that a better level of service could be provided while achieving an overall reduction in costs. With those key advantages in sight, the project team recommended the responsibility for the bulk of the services should be transferred back in-house. The plan called for the services then to be subjected to market testing exercises at regular intervals to ensure continuing value for money.
The data communications function was excluded at this stage because of the technical complexity involved in separating the networks. The project team recommended that the MoU arrangement be continued for the data communications and Internet gateway functions, with the option of reviewing these at a later stage.
While the issues of cost and quality of service could be argued and agreed on with some certainty, it was extremely difficult to reach consensus on the issue of risk.
Almost without exception, key IT advisory groups argued strongly against the department pursuing the in-house option. The primary concern was that the department would not be able to build the organisational capability required to implement a transformational change of this nature.
This was potentially a valid issue. The department did not have appropriately qualified people to take on the new functions and a major recruitment exercise would be required. But this concern was not enough to temper the determination, conviction and enthusiasm of several key staff within the IT group.
Positioning for a Clean BreakIn a happy circumstance of timing the IT group within DEST had already adopted some key strategies that were to have a significant bearing on the viability of bringing infrastructure services back in-house.
For several years it had been moving to consolidate all IT applications onto a homogenous mid-range platform, where previously, the department had operated a mix of applications running on mainframe and mid-range systems. This process of consolidation had been completed by early 2001 with all applications now running in a Windows server environment.
Second, a project to upgrade the department’s underlying network operating system from NT4 to Windows 2000 had provided the opportunity to separate all the DEST mid-range systems at the logical and physical levels. This project had been completed by January 2002.
Finally, a complete refresh of the desktop had been planned for mid-2002. Over a period of several months, new desktop computers were rolled out across the department with a software image that had been configured specifically for DEST.
A Two-Phased ApproachWith a clear strategy now in place, preparation for the transition began in earnest in August 2002.
The transition was divided into two distinct phases. Responsibility for the mid-range systems support would be transferred in November of that year. The help desk and desktop support services would then follow in March 2003.
Administratively, a new organisational unit would be created within the IT group comprising three sections, one to cover each of the new infrastructure functions. Each section would be led by a manager at the EL2 level.
The transition of the mid-range environment was recognised as a critical success factor for the overall strategy. These systems provided for LAN access, e-mail services, data storage facilities and IT security. At the time of the transition, the mid-range environment in DEST comprised some 280 servers that were geographically spread across several data centres.
In defining roles and responsibilities within the new organisational structure, the mid-range systems support function was partitioned as follows:
—server software image maintenance
Because of the specialist skills required to perform each of these functions, the mid-range section was staffed entirely by contractors, each having pre-eminent expertise in one or more of these areas. A manager with extensive experience across all areas was appointed to lead this team.
Finding a new facility to house the department’s fleet of servers was a priority and it was decided to upgrade the department’s existing premises to meet the standards required for a robust, reliable and secure data centre. In addition, a secondary data centre was constructed in a separate building to provide redundancy for some of the more critical systems and offsite storage of data backups.
The most expensive and time-consuming component of the mid-range transition was the design and implementation of a set of disaster recovery systems and procedures. New data storage systems were commissioned in the three months prior to the transition date to allow for sufficient overlap with the existing backup services that were being provided by DEWR.
With the transition of support for the mid-range systems complete, it was time to focus on the help desk and desktop support functions. As with the mid-range support function, it was important to establish clear roles and responsibilities.
The replacement for the help desk services that were being provided by DEWR would be known within DEST as the IT service desk, and would be the first point of call for most people requiring assistance with IT services.
By a simple pro rata of the calls being received currently at DEWR, the team assessed that approximately 200 calls per day could be expected by the new DEST team. It set up an automated call distribution system to manage these calls, which would then be logged by a new electronic problem management system.
The service desk team also assumed responsibility for providing a range of training services to help people become more productive in the use of the department’s IT systems.
The new desktop support team would be responsible for managing a range of devices within DEST including desktop PCs, notebooks, handheld PDA devices, printers and scanners, as well as providing second level technical assistance for problems that could not be resolved at the IT service desk.
A recruitment exercise conducted over a period of two weeks in February 2003 saw approximately 20 staff being selected to fill the service desk and desktop support positions.
The transition of services was competed as scheduled and on budget by March 2003. The main success factors were proper planning and the recruitment of good people. All up, approximately 30 positions had been created to support the new functions, with the bulk of these positions being filled from external sources.
Jamie Macgregor was the acting CIO in DEST for the period over which the transfer of responsibility for IT infrastructure services took place. He now works in the eBusiness Division of the Department of Industry, Tourism and Resources
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.