Pretty much all outsourcing contracts in the IT Service Management world rely on, or at the very least, utilise the Service Level Agreement (SLA).
Certainly they are important as they are the physical representation of performance of the contracting party and used as the measure by which trends in supplier performance is understood.
But is there too much reliance on SLAs as a measure of performance and are they often inserted by the eager contract or procurement manager to mitigate risk or provide a means for the insertion of penalty / reward clauses because “that is what is expected in a contract”?
In my personal experience SLAs are often poorly defined or their alignment to the realities of IT service delivery misunderstood. Because of that, there has been many mitigating circumstances offered as to why an SLA has been failed by the contracted organisation, followed with significant discussion as to whether the mitigation can be accepted. This has a tendency to suck up both time, effort and therefore money, from both organisations into managing the performance measures, drafting contract change notices and often not looking at the root cause of why SLAs are being missed or in one case I have dealt with perpetually exceeded (plainly in that case the SLAs were too generous or measuring the wrong outcomes).
The vendor will look to win the opportunity and subsequently concern themselves with making the delivery side work (especially when the bid team is not going to be involved in delivery). They will obviously try their utmost to meet the targets set, but also expect to provide mitigations in the event of failed SLAs. They have the experience of dealing with a number of clients and so have reference points to support them, whereas the contracting organisation does not have the same level of experience or number of reference points.
A common resolution is the instigation of a continuous performance improvement plan, and when that has been met redrafting of workable SLAs agreed by both parties, or if it fails penalty clauses or litigation.
Poorly defined requirements from the customer. Either they are unsure what they want, have over / under specified the level of service really needed, they are looking to outsource a problem, or the business units have been poorly engaged, if at all, by procurement through the tender process. In such circumstances the supplier is almost being set up to fail from the outset (which from their experience they will probably realise) and therefore they will look to manage their way around the issues as they arise.
The common resolution is a redefinition of the SLAs probably with an element of contract renegotiation once the customer has determined the service it expects or requires.
So what should a good SLA really be about? A well constructed SLA should be seen as an important measure to support a positive contractual relationship, it should also be periodically reviewed for its applicability in light of changing business demand. However, the SLA should not replace or overshadow the development of the relationship between the customer and supplier. Rather, the SLAs in place should support a collaborative attitude towards delivering a contract outcome that benefits both parties; the customer receives the service they need, the supplier makes the profit margin they expected and the customer is satisfied with.
Neither party should be wasting time and money negotiating mitigations, instead the time saved can be spent on delivering future value. Unfortunately developing proper, mutually beneficial collaborative relationships in a business environment is not easy where customer and supplier aspirations are not aligned.