The cloud outages of recent times highlight just how import vendor and SLA management will become and also having the right legal frameworks to support these efforts. I have reviewed a number of agreements that vendors have handed to clients and have been somewhat concerned with the standard of some of the contracts and clauses. It seems in some cases that there has been such a focus on getting cloud product ready for market on the technical front that the legal side has been left underinvested.

A number of the agreements that I have reviewed were simply the same managed services agreements that had been developed some time ago, and these didn’t reflect the special nature of the cloud arrangements being proposed. In these cases what was proposed was a large scale enterprise transformation from on-premise to cloud services for everything ICT and not just the provision of some servers or storage. These are significant transformations with important milestones and joint responsibilities. In my view the vendor should make the investment in developing a solid legal framework that will both satisfy and impress the clients. And it will be reusable for future clients, why it may even become a differentiator.

I could blog quite a bit about the terms and conditions for cloud computing but in this case I want to discuss the one area of service credits. The consequences of failing to meet the uptime or availability warranty should be clearly set out in the agreement.  Some vendors offer service credits for outages outside of SLAs. These are often a percentage of the monthly charge for the service up to a maximum percentage.

Whilst a lot of vendors are not even offering service credits, those that do mostly provide a token amount and in a lot of cases these are hardly a disincentive.

Here are just a few questions that CIOs might ask about service credits and service management generally:

• If a vendor is quoting a lot of “9’s” in their availability and making the claim that it is safe then why aren’t they offering financial restitution if they fail to meet the SLAs?
• Why are there only limited service credits on offer and no considerable and sizable penalties for any downtime outside of the SLA?
• Equally why would the vendor not also consider bonus payments from the customer for achieving above the SLAs? This would act as an incentive to achieve higher service standards.
• To what extent has the vendor invested in people, process and technology that focuses on service management, service assurance and security?
• If the vendor experiences a failure in one of their cloud zones do they have a process for telling you as soon as it happens, which application is going to be affected, what the impact will be and which users are going to be impacted and when it will be fixed?

Of course the vendors aren’t able to or perhaps not interested in considering restitution, because they know that outages are very likely and that some outages are likely to be caused by something out of their control, such as an ISP. So for now restitution remains just an idealistic notion, however I cant help but think that the first vendor that does offer a serious restitution arrangement may well gain a leading market differentiation.

Tweet this article
Follow @Cloud_CIO