The Architecture Graveyard: Trap #2 Not Making Progress

February 02, 2017

There is a frankly alarming trend in the EA profession: short-lived EA Teams. The average lifecycle of an EA team is 2 years. Most organizations have a history of standing up, disbanding, rebooting and sidelining EA teams. Our practice that establishes & enhances EA teams sees more of this than most. When we ask leadership teams what didn’t work before we arrived, it is always the same: nothing consumable was delivered. Worse when the architecture was provided it was irrelevant in the drive to realizing strategic goals. This is a classic symptom of Trap #2 Making Progress.

Everyone who has attended our EA with TOGAF & Navigate or TOGAF Certification courses knows the Angry Duck. A simple piece of clipart showing a duck with its arms impatiently crossed and eyes lowered in contempt. The Angry Duck is a reminder to all architects that despite the multi-pressures around producing useful architecture and guidance, there is never enough resources and information for an architect to be 100% confident or 100% complete. Instead, we must get done. We must deliver sufficient for the purpose.

While EA teams are chasing the incidental or being paralyzed by another round of analysis organizations are moving forward with deliberate and accidental change. Useful architecture is sufficiently complete to provide sufficient confidence to support the decision that has not been made.

Our profession’s TOGAF framework includes the tools for getting to done: iteration and varying levels of detail in the Enterprise Landscape. High functioning EA teams align their deliverables to the business cycle and use a structured repository.

Decisions that will frame a project, approve the critical specifications that deliver measurable value typically require less detailed analysis than what is necessary to guide & constrain a change team implementing a solution. Frankly getting to done means stopping work when the current iteration can address the problem at hand.

We use the agile concept of backlog to hold the work of future ADM iterations. In practice it is simple: the roadmap will identify initiatives, or programs, that need more work; programs will have projects that require more work, and projects will hold solutions that require more work. Each iteration extends the previous work, validates & refine specifications and controls. Filling in the EA Landscape.

Purposeful architecture development drives delivery. TOGAF’s concept of an architecture project is a card in our Kanban backlog. At the start of every Predictable EA sprint, we pull cards and break down what is needed in the repository and what will be delivered to the consumers (decision makers and stakeholders).

With an expanding repository, time-to-market accelerates with re-usable work. When the portfolio architecture did the stakeholder analysis and defined the preferred approach to the work packages the architect walks in the door knowing priority and whether the current work should start with the assumption of rip & replace, evolution, or revolution. More importantly, it drives convergence within the set of Stakeholder demands. Stakeholder clarity on what must change drives convergence on strategic benefit.

All of a sudden, without a mandate, without complex governance structures and their imaginary decision authority the EA team is delivering value. Directional decisions have visible traceability to value. Implementation decisions conform to specification and control that are obviously defensible. Across the board, the organization starts to change with more success and less friction.

All because the EA Team dodged the trap of Not Making Progress