The reality of grey areas when working with cloud providers?
Joint management requires clear definitions, acknowledging that cloud services are complex and not often clear-cut.
When an organisation contracts services from a cloud provider, it is outsourcing its mission-critical applications to a company. It needs a partner that can ensure the availability, security and scalability of those applications 24x7. This is easy to promise – but hard to deliver.
What is needed? The key elements in delivering this support are the systems and experience of delivering a quality service in an environment with shared responsibility and authority between organisations and their clients.
We know cloud services are complex. Solutions are not only technical but require good management and good processes. The simple reality of offering cloud services results in many shades of grey. For example, a cloud provider is responsible for platform performance, but doesn’t control all aspects required to ensure that performance. While a customer will want its provider to resolve outages and problems, they’ll still require certain access and rights in the platform for its teams to operate, such as accessing the database and deploying new versions of the application.
This poses a conflict. When a platform fails, it might be an infrastructure or storage problem but it might be something from the customer’s side. Service providers need to recognise and address the issue of joint access to platforms, acknowledging the grey areas for they know that areas of unclear responsibility can hurt their customer relationships.
- Define responsibilities.
- Adopt best practices for working with customers.
- Work closely with organisations so that the way the two companies work together accurately reflects the situation and infrastructure.
A good company will adapt their processes and, if necessary, work to adapt those of their customers too so that neither party contributes to issues with platform performance.
All of this should be reflected in an ever-evolving Service Level Agreement (SLA). True service providers will be able to explain how their SLA has evolved over time – how they’ve worked to define what they guarantee, based on what they actually control, and what is required from their customers.
In my experience, service providers rewrite their SLAs about every two years so that they can incorporate what they’ve learned in the last two years about what constitutes a great service.
Some customers find such an SLA unattractive; they want one SLA that guarantees performance without any limitations and requires them to do nothing. This is unrealistic: an organisation can only guarantee what they control and a provider can only deliver a quality service if it works well with its customers.
I’d be wary of the blanket SLAs often offered by some of the new cloud providers. It’s easy to write promises on a piece of paper, but an SLA that reflects what can actually be delivered and guaranteed is in the best interest of all parties. A company that has previously faced the problems that result from joint management of a platform knows how to avoid issues, and how to bring its long knowledge and experience to each of its customer engagements.
It’s this ability to recognise the grey areas that naturally result from the customer and provider sharing operation of the same platform that will ensure customers get the service they require and deserve.