Efficiency Is Not Durable Value

For years, I have watched organisations pursue efficiency.

Reduce costs.

Automate work.

Simplify platforms.

Consolidate suppliers.

Most of the time, those objectives were entirely reasonable. Some of the strongest sustainability outcomes I have encountered also began as efficiency initiatives: less waste, less unused infrastructure, less duplication and lower energy consumption.

The assumption, however, can become deceptively simple:

If something is more efficient, it must be better.

What has always interested me is what happens next.

Some consequences only become visible much later:

• Critical skills disappear.
• Dependencies become concentrated.
• Exit options narrow.
• Resilience declines.
• Future choices become constrained.

None of these consequences automatically makes the original decision wrong. They may be manageable, accepted or outweighed by other benefits.

But they reveal an important distinction:

Efficiency and durable value are not always the same thing.

A recent investor briefing from the University of Cambridge Institute for Sustainability Leadership, AI and the Future of Work, sharpened that distinction for me.

Although written for investors, it examines four tensions that feel equally relevant to digital architecture:

• Short-term efficiency versus long-term organisational resilience
• AI adoption versus durable value creation
• Concentration versus diversification dynamics
• Company-level impacts versus system-level risks

These are not unfamiliar architectural concerns. What interested me was how clearly the briefing connected them to governance, organisational preparedness and execution.

The Efficiency Trap

Technology projects frequently promise:

• Lower costs
• Faster delivery
• Greater automation
• Improved productivity

Those outcomes can be entirely real.

The difficulty is that they do not always represent durable value.

A decision that appears highly successful today may create new dependencies, concentrated risks, reduced resilience, knowledge loss or future constraints.

The challenge is not to reject efficiency or optimisation. It is to ensure that their wider consequences remain visible.

When Good Decisions Produce Unexpected Consequences

Many organisational disappointments do not emerge because people made poor decisions.

They can occur even when capable people perform apparently sound analysis because the assessment boundary did not make every material consequence visible.

A business case may have been correct. The financial model may have been accurate. The environmental analysis may have been well executed.

Yet a critical dependency, organisational consequence or resilience concern remained outside the assessment boundary.

Years later, that omission becomes visible. By that stage, the architecture, process or investment may already be difficult to change.

This may not begin as a failure of intent, expertise or analysis. It begins as a visibility gap and becomes a governance problem when that gap remains unchallenged.

Four Tensions Worth Examining

The CISL briefing identifies four tensions that are equally relevant to architecture decision-making.

Efficiency Versus Resilience

An architecture can become more efficient while simultaneously becoming more dependent on a smaller number of providers, skills or platforms.

Reducing duplication and consolidating services may produce genuine benefits. It may also reduce redundancy, recovery options or organisational independence.

Adoption Versus Durable Value

Rapid implementation does not necessarily translate into sustained organisational benefit.

Deployment is relatively easy to observe. Durable value is harder to establish and may only become apparent after operating costs, workforce effects, governance obligations and longer-term outcomes are understood.

Diversification Versus Shared Dependencies

Multiple solutions can still rely on the same underlying services, models, data sources, infrastructure or supply chains.

Apparent diversification may therefore conceal common points of dependency.

Company-Level Impacts Versus System-Level Risks

A decision may appear beneficial for a project, function or organisation while contributing to wider risks or dependencies that only become visible when viewed at a broader level.

These tensions do not invalidate the original decision.

They remind us that important consequences may sit beyond its primary optimisation objective.

Preparedness, Not Technology Alone

The briefing goes beyond identifying tensions. It describes six areas of organisational preparedness:

• Strategic alignment to business strategy
• Workforce involvement
• Oversight and accountability
• Workforce and transition capability
• Execution and realised value
• Dependencies and resilience

Together, they reinforce an important distinction: possessing or adopting a technology is not evidence that it will create durable value.

Outcomes depend on whether an organisation can govern, implement, evaluate and adapt the change effectively.

The parallel with sustainable digital architecture is useful.

Environmental data, assessment tools and sustainability commitments do not produce better outcomes by themselves. Their value depends on whether they influence decisions, remain visible through governance and are verified after implementation.

This does not make the six preparedness areas new SDAF scoring dimensions. It suggests that organisational preparedness may constitute relevant evidence when it materially affects the durability of an architecture decision.

What SDAF Already Addresses

SDAF was created to address the limited visibility of environmental and lifecycle consequences within digital architecture decisions.

The framework already places emphasis on evidence quality, assumptions, lifecycle impacts, financial considerations, sensitivity analysis, governance conditions and outcome verification.

SDAF Studio and the Enterprise Integration Playbook also distinguish between evidence coverage, evidence quality, governance readiness, approval status, implementation conditions and verification obligations.

The purpose is not to calculate a perfect answer.

The purpose is to make important consequences visible before commitments are made and preserve the reasoning needed to review the decision later.

What Should Become More Visible

Reading the CISL briefing raised a question for me.

If environmental consequences can disappear from decision-making, what other consequences may also disappear?

Examples might include technology concentration, knowledge retention, critical skills dependencies, workforce-transition impacts, exit capability, deferred costs and transferred consequences.

These are not environmental metrics.

They should not be converted into carbon values.

They should not be hidden inside a composite sustainability score.

Yet they may still be material to the durability and success of a decision.

Material Consequences and Dependencies

One possible way of thinking about these issues is as material consequences and dependencies.

The objective is transparency.

Once identified, a consequence can be examined, supported by evidence, assigned an owner and referred to the appropriate governance or specialist process.

It may ultimately become:

• A recognised benefit
• A dependency
• A trade-off
• A mitigation
• An approval condition
• An accepted residual consequence

Visibility does not prescribe the outcome.

It ensures that the consequence remains part of the decision.

Visibility Without Conflation

It is important not to confuse visibility with scope expansion.

SDAF is not an attempt to absorb every discipline into a single framework.

Environmental impacts remain environmental impacts. Operational resilience remains operational resilience. Enterprise risk remains enterprise risk. Workforce assessment remains workforce assessment. Responsible AI governance remains responsible AI governance.

The objective is not to merge these concerns or translate them into a universal score.

The objective is to prevent a material consideration from disappearing merely because it sits outside the primary environmental calculation.

Related consequences can remain connected without being treated as equivalent.

Why This Matters to Me

One of the original motivations behind SDAF was the observation that environmental consequences often become visible only after decisions have been made.

While reading the CISL briefing, I found myself wondering whether the same visibility challenge applies to dependency, resilience, knowledge and organisational consequences.

I do not yet know how far that idea should extend within SDAF.

I believe it is worth exploring.

The purpose is not to transform SDAF into an enterprise-risk, workforce or responsible-AI framework. It is to examine whether a defensible architecture decision record should preserve material consequences that might otherwise disappear beyond the assessment boundary.

Clear Boundaries

SDAF is not a substitute for enterprise risk management, workforce assessment, operational resilience analysis, knowledge-management disciplines or responsible AI governance.

Its role is more modest.

Where these concerns become material to a sustainable digital architecture decision, SDAF may help preserve their visibility, evidence, ownership and traceability while directing them to the appropriate specialist or governance process.

Those specialist methods should remain authoritative within their respective domains.

An Invitation

This remains an exploration rather than a formal extension of the framework.

The CISL briefing prompted me to consider whether the same visibility discipline used for environmental and lifecycle impacts could help prevent other important consequences from disappearing from architecture decisions.

I am particularly interested in hearing from practitioners working with AI-enabled transformation, enterprise architecture, operational resilience, knowledge management, workforce transition and digital sustainability.

What consequences do you routinely find appearing too late? What remains invisible during decision-making? Where has efficiency produced unexpected longer-term consequences? And where do specialist methods need to remain authoritative?

SDAF does not need to measure everything that matters. But a defensible decision record should not allow something material to disappear merely because it sits outside the primary environmental calculation.