The Impact of Service Performance Patterns on the Design Elements of Service Level Agreements

With businesses becoming more and more dependent on IT services, business operations are highly depending on the stability of IT services, and the predictability of IT service performance. The Service Level Agreement (SLA) is a core document that prescribes service performance over time. The Service Level Management (SLM) function designs the SLA and uses the SLA for defining the terms of inter- and intra-firm service provision. SLAs are a procedural IT governance mechanism that moderates business-IT alignment. Consequently, SLAs are an essential tool for managing IT customers’ expectations of IT performance and value. Firms often use one standardized SLA design for all IT services included in their portfolio. Research has further contributed to the development of standardized SLAs. Assuming that IT service performance varies across services, we must, however, ask how those performance patterns vary across IT services. This discussion adds to our understanding of critical design requirements for service performance definitions and metrics. Both performance definitions and metrics are the core element of every SLA. The objectives of this research are to test if IT services have varying performance patterns, and if so, how IT service design characteristics impact those performance patterns. We hypothesize that IT services perform contingent on IT service design. Contingency factors include technical, process, and human
resources characteristics. Through a case study approach, we analyzed quantitative and qualitative IT service data for one year. The results of this study are a classification of performance patterns. We discuss the managerial implications of performance patterns and present metrics for each pattern that can be used in SLA design.