Modern Enterprise Architecture: Architecting for Outcomes • Simon Rohrer • GOTO 2025
Simon Rohrer challenges traditional Enterprise Architecture (EA) principles, proposing a modern approach focused on outcomes, continuous evolution, and socio-technical alignment. He outlines five key tenets: Aligning Value, People & Technology; achieving Better Value Sooner, Safer, Happier; implementing Continuous Conversational & Automated Governance; scaling DevOps across the enterprise; and fostering Evolutionary Enterprise Architecture.
Simon Rohrer challenges prevailing notions of Enterprise Architecture (EA), advocating for a ‘modern enterprise architecture’ that prioritizes outcomes and breaks with traditional, phase-based approaches. Inspired by Dave Farley’s modern software engineering, this approach is fundamentally about achieving desired results.
The Evolution of Enterprise Architecture
Rohrer provides historical context, tracing software development from the 1990s—marked by the Standish Group’s 70% project failure rate, the rise of Netscape and the internet, and Kent Beck’s Extreme Programming Explained—to the present, where “software ate the world.” This shift, driven by Agile, DevOps, cloud, and systems thinking, transformed development from months-long, phase-gated projects to continuous, product-centric evolution, delivering ideas to production in days or hours. However, many enterprise architects often failed to adapt, clinging to phase-based thinking, leading to increased complexity. Roger Sessions’ observation that “complexity is the enemy” highlights the critical need for EA to prevent systems from devolving into tangled webs.
The ABCDE of Modern Enterprise Architecture
Rohrer distills modern EA into five core tenets:
A: Aligning People, Value, and Technology
Traditional EA often separates business and IT, but modern EA recognizes the profound interrelation between organizational and system architecture. Drawing on Conway’s Law, Ruth Malan’s insight that “If the architecture of the system and the architecture of the organization are at odds, the architecture of the organization will win,” and Alan Kelly’s counterpoint, Rohrer emphasizes a feedback loop: organizations shape systems, and then systems constrain organizations. Effective modern EA consciously designs both. This alignment aims for a sustainable flow across value, organization, technology, and process, operating fractally at self-similar levels—from two-pizza teams (as per Team Topologies, designing “software that fits in your head”) to larger “teams of teams.” Architects within these teams act as coaches, facilitating decision-making rather than dictating it. This creates a sociotechnical architecture that balances coupling across interconnected systems.
B: Better Value Sooner, Safer, Happier
Traditional EA focused on outcomes like standardization, consistency, or cost reduction, which often manifest as outputs (e.g., migrating to cloud, using Kubernetes). Modern EA, however, prioritizes business agility and a balanced scorecard of quality:
- Better: Value that makes the organization a “special snowflake.”
- Sooner: Optimizing for flow, reducing time from concept to cash or learning.
- Safer: Building an agile, not fragile, system, optimizing for security, privacy, and minimum viable compliance.
- Happier: Ensuring satisfaction for customers, colleagues, citizens, and the environment.
Rohrer illustrates this by contrasting inefficient layered architectures (UI, API, service, database teams requiring multiple handoffs for a simple feature) with restructured, cross-functional teams owning entire business capabilities. Eliminating handoffs significantly reduces coordination costs, improving efficiency, reducing risk, and lowering costs.
C: Continuous, Conversational, and Automated Governance
Complexity has an asymmetrical cost: adding complexity can be cheap, but removing it is often far more expensive due to widespread changes in distributed systems. Traditional Architecture Review Boards (ARBs), designed for months-long waterfall phases, are an anti-pattern in agile, continuous environments. Instead, modern EA implements governance on realized architectures. This means:
- Conversational Approval: Requests (e.g., for a new service, data access, or synchronous communication) are submitted as pull requests, triggering rapid, conversational reviews within an architecture community of practice.
- Paved Paths: Standardized, well-vetted solutions are provided via platform engineering, allowing teams to proceed without explicit approval if they use these compliant tools. If a team uses the platform, many architectural and security checks are pre-fulfilled.
This approach replaces phase-gated, document-led, adversarial governance with continuous, conversational, and automated oversight, fostering collaboration rather than bottlenecks.
D: DevOps at Enterprise Scale
Modern EA embraces the DevOps loop at all fractal scales, from individual teams to the entire enterprise, aiming for fast feedback everywhere. The Architecture Center of Excellence functions as a DevOps team, focusing not just on future changes (dev) but also on operations, learning from deployed systems. This requires close collaboration with service managers and Site Reliability Engineers (SREs), integrating their data-driven insights on non-functional requirements (SLOs, SLIs) into architectural considerations. John Cutler’s framework, mashed up with the DevOps loop, highlights that a full product team encompasses everything from opportunity selection to continuous operation. This “you build it, you run it” philosophy is essential for tackling complex problems through rapid feedback and continuous adaptation.
E: Evolutionary Enterprise Architecture
Finally, modern EA emphasizes the curation of continuous evolution in both technical and organizational architectures. Referencing Building Evolutionary Architectures, Rohrer notes that measuring fitness functions in distributed systems is challenging but critical. He draws an analogy to “punctuated gradualism” in biological evolution: most changes are slow and steady, but occasional “step changes” are necessary for significant shifts. Enterprises must continuously address complexity and technical debt (e.g., allocating 20% funding for continuous improvement) to combat entropy. While large-scale rewrites are costly and often ill-advised (e.g., “second system syndrome”), strategies like Martin Fowler’s Strangler Fig pattern allow for continuous evolution by gradually replacing legacy components. Simon Wardley’s work on Wardley Maps further underscores that evolution is driven by competition, necessitating continuous adaptation. The ultimate lesson, echoing the final Standish Report, is to “forget projects” and instead focus on continuous development and implementation—“just do it”—to foster constant evolution.