Service Level Agreements

Background

A key objective of the AIMS project was "to ensure appropriate quality and efficiency in the provision of services, and to determine the best location of services in a devolved structure". For services to be of the appropriate quality and efficiency, there needs to be consensus between the internal supplier of a service (typically a support department within the University) and the internal customer of that service (typically a Faculty) as to what services are required, and to what level of quality these services should be provided.

Debates on these aspects of service provision had not occurred routinely within the University, and lack of communication had led to acrimony between the support departments and ther internal customers. Internal customers complained about inappropriate and poor service, while support departments were demoralised and disempowered by the negative messages they received. The cycle underpinned a prevalent sense of "us" and "them" within the University and weakened alignment between support services and the core business of the institution.

The Organisational Design & Governance report highlighted poor alignment not only as a problem for relationships within UCT, but also as a constraint on overall institutional performance. It recommended

• An agreed system of institutional performance criteria and measures (linked to the budget as necessary)
• The establishment of Service Level Agreements (SLAs) as the basis of negotiating, pricing and measuring the provision of services to the Faculties and central administration
• The establishment of a Management Services Committee (MSC) to set the generic terms for SLAs, monitor routine performance and deal with major failures in agreed service levels
• Linking of issues arising within the MSC and relevant to Internal Audit, via the Vice-Chancellor.

These recommendations were noted by Council, in adopting the ODG report in November 2000. Council commented that implementing SLAs was likely to be a complex exercise, requiring effective management coordination.

^ Top

Functioning of Service Level Agreements

SLAs allow internal customers to state their service level needs clearly and unambiguously. In turn, suppliers are enabled to focus their resources into providing services at the required level. Key performance indicators (KPIs) measure the quality of the service provision to allow for timely remedial action where necessary.

Where services, over time, are not delivered to expected levels of quality and efficiency, two courses of action can be taken. Either the service capacity can be increased (often accompanied by a cost increase), or the customer's expectation as to level of service can be decreased. This decision should be taken transparently, so that both customer and supplier are aware of new expectations and/or service targets. Any additional costs would also have to be negotiated. Options for closing the service gap are illustrated in Figure 8A.

Fig. 8A

^ Top

Approach for Developing SLAs at UCT

The objective of the AIMS Service Level Agreements workstream was to build exemplar SLAs and, in so doing, build capacity within UCT to establish and manage SLAs across the University.

In October 2000, five key areas within five support departments were identified as learning cases. It was envisaged that SLA project team members involved in these exemplar SLAs, would continue to play a prominent role in the ongoing establishment of SLAs in their own departments. They would also participate in training staff from other departments to apply the SLA methodology modelled in the AIMS exercise.

The five exemplar SLAs chosen were:

• Services provided by the General Ledger section of the Finance Department
• Campus Protection Services (to measure the performance of both security service providers: CPS and Gray Security)
• Central Academic Administration
• HR Administration services around the Terminations process
• Desktop hardware and software acquisition and installation services, provided by the Information Technology Services Department.

A generic process was followed in compiling the exemplar SLAs, comprising the following steps:

• Compiling a dictionary of services in the area specified
• Conducting a workshop with the supplier and establishing possible KPIs
• Conducting a customer workshop and validating the proposed KPIs
• Compiling a draft SLA
• Agreeing the SLA in a "reconciliation workshop" with supplier and customer
• Signing the SLA
• Drawing up an implementation plan.

Once completed, the exemplar SLAs would have to undergo formal ratification by the Management Services Committee, as recommended by the ODG Report. In some instances, finalisation with other relevant governance structures would also be important. It was envisaged that some of the exemplar SLAs could become part of a larger SLA encapsulating a bundle of closely related services. This was likely to be the case, for example, with the Terminations process, which is just one of several related HR Administration service areas.

^ Top

Exemplar SLAs

Exemplar SLAs were grounded in some key principles:

• Internal transactions and agreements between internal suppliers and customers should be seen as part of a larger process whose final participant is the external customer
• Internal customers have a responsibility to articulate needs clearly and precisely, to recognise the cost impacts of requirements, and to clearly distinguish demands from wishes
• Internal suppliers have a responsibility to ensure that the University aspires towards an agreed level of service and to ensure that agreed service levels are delivered to the customer
• Defining customer-focused KPIs is a critical first step in measuring performance
• The SLA needs to cover certain basic elements including: participants in the agreement, an agreed review process, services to which the agreement applies, how service provision will be measured, customer obligations and supplier constraints. (These elements are illustrated in Figure 8B.)

Fig. 8B

By December 2000, progress had been made in all five of the exemplar SLAs:

• General Ledgers
  • Key services on offer, as well as high-level KPIs, customer obligations and supplier constraints had been identified.
• Campus Protection Services
  • A service dictionary had been compiled, and these services needed to be grouped and divided between CPS and Gray Security. Some progress had been made in beginning to define KPIs.
• Central Academic Administration
  • A service dictionary and some high-level measures had been compiled, and activities and measures were to be detailed behind each of these services. Attention was being focused mostly in the areas of administration of doctoral degrees, and production of HEMIS student data for submission to the Department of Education.
• Terminations process
  • Services, or outputs, for an effective HR Terminations process in such instances as resignation, retirement, disability and retrenchment, had been identified, with corresponding KPIs. The process related to deaths and ill-health retirement had also been investigated.
• Desktop hardware and software acquistion and installation services
  • Services on offer to customers had been validated and KPIs identified. The reconcilation workshop would need to resolve conflicting customer and supplier points of view around choice: for example, while customers wanted to acquire computers as cheaply as possible in light of their specific needs, suppliers felt that choice without standards would jeopardise connectivity.

Readers interested in seeing more detail on the developing exemplar SLAs are referred to the Progress Report to the AIMS Steering Structure in December 2000.

The exemplar SLAs still had to be finalised, and it was apparent that each case was likely to differ in accordance with the specific needs of the area under investigation. In some instances, the work was proceeding relatively slowly, as the work of developing SLAs was revealing problems in the process underlying service delivery (ultimately, such process "fixes" would, of course, be beneficial).

In December 2000, the SLA workstream suggested a process and timeline for the overall way forward in 2001.

^ Top

The Way Forward

AIMS' role in the SLA process came to an end in December 2000. The Registrar then became accountable for taking the process forward, with the following supports in place:

• Members of the AIMS SLA team were to become a "core SLA team" with the following roles:
  • Finalising the process of compiling the exemplar SLAs
  • Facilitating compilation of remaining SLAs in their respective support departments
  • Providing initial support and training to new "SLA resources" in the support deparments
  • Maintaining a forum where SLA resources could meet to exchange ideas, learn, coordinate and integrate activities.
• SLA resources in each support department were to be identified early in 2001, and thence to form a "greater SLA team", facilitating the compilation of SLAs in their respective areas and establishing an SLA capacity in the University.
• The Management Services Committee (MSC) was to approve at its inaugural meeting in 2001 the SLA process and templates modelled by the AIMS work, as well as the SLA implementation timeline, and proposed customer signatories for SLAs in the University. (Figure 8C illustrates relationships between support departments and possible customer signatories.)

Fig. 8C

Progress in SLA implementation was unfortunately delayed, as the MSC met for the first time only in May 2001. At this meeting it reviewed "external" SLAs already in place for the pension fund, and between Properties & Services and Gray Security. It also specified those areas in which UCT's support departments should develop SLAs, with a focus on developing agreements in areas where service deficiencies existed. Heads of support departments were asked to draft such SLAs for MSC review within three to four months.

^ Top

Team

^ Top

Links to Additional Material

• File of Printable Figures on this page (pdf 19 kb)
• Progress Report on Exemplar SLAs to AIMS Steering Structure, 1 Dec 2000 (pdf 208kb)

To navigate from this page to additional pages on the AIMS website please make use of the buttons in the navigation sidebar near the top of this page.

^ Top

Last updated: 2001-09-14