The Road to the Cloud

Recently I was asked to reflect on what kinds of cloud technology projects I expect to see in the next eighteen months , and what kind of skills will be required to deliver those projects.

After having a lot of conversations with different companies around their cloud roadmap I think we can classify the current phase as “getting ready”, rather than “fully committed”. So, particularly in the infrastructure space, the types of projects and engagements I expect in the next twelve to eighteen months are not that different from what we have been focusing on the last two years or so.

In order to get ready to move to cloud services many organisations need to clean up and consolidate what they currently own. Directory consolidation, server virtualisation and upgrading to the latest platform are certainly critical to help position for a jump to the cloud. I'm also getting the feeling that customers are willing to trade off some of the functionality they currently enjoy for simpler and cleaner technology environments. This is a real fundamental shift from the way that IT departments have deployed infrastructure in the past.

I am also talking with several companies who look to the cloud for solving the integration challenge of mergers and acquisitions. I expect organisations to get their toes wet with cloud services by migrating smaller and more autonomous divisions or companies within their organisation. The key to success will not be having a small department running in the cloud, but it will be measured by the success of follow-on integration projects. The problem is that these small projects will probably not have the time or the resources provided to ensure successful future integration efforts. For large customers that go down this path the crucial engagement will be defining the overall architecture of the end state.

I am also having a lot of conversations about the idea of a private cloud. In the minds of most IT people the concept of the private cloud is synonymous with virtualisation. This has a lot to do with marketing strategies of virtualisation vendors, but it is also a failing on the part of technology leaders to clarify what cloud infrastructure really consists of, irrespective of whether it is hosted on premise on in a vendor's data centre. Once again it will require architecture skills to define a private cloud strategy that is not simply a virtualisation migration exercise.

Finally companies see the cloud as a way of achieving selective sourcing. The whole of IT approach to outsourcing was not very successful in managing costs, and had the significant disadvantage of hampering flexibility. The cloud offers a way to develop a services-based approach to selective sourcing. What will be required by customers are projects to decouple existing systems and processes, and to begin to define very clear service catalogues and models. The engagements that will come out of migrating to new sourcing models will be primarily operations process engagements.

In this new world the required skills are clear, they are primarily architecture and operations skills. Restructuring and decoupling existing processes is clearly a major activity that is ideally suited to operations skills. Defining a clean services models and architecting an environment which does not limit future cloud migration and integration efforts requires very high levels of architecture skill and experience. In fact the architecture profession will need to step up in a big way and start to develop some real discipline and maturity.

As for technical skills, there will certainly be opportunities in the short to medium term for experts in identity management, security and virtualisation. However in the long run the move to the cloud reduces the need for straight technical skills, and increases the need for architecture, operations and business expertise. The cloud is simply another service delivery option, which must be integrated into a cohesive technology strategy. Life is not about to get any easier.