Service Credits in the Cloud

Consider the following example of a services-based agreement governing multiple service features with a five-year term.

The agreement subjects Vendor Co. to a number of individual performance service levels. The agreement provides for service level credits, as well as termination rights for material breach, abandonment of the agreement and “chronic service failure” (i.e., failure to meet service levels a certain number of times over a long period of time, say six months to one year). Over the first three years of the term of the agreement, Vendor Co. misses a few service levels each year and pays a few service level credits, but performance is generally acceptable. Vendor Co.'s performance then drops significantly, triggering Customer Co.'s right to terminate for chronic service failure. Although Customer Co. elects to terminate, it also determines it has a strong factual case for damages due to Vendor Co.'s performance failures. Vendor Co., seeking to avoid a public litigation, argues that Customer Co.'s damages claim is barred because Customer Co. accepted service level credits as liquidated damages. Customer Co.'s general counsel frantically reviews the agreement, tries to recall first year contracts law, and calls external counsel. Is Customer Co.'s damages claim lost because it accepted a few service level credits in the first three years of the contract?

Service levels and service level credits are routinely the source of heated negotiation in various types of commercial agreements with technology and services vendors, including rigorous debate about the very purpose of such credits. From the customer's perspective, although a decrease in payment liability likely does not fully compensate the customer for the particular harm suffered due to a vendor's service level failure, the credits create an incentive for the vendor to perform at the agreed-upon levels. That incentive is the customer's primary objective for service credits. Vendors may also have a stake in the service credit model, preferring a set of fixed and achievable performance expectations and consequences instead of an amorphous and subjective standard for service performance.