Enterprise Architecture Governance: From Requirements to Delivery – A Complete 6-Step Process

Why Enterprise Architecture Is Needed

Many large enterprises accumulate numerous systems and platforms, yet suffer from information silos, data duplication, and broken processes. Sales cannot access complete customer data, factories lack real-time inventory visibility, and executives rely on manual Excel collection. The root cause is not technical deficiency but the absence of holistic design and coordinated planning between strategy and execution. Enterprise Architecture (EA) acts as a city master plan, aligning business needs, IT capabilities, and data standards so that every IT action advances a unified goal.

The 4A Architecture Framework

EA comprises four domains, known as the "4A" architecture:

Business Architecture (BA) : Defines how business operates — processes, organization, roles.

Application Architecture (AA) : Defines system layout — functional modules, inter-system relationships.

Data Architecture (DA) : Defines information flow — data items, interfaces, quality standards.

Technology Architecture (TA) : Defines technical support — infrastructure, platform capabilities.

Some consultancies condense this into Business Architecture ("what business does") and IT Architecture ("how technology enables it"). For example, to improve order-to-delivery efficiency, EA coordinates process redesign with IT to ensure ERP integrates inventory and provides real-time order tracking. EA serves as the central coordinator, setting direction, system roles, and data exchange patterns without writing code or directly changing processes.

Six-Step Process from Demand to Delivery

Step 1: Unified Demand Entry with Tiered Classification

All new demands enter through a single portal or form, specifying objectives, scope, value, and affected systems. Demands are classified by value, impact, and complexity into three tiers:

Strategic demands : Support multi-year plans (e.g., five-year strategy); led by EA as major programs.

Business improvement demands : Managed via "Product + Version" mechanism.

Quick-response demands : Small changes handled by routine IT operations, outside EA governance.

This ensures important initiatives receive dedicated leadership while minor requests are processed rapidly.

Step 2: Unified Review and Prioritization

Submitted demands undergo a centralized review to assess priority, impact scope, and strategic alignment. Long-term initiatives like "14th Five-Year Plan" or "15th Five-Year Plan" projects — data platform construction, global process restructuring, business model transformation — require early EA involvement during the demand phase.

Step 3: EA Team Designs Blueprint (TOGAF Phase E & F)

Once approved, the EA team engages deeply, analyzing:

Gaps between current processes and target state.

Data connectivity and interface breakpoints.

Systems requiring reconstruction or new development.

This corresponds to TOGAF Phase E (Opportunity & Solution Design) and Phase F (Migration Planning). EA decomposes strategic goals into executable "project groups", sets priorities, and plans an "annual advancement rhythm". For instance, a "Global Unified Supply Chain Platform" might be phased as: Year 1 — process standardization; Year 2 — data platform build; Year 3 — core system launch; Year 4 — replication and rollout. EA acts as both engineer and planner.

Step 4: Version Cadence — Major, Medium, Minor Releases

To maintain orderly progress, a "Product + Version" mechanism schedules delivery:

Major version (X) : Functional restructuring, system upgrades; maximum once per year.

Medium version (Y) : Feature additions and optimizations; fixed three times yearly (e.g., May, October, pre-Spring Festival).

Minor version (Z) : Rapid delivery for small needs; handled by IT operations, bypassing architecture process.

This balances rhythm for large projects with agility for minor enhancements.

Step 5: EA Sinks into Projects for Dynamic Governance

During execution, EA does not merely monitor from afar but embeds into projects:

Participate in design to ensure system selection and data structures align with architecture.

Participate in development to detect deviations, technology misuse, interface mismatches.

When major deviations arise (e.g., MES system cannot support complex material combination logic), immediately adjust design rather than deferring to post-launch optimization.

This "dynamic governance" ensures projects follow the blueprint, refine the blueprint during implementation, and re-plan when necessary.

Step 6: Accumulate Architecture Assets for Reuse

Project completion triggers asset accumulation. EA archives:

Business process diagrams.

Data models and standards.

System interface specifications.

Decision records and design cases.

Project retrospectives and lessons learned.

Stored in an "Architecture Knowledge Base", these assets prevent redundant design, shorten future analysis cycles, and increase overall reuse. The ultimate EA value: making the next initiative easier, faster, and more accurate.

Annual Planning Drives EA Execution

To avoid fragmented, reactive development, enterprises conduct annual planning defining three objectives:

What strategic directions to achieve? (e.g., "Channel Digitalization", "Intelligent Factory")

Which capabilities to build? (e.g., "Unified Customer Profile", "Real-time Inventory")

Which projects to launch? How to map them into Product + Version tracks?

EA leads this planning, translating strategic goals into executable projects and plannable versions, closing the loop: Strategy → Architecture → Project → System.

Conclusion: EA Is Not Paperwork

EA is often mistaken as diagram-drawing and document-writing. True EA penetrates business, process, data, and IT to drive strategic landing, system collaboration, and data fluidity. It does not solve specific small problems but determines whether the organization can systematically solve problems. Only with solid EA mechanisms can an enterprise achieve:

Strategic goals with clear paths.

Project execution with rhythm.

System delivery with evidence.

Data usage with standards.

Change processes with control.

Outcomes that are reusable.

Business leaders, IT heads, and architects should ask: Does our company have such a mechanism? Is the EA team truly on the critical path of strategy execution? If not, now may be the time to build it.