It is a contract that exists between customers and their service provider, or between service providers themselves. The SLA records the common understanding the parties have regarding services, priorities, responsibilities and guarantees that can be taken to represent the level of service itself. An SLA may specify the levels of availability, serviceability, performance, operation, or other attributes of the service such as billing and even penalties in the case of violation of the SLA.
SLAs have been used since the late 1980s by fixed-line carriers as part of contracts with corporate customers. More recently, IT larger enterprises have adopted the idea of using service level agreements with their customers – users in other departments within the same enterprise – to allow for comparing the delivered quality of service with the one promised and potentially consider the alternative of outsourcing IT services to an external company.
The technical specifications of an SLA are described through either a service level specification (SLS) or a service level objective (SLO). An SLS is intended as an operational guideline for the implementation of the service, and an SLO is a subset of an SLS containing some service parameters and goals to be achieved by the SLS.
Common metrics and penalties
Uptime agreements are another very common metric, used for data services such as shared hosting, virtual private servers and dedicated servers. Common agreements include percentage of network uptime or service availability. Figures such as 99.999% uptime are typical and service providers would face some form of penalty were they to drop below that figure if it were specified in an SLA. Generally, stepped structures for penalties are negotiated between service providers and customers under which missed targets incur fines. These are typically expressed in dollar terms and see customers receive either payments, reductions to their bills or credit notes towards future service consumption. However, customers now look for increased protection and often seek a “right to terminate” clause. This could be enacted if a certain number of service affecting incidents occur in a given month. Service providers are understandably reluctant to open up the possibility of a contract being terminated but customers feel justified in requesting such get-out clauses. Measurement of performance, however, is just one way in which service providers now conform to modern SLAs. Pro-active notification of incidents and quick resolution of problems are part of the mix and performance metrics for these will be included in many SLAs. It is worth noting that few customer organisations actually want to claim service level agreement failure penalties. They would far prefer to receive the service they specified at the price they agreed. In addition, in the case of a financial institution, for example, the penalties agreed may be insufficient to compensate the organisation for businesses losses entailed by the failure.
Penalties are not always expressed in cold cash terms and some providers offer innovative or quirky compensation packages designed to underscore their belief in the reliability of their services. One hosting company in the US reputedly offered compensation of a single one dollar bill – with the proviso that it would be hand delivered anywhere in the world by the provider’s chief executive.
So what is service assurance?
Service assurance is a procedure or set of procedures and systems designed to optimise performance. Service assurance is predicated on the concept that maximising customer satisfaction maximises profit in the long term. Service assurance can involve quality assurance, quality control and service level management. Quality assurance is intended to ensure a service meets specified requirements at all stages in the process. Quality control ensures that a service adheres to a defined set of quality criteria or meets the requirements of the client or customer.
Service assurance has emerged as the means by which service providers seek to address the quality requirements of users as well as the demands of pro-active notification and quick problem resolution. It also represents a means for service providers to differentiate themselves from their competitors – potentially reducing churn. The imperative for carriers to address not just intra- but inter-network service performance required by the delivery of complex services such as content served by third party providers, has driven demand for network management systems that address service assurance. Service providers now need to support carrier-to-carrier services, carrier-to-subscriber services and carrier-to-content provider service. These all require service performance and quality to be assured.
