One stumbling block to implementing service management initiatives in government is how to go about crafting effective internal and external service-level agreements (SLAs).
In government shops, even though service commitments exist between IT and publicly funded agencies, those SLAs are generally considered internal to the state or government operation. External SLAs are agreements with outside service providers used to outsource some aspect of IT functionality.
Internal SLAs must be measured and monitored in the same way as external SLAs. The difference is gaining access to real metric information collected by third parties, though this is becoming easier as toolsets have matured. In addition to monitoring and measuring SLAs, addressing consequences of degraded service is also quite different for internal versus external SLAs. Internal SLAs can be difficult to enforce, especially in some government organizations that might have limited funding to guarantee resources for adequate IT infrastructure investments. External SLAs, on the other hand, must be monitored proactively to make sure expected value is being delivered — a critical factor when spending taxpayer dollars.
Service level management (SLM) seeks to tie the organization’s objectives to IT and then to measure those objectives in a tangible way. More than other areas of technology, this forces IT shops to understand the business requirements in detail before considering any SLM solution. Services in state and local government might involve supporting the distribution of food stamps to needy families, processing vehicle registrations and tax bills, or providing access to police records. The nature of these activities differs greatly from commercial operations, and many are personal in nature. Service quality is perhaps even more important than in other endeavors
This year, 82 percent of respondents report they’ve implemented SLAs, either internally or externally, or both. Government organizations are right in line with other industries, with 75 percent saying they have implemented internal or external SLAs — 25 percent with external SLAs. Most government shops either rely on service provider data (37.5 percent) or do not measure (50 percent) service performance at all for these outsourced services. This finding stands in contrast to the majority (75 percent) of government entities that indicated it is important to govern outsourcing relationships.
For many years, educating organizations has been the top challenge for SLM. In Enterprise Management Associates’ 2007 research, this remained the case for all vertical segments including government responses, with 37.5 percent concurring with this assessment. Other SLM deployment challenges for government respondents were integrating toolsets to measure SLAs (25 percent), articulating SLA language (25 percent) and securing funding for SLM initiatives (12.5 percent). Most verticals show similar results.
The interesting contrast in the government group arises when comparing challenges to priorities. While educating the organization is the biggest challenge, it is not the top priority. Defining and prioritizing services, articulating SLA language, integrating toolsets for SLA measurement, and instituting service costing were all higher priorities. This is interesting and suggests that government groups are trying to approach service management from a number of different perspectives — taking on the very difficult deployment tasks at the same time they are articulating the need to government employees and leaders.
Can the SLA itself get in the way? Perhaps. And that’s what the Alan, IT service portfolio manager with the State of Colorado has been contemplating. Ashurst is responsible for service management in a newly created, cross-functional role responsible for managing service deployment and quality.
At the state of Colorado, the IT Infrastructure Services group consists of approximately 250 staff members. Its customers are state agencies comprising around 22,000 state employees. Ashurst’s group primarily supports two major applications that are used by a variety of state agencies — statewide financial accounting systems and statewide human resources. The Division of Information Technology also handles infrastructure support for systems, network, applications and so on.
In Colorado, Ashurst is driving an effort to move away from formal SLAs in favor of a service catalog. He has been working with consultants to address ineffective SLAs that have been in place at the state. In his organization, he believes that it is better to demonstrate to his customers (agencies) what is available and what to expect for quality. Formal agreements have been sitting on the shelf with unenforceable consequences for service degradation.
A catalog of services will enable customers to see and order appropriate services, and it will give them access to a real-time system that reflects current capabilities. The catalog itself will be the representation of the SLA negotiation and commitment. Perhaps this approach will become the wave of the future, especially when contemplating internal SLAs. Ashurst’s movement away from SLAs is by no means a step away from service quality, but rather it’s a strategy for making service quality commitments more meaningful in his organization.