An organisation adopts a formal EA framework and produces extensive deliverables while decisions remain slow. How should such a framework be used?

The failure

The framework has been adopted as a process to be executed rather than as a toolkit to be drawn from. Formal EA frameworks describe a comprehensive method with many phases and deliverables, intended to be tailored. Organisations that execute them completely produce a large volume of artefacts whose maintenance consumes the function that produced them.

The frameworks say to tailor. Almost nobody does, because the complete version is legible and the tailored version requires judgement.

What to take from it

The genuinely valuable parts:

• A vocabulary. Shared terms for business, data, application and technology architecture, so conversations across a large organisation are about the same things.
• The discipline of tracing from business need to technology decision. Requirements to capabilities to applications to technology, so a technology proposal can be argued in business terms rather than technical ones.
• A repeatable method for a major change programme, where the rigour genuinely pays.
• The idea of an architecture repository with reusable, versioned assets.
• Governance and compliance concepts, applied selectively.

What to leave

• Producing every deliverable for every initiative. Most changes need a fraction, and asymmetric depth is what makes the practice affordable.
• Uniform depth across the estate. Two levels of decomposition everywhere, five where a decision is pending.
• A full architecture development cycle for small changes.
• Maintaining models nobody consults. If no decision depends on an artefact, it is cost.

The reorientation

Start from the decision, then pull in only the framework elements that inform it. Should we consolidate order management across regions? That needs a capability map at some depth, an application overlay, a cost overlay and a target architecture for one area — not a complete enterprise model.

The framework is then a checklist of what one might consider, used to avoid missing something, rather than a sequence of deliverables to be produced.

The honest assessment

Formal frameworks are most valuable in organisations that are large, regulated, and undertaking major transformation, where common vocabulary and traceability genuinely reduce risk. They are least valuable in fast-moving product organisations, where the deliverables age faster than they can be produced.

Adopting the vocabulary and the traceability discipline while declining the deliverable set is a legitimate and common choice — and one that many practices arrive at eventually, after producing several years of artefacts nobody read.