|
|
|
|
|
Service Level Management - Best Practices Introduction
The following three factors stand out as key to the high performance of a service.
Three distinct
groups of responses were observed. Their differences are described below:
The average SLM rating for all respondents was 71. Solid Performers (42% of respondents) delivered on users' basic needs, and their performance on 19 of 22 factors closely followed the Stars (48% of respondents). Both groups scored high on availability status and trend, performance status and trend, had short recovery times and effectively managed their operations. The primary difference between Stars and Solid Performers was their implementation of "customer care" service based on three factors:
Low Performers (10% of respondents) did not meet basic requirements, and in general did not have significant SLM implementations - although some elements described above as customer care were present in several members of this group. We have inferred that the Low Performers are just that - not merely respondents discussing low priority services. This is based on our request that respondents "focus on what you consider to be the most mission-critical application(s) group that you service." As these are all mission critical services, we see stages of development in SLM implementation - and clear evidence that meeting basic user needs should be everyone's first step in the evolution. The elevation of the IT organization to business leadership is achieved from extending consistency and stability into high quality customer care.
Above Average Performing Industry Sectors:
Below Average Performing Sectors:
The above average group consists predominately of high-technology service providers, other professional services providers (e.g. facilities, logistics, etc.), large-scale healthcare organizations, healthcare services organizations, and insurance and financial services organizations. These business types are generally characterized by a high reliance on information. Hence their information technology organizations are not a business adjunct, but rather a significant part of the core business. They have high SLM ratings (71 to 78 points) and the group's leading scorers are well on their way to implementing customer care. They are managing their information assets as a business function, and are managing their relationships with their end users. Most of the Stars are from these sectors. The below average performing sectors (65 to 70 points) have demonstrated limited to no customer care implementation. They are generally meeting basic user needs. These businesses are also dependent on information technology but also tend to manage large physical assets. Information technology is more frequently a business adjunct than a core business. Generally the Solid Performers are prevalent in these sectors. It must be noted that the preceding statements are about tendencies. Certainly there are high performing distributors in terms of SLM and user satisfaction. Likewise there are low performing services organizations. However the industry relationships discussed above are generally statistically significant. It is clear that the basic availability and performance needs of users must be met for user satisfaction to be high. Service Level Management, and more specifically, its customer care aspects, are a leverage point to a higher level of user satisfaction. One of the many questions left unresolved by this baseline study is the causative relationship between the elements. That is, if you attack basic availability issues first, and then address customer care are you any better off than using the converse strategy or a parallel strategy? The research only suggests that there is a hierarchy of needs that must be met for high satisfaction and that our respondents tend to cluster around "stations" on that need hierarchy. There are strong indications that implementations are clustered in their self-assessment of SLM, inferring that there are stages of SLM development and that success (as measured by user satisfaction) is achieved in steps. Industry comparisons point to a relationship between user satisfaction, the importance of information to the organization and the steps taken to manage information as a business asset. The study sponsors have found from their experiences in SLM service engagements that their clients tend to follow four distinct stages of evolution in Service Level Management.
Service Level Benchmark - Key Findings
Over 60% of respondents reported having service level agreements in place for the particular service on which they reported. Generally, these agreements were broad, including definition of roles and responsibilities, goals, reporting polices and support availability. A significantly smaller percentage reported performance penalties or incentives. About half included procedures for adjusting to changing conditions.
Over half (62%) of the respondents with agreements in place provided reporting based on aggregated indicators of infrastructure or components. Thirty-six (36%) percent reported having automated reporting systems that provided customized reports to users. Twenty-two (22%) percent of those with agreements did not provide reports to end-users.
Generally, respondents felt that their problem documentation was somewhat effective or very effective. Only 8% felt their documentation was ineffective. Key reasons why documentation was not very effective included lack of sufficient details, ineffective use of documentation, and out of date or difficult to access documentation.
Respondents reported availability as meeting user requirements 92% of the time. Only 5% reported declining availability. Similarly, 87% of the respondents indicated performance as meeting requirements and 11% reported performance as declining. Use of staff resources to support availability and performance issues was relatively low. About half (45-55%) of respondents reported less than 20% use of their systems administrators', applications administrators' and database administrators' time to support availability or performance issues. Only 5% of the respondents indicated that 60% of their administrators' time was engaged in availability and performance issues.
Infrastructure implemented to support high service levels was prevalent. Approximately 70% reported a high availability server configuration or a parallel sysplex, more than 50% mirrored disk and over 42% reported network failover configurations. Only 11% reported no specific high availability infrastructure features.
Respondents were very positive about their end users' satisfaction. Only 17% reported neutral or dissatisfied users. In general, user satisfaction was positively related to all elements of sophisticated service level management. Fifty-six (56%) percent of respondents reported having a formal user satisfaction measurement program. Respondents with formal programs were more likely to report higher user satisfaction than those without a formal measurement program. Respondents and Sites The 182 respondents came from a broad spectrum of industries - dominated by commercial, financial and services sector organizations.
Sixty (60%) percent of respondents were from sites with over 20 terabytes of on-line data. Thirteen (13%) percent had over 80 TB.
Approximately 50% of the sites had more than 7,000 users, and half of those had more than 15,000 users.
About a third (31%) of the respondents were in technical management roles, 20% were in applications management roles, and 15% were IT executives.
This survey was conducted by CustomerSat.Com using a web-based approach. The study was sponsored by BMC Software, Sun Microsystems and PricewaterhouseCoopers. |
sponsored by
|
||||||||||||||||||||||||||
| Copyright (c) 2000-2003, nextslm.org. All Rights Reserved. Legal Statement. | |||||||||||||||||||||||||||
