This step involves the analysis of the architecture to extract dependencies between components, for example that a data entity is central to a particular process, or that a process/services deployed to a particular location. Dependencies are only noted if they will play a part in defining the migration strategy, that is, they must be dealt with during implementation of a new component.

It is important to note that the focus of this dependency analysis is the architecture and it's structure, not implementation dependencies (this is done in the next step). Also, both the exiting architecture and the target architecture dependencies are identified.

Capturing this information in a way that is useful for the next steps is an issue since the picture can get quite complex. Two approaches are possible:

1. Network (Location) Centric - this approach uses the location or network view of the architecture as the common thread for organizing the dependency information;
2. Process Centric - this approach uses the business process flow as the common thread for organizing the dependency information.

Either way, the information is extracted and summarized twice, once for the existing architecture and once for the target architecture.

This step involves the analysis of the architecture to extract implementation dependencies between components, for example that a data repository is needed before a process or service that uses it can be constructed. Implementation dependencies are only noted if they will play a part in defining the migration strategy.

It is important to note that the focus of this dependency analysis is on the implementation steps, not the architecture and it's structure that was done in the previous step.

Again, capturing this information in a way that is useful for the next step is an issue since the picture can get quite complex. Since time is the key factor for implementation, a time-dependency view of the information is recommended.

This analysis is done once, for the target architecture, since "implementation" in this context means the implementation of new or changes capability that replaces or augments existing capability. The complete picture (i.e. existing and target) is defined by default. 

This is a most difficult step since it involves combining information from two different analyses (i.e. Architectural structure, and implementation timeline) into a cohesive migration plan. These two views need to be brought together to define the target architecture migration plan. A timeline presentation is needed, similar to a project plan.

It is recommended that the timeline defined for the target architecture implementation (defined in the previous step) is used as the starting point. This timeline is then adjusted as required based upon:

• Existing architecture dependencies;
• Priorities assigned to gaps.

The resulting timeline will describe the migration plan as influenced by constraints imposed by the existing architecture dependencies, the new architecture implementation timeline, and mitigated by the prioritized gaps between them.