CIOs spend a lot of time negotiating IT services agreements. Traditionally these have been for the acquisition of services such as consulting or development services, although more recently with Cloud Computing (formerly known as Software-as-a-Service or ASP) CIOs have been confronted with services agreements that relate to more abstract concepts such as software functionality or remote hardware capacity.
By their nature these services contracts often have broad descriptions of the services required to be provided under them. For example, a hosting agreement might provide that the ISP is to host the customer's Web site, or a desktop services agreement might oblige the provider to make desktops available for use during the term.
All well and good, but the expectations of both CIO and supplier when it comes to performance of the obligation to provide the services may not always be aligned. What does it mean exactly to say that a Web site will be hosted? Does that mean that it will always be available, or that some outages are acceptable? When will the supplier be in breach of the agreement?
While services agreements often contain warranties as to service quality, specifying service levels with respect to certain aspects of the services provides the parties with certainty as to the level of performance required with respect to those aspects of the services. To be effective, service levels must be:
(a) important to you; and
(b) capable of measurement.
Outcomes based measurement
Service levels are often outcomes based. For example, in a hosting context, as well as the obligation to provide the hosting services described in a schedule with due care and skill or to best industry practice the supplier might also have to ensure that over each specified period:
(a) the Web site is available a certain percentage of the time;
(b) the Web site has no more than a certain number of outages;
(c) the supplier responds to an outage within a certain timeframe; and
(d) the supplier resolves an outage within a certain timeframe.
Including service levels of this nature clarifies the expectation of the parties with respect to the aspects of the services that are subject to the service levels. To be meaningful the service levels have to be stated as being measured over a certain period e.g. to say that a Web site is to be available 99 per cent of the time is meaningless without an associated timeframe such as each month.
Although service levels are often technical and need to be drafted by experts in the relevant field, it is important that they be clearly expressed. Lawyers can assist by asking if this is the case, and particularly drawing attention to terms such as ‘availability' and ‘outage' , and what constitutes a ‘resolution' of an issue. For example, is a Web site available if it is operating at a reduced capacity, or can only accept a limited number of concurrent users?
An ‘outage' is generally defined in terms of something not being ‘available' (however available defined). Similarly, a problem may be resolved if the item that is the subject of the service level is available following an outage. Definitions for service levels have to be considered with as much care as any definitions in a binding agreement.
Join the CIO Australia group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.