7 Key Attributes of Strong Service Level Agreements

When outsourcing managed services, well-structured Service Level Agreements (SLAs) help ensure supplier and business interests align to achieve your desired outcomes at a fair cost. Every agreement should have strictly defined contract terms, which commonly include:

• Scope of services
• Responsibilities of each party
• Charge calculations for services
• Allocation of risk
• Legal recourse in case of breach

Another crucial but often neglected function is determining how service quality is measured – and consequently, any penalties suppliers incur should performance fail to meet expectations.

Performance targets are typically left undefined well into the RFP process. Further, suppliers can often suggest service metrics as part of their bid submissions. Leaving these vital metrics to suppliers is a sub-optimal quality control strategy that leaves the door open to efficiency-sapping disruptions, such as:

• Solutions that fail to meet the actual needs of downstream IT or business functions
• Paying more for over-engineered solutions with unnecessarily high performance levels
• Unfair distributions of risk, leading to legal complications in the event of breaches

Precise and well-structured SLAs eliminate such issues before they arise by clearly defining delivery expectations. They also provide the data for proactive measures to continuously improve performance, as well as corrective actions for in-flight errors.

Let’s examine 7 key characteristics of effective SLAs and how they can streamline the delivery of managed services.

The design of an SLA agreement must be associated with the risk and business impact caused by successful value delivery and projected sub-optimal performance. Engaging service recipients to develop these metrics ensures they understand the consequences of a missed service level, enabling them to make decisions and respond appropriately.

An SLA agreement must consider the audience that will track and make decisions based on captured performance data. A common mistake is creating service level reports that only provide a partial view of services delivered to end users, leading to confusion and conflicts.

Clients should develop comprehensive visibility of performance targets, how they are affected by the service provider and other metrics, and the impacts on end-user experiences. Achieving this may require two different sets of reports:

• Service level reports that measure service provider performance
• Service level reports that measure end-user experience of delivered services as a whole

An IT outsourcing SLA should reflect the precise responsibilities and risks assumed by each party. For example, a client may be wholly responsible for ensuring effective end-to-end application performance while server and storage functions are allocated to a service provider.

Poorly designed, overly general service levels leave overlapping responsibilities and gray areas that can lead to conflict. In this example, holding a data center services supplier accountable for application performance can mitigate supplier risk by allowing them to finger-point, leaving the client with the risk of limited resources should a major service issue arise.

Service levels should not be subject to subjective interpretation. They should be measured and reported based on factual data. Data sources and formulas used to calculate metrics should be as sharply defined as possible – especially in situations where tracking involves correlating data from multiple sources.

Service levels must be aligned with the responsible party’s contracted capabilities. Commitments should be based on minimum acceptable performance thresholds, not aspirational targets, to prevent unrealistic expectations from derailing delivery. IT departments should design service levels to align with end-user expectations, and publish the service commitments to manage them after.

The value of information produced from tracking service levels should always supersede the cost to measure, report, and track it. Otherwise, it represents a source of value leakage that should be eliminated.

It is a significant investment to support real-time reporting of complex service levels, and providers must both quantify and justify the effort required to track them so clients can determine if the view is worth the climb.

Deploying over-engineered solutions should be avoided. Options like slickly designed real-time dashboards offer excessively high performance for their intended purposes, and only amount to additional OPEX costs.

SLA information should always lead to action. Actions can span several functions:

• Preventive – Getting ahead of a negative trend
• Corrective – Fixing performance issues
• Proactive – Defining performance thresholds that could worsen risks if breached

Though these 7 characteristics are good rules of thumb for SLA design, the specifics of effective SLAs change with the state of your business enterprise, business objectives, and operational needs. Consult expert advisory to clarify these unique variables and formulate airtight SLAs to guide your managed services strategy.