A service level agreement or SLA is a formal document that defines a working relationship between parties for a service contract. It is generally more applicable to businesses than to consumers and involves one or more end user parties and a service provider. An SLA does not define how the service will be delivered but rather provides a measurable framework for a service. Typically, an SLA could be developed with Internet and broadband service providers, web hosting, cloud computing and outsourcing arrangements with companies such as managed services providers. Some metrics that are used to define the SLA might include items such as the mean time between failures, an agreed accepted level of downtime across an agreed time frame, support response and resolution times.
In most cases with telecommunication services, the service level agreements are usually made back to back with their suppliers who are carriers. Working with providers who are wholesalers of telecommunications infrastructure means that the actual service level will be different from service to service. For example, whilst an end customer may have financial negotiation strength as a whole with a provider, if low cost telecommunication services are procured for some sites, it is unlikely there will be a consistent business wide SLA, as each service will come with its own SLA back to back with the carrier.
In essence, the SLA is some sort of guarantee of performance, negotiated between the customer and the service provider. But to make the service level agreement actually meaningful it needs to be measurable. Service level agreements are basically pointless without the ability to test and collect service metric data. In some cases, it makes sense to include a service performance reporting mechanism where the performance of the service is measured.
Exact metrics of a service level are different among providers. Additionally they are different for individual services acquired by the end customer. The SLA commonly defines the responsibilities of the service in regards to availability and performance. Some of the performance metrics include:
Agreed service availability.
Agreed service performance metrics.
Agreed support availability.
Agreed support performance metrics.
Clearly defining the service levels expected of the service provider helps both parties eliminate the grey areas often involved in the provision of services. The SLA defines the level of service expected. Within a service level agreement, it’s possible to include financial incentives for exceeding goals and penalties for failing to meet a certain level of service. Typically, telecommunication services are set with penalties for the provider in the event that they are outside the SLA agreed metrics.
Whether you are using a service provider for your broadband, private IP network, managed firewall, help desk support or cloud computing platform, each service provided by the third party is an independent service. It means that each service would actually require its own SLA with its own terms of service plus metrics negotiated on a case by case basis.
Having meaningful service level agreements in place helps to define the services and the performance that is expected from third party service providers. In most cases, the penalties or rebates that accompany the SLA are more trouble than they are worth. The uptime, availability and performance of the service procured is far more valuable than the rebate.
Most service providers will promote the fact that they have service level agreements in place for their services however, few will have the ability to effectively measure their service levels and even less will proactively provide service level performance information on a proactive basis. A well designed, implemented and maintained service level agreement can make a profitable investment made with a third party service provider.