Perspective-Based Architecture


Perspective-Based Architecture (PBA) is a method developed by Lewis Curtis, George Cerbone and Ryan Plantm of Microsoft Corporation.

The ultimate goal is to promote high-quality decision making. It is founded on the premise that if you don’t ask the right questions you will not look for the right answers. The method describes three stages of structured, focused questions. Questions that:

  1. Capture perspectives
  2. Examine impact of alternatives
  3. Examine impact of final proposal

The motivation for PBA is to address the complexity of aligning solutions with existing, current and projected IT environments. The complexity typically faced by an architect in a large enterprise. The solution focuses on addressing these critical issues:

  • What are good questions to ask when making good, discriminating architectural decisions?
  • How do I promote cohesive, well-thought-out decisions decisions being considered on my project?
  • How do I avoid common pitfalls when making architecture decisions?

The method has evolved from examining the conduct of successful senior architects. It has been observed that they usually posses three fundamental capabilities: knowledge, experience and perspective. Knowledge and experience are well-known and understood. However, perspective requires some explanation.

Perspective represents a frame of understanding that enables the architect to understand the impact of the decisions proposed on the environment. This is the most difficult of the capabilities to acquire, and thus is the focus of the PBA (and hence the name).

The method looks at the problem domain from four dimensions, and organizes questions accordingly to allow the architect to understand each one:

  1. IT Environment
    What are the other systems? what are the standards? What are the time/financial pressures, etc.
  2. Business Environment
    What are the strategic goals? What is the customer market? What is the political environment, etc.
  3. Requirements
    What is the currently utilized solution? What are the non-functional requirements? What are the project constraints?
  4. Trends
    What are the corporate trends? What are IT environment trends? Industry trends, etc.

While a senior architect looking at the above may remain unimpressed, for someone new to the game, such a list of questions represents a solid baseline of learnings. For those of us approaching the retirement corridor, is provides a training tool for those who may succeed us.


Leave a Reply

Your email address will not be published. Required fields are marked *