After a comfortable flight we arrived safely in Las Vegas, where the 23rd floor of ‘The Mirage’ -with Strip view- will provide us with some good night’s rest to be able to absorb all that re:Invent 2017 has to offer. With 50.000 visitors (twice as much as the largest VMworld USA conference) the event has completely sold-out, which makes us already wondering what next year’s event will bring in terms of location and logistics.
An AWS view on People, Organization and Operating Model Transformation
Being wide-awake early as a result of a small jetlag, we started our first day with a workshop that we did not want to miss: People, Organization and Operating Model Transformation, hosted by AWS Professional Services. AWS agrees that technology is not the challenge anymore; a successful transformation depends heavily on an organization’s ability to (1) start on the right foot, (2) design new operating models for cloud and (3) execute proper organizational change management.
Use Application Portfolio Analysis to get started on the right foot
When it comes to ‘starting on the right foot’, AWS mentioned using Application Portfolio Analysis as the major driver to determine where and how to get started with cloud. Think for example about the strategic benefits you are trying to achieve with the transformation to cloud (innovation versus cost reductions).
Funny detail was that those applications that lack ownership, have a very low rate of change and do not really provide added value were referred to by AWS as being ‘craplications’. Definitely a term that we are willing to adopt!
Design Cloud Operating Models that fit the Application Portfolio
With regard to Operating Models, AWS presented the Four Box Approach which they use to help their clients visualizing and building Operating Models for cloud.
Although we believe that this Four Box Approach is too simplistic in some areas, it turned out to be a powerful model to kickstart the design process. According to AWS, in the first period of moving to cloud in particular two different Operating Models apply:
- DevOps: is most easily described as "build it, run it", whereas the application and infrastructure is built and operated by the same team that are usually organized by product or outcome.
- Traditional: is most easily described as "operating applications and infrastructure separately" and is usely performed by a shared services or central IT Organization that is organized by function, not by product or outcome.
Later on, we'll talk some time to reflect on these different Operating Models proposed by AWS.
Establish a "two-pizza" Cloud Team
What the Operating Models have in common, is that they include a ‘Cloud Team’. According to Jeff Bezos, a Cloud Team should not be bigger than you can feed with two pizza’s. We thought Americans to be big eaters, but apparently this means that a Cloud Team should consist of 8 to 10 people! A Cloud Team is:
- Hands-on (an advisory role alone is not enough)
- Co-located and dedicated (and preferably not scattered around the globe)
Typical roles that you may find in a Cloud Team are:
- Product Manager (responsible for the Cloud Backlog)
- Enterprise Architect
- Application Engineers
- Security Engineers
- Infrastructure Engineers
Later on, we further elaborate on these roles and responsibilities and how to address scaling the Cloud Team.
Execute organizational change management
The workshop was concluded with a comprehensive model for executing organizational change management, which is a key ingredient for successful cloud transformation.
Want to learn more?