- Service Level Agreements (SLA) are the intersection between the provisioning of bulk data center services at lowest cost, and the measurable (and unmeasurable) business service expectations. Data centers are structured to provide lowest-cost, reliable infrastructure services, repeatedly.
- Data center managers are being asked to agree to SLAs with the business units, which have stated expectations beyond the structured delivery of delivered services.
- Most internal data center SLA efforts are focused on the failure of services. This creates an SLA process that is oriented towards penalties after the service failure rather than continuous improvement to prevent service failures.
- Tension between traditional SLA wording and business expectations results in most internal data center SLAs being ineffective at providing a framework to facilitate a dialog about service delivery.
Our Advice
Critical Insight
- An SLA in its simplest terms provides a definition of performance for a service, so the provider and customer have clear expectations within the relationship. But in reality, the key to a sound SLA goes beyond strict performance measurement of quantitative metrics.
- While 60% of organizations do NOT have a formal data center infrastructure SLA implemented, they agree that the business has informal service level expectations. SLA or not, organizations that are successful in managing service levels have a formal process in place to measure, monitor, and report service levels.
- This solution set will provide guidance on the governance of business alignment through the use of an SLA. It provides actionable tools and insight to help the data center through phases of SLA management.
Impact and Result
- Ensure that an SLA is necessary before making the commitment. Consistent service levels can be accomplished without a handshake with the business.
- Identify and implement metrics that match service level and business alignment goals to maximize effectiveness. Create reasonable targets that are based on historical service level data, to avoid making promises that can’t be kept.
- Conduct a quarterly review meeting within IT to review and update performance targets or the SLA.