WHY BIG SYSTEMS FAIL: LESSONS FROM GALL'S LAW

Across governments, corporations, and international organizations, there is a recurring temptation to pursue grand systemic reforms: new national healthcare architectures, sweeping education reforms, continent-wide digital infrastructures, or comprehensive social protection schemes. These initiatives are often motivated by noble ambitions to transform entire systems in a single leap. Yet, despite their scale and resources, many such endeavors collapse under their own complexity or fail to achieve their intended impact.

This phenomenon is not coincidental. It reflects a deeper principle captured succinctly by Gall’s Law, which states:

“A complex system that works is invariably found to have evolved from a simple system that worked. A complex system designed from scratch never works and cannot be patched up to make it work. You have to start over, with a working simple system.”

Gall’s Law offers a profound insight into why many large-scale system reforms falter. It suggests that functioning complexity is emergent, not engineered wholesale. This article examines the implications of Gall’s Law for system design and reform, highlighting why starting small and evolving gradually is not merely a cautious strategy, but a structural necessity in complex environments.

Understanding Gall's Law

Gall’s Law originates from John Gall’s 1975 book Systemantics, which explored the inherent dysfunctions of complex systems. Gall observed that functional complex systems are almost always the result of iterative evolution rather than upfront design.

This principle resonates strongly with findings from complexity science, systems theory, and evolutionary biology:

• Complex adaptive systems are path-dependent: Their current behavior is shaped by accumulated interactions over time, not by any central blueprint.

• Emergent order cannot be fully predicted: Interactions among system components produce novel behaviors that are often unexpected and irreducible to their parts.

• Incremental variation enables learning: Small experiments reveal information about what works; large jumps obscure causal feedback and magnify risk.

Gall’s Law therefore warns against the hubris of attempting to design functioning complexity ex nihilo. Systems that endure are those that build on a foundation of simpler, already working subsystems..

Why Big Systems Fail

When organizations attempt to launch fully-formed complex systems from the outset, they encounter several recurring failure modes:

1.  Cognitive Overload and Design Blindness:

The sheer scale of a complex system exceeds human cognitive limits. No design team can anticipate all interactions, dependencies, and potential failure points. Gaps and mismatches between components emerge only after deployment—by which point they are expensive and politically difficult to correct.

2. Fragile Interdependencies:

Large systems often integrate many interdependent modules that must all function correctly at launch. If any critical element fails, the entire system can collapse, as failure cascades through interconnections. This is especially true for digital platforms, where tightly coupled architectures amplify single-point failures.

3.  Absence of Real-World Feedback:

Designing at scale bypasses the iterative feedback that reveals user needs, contextual constraints, and unintended effects. This deprives designers of opportunities to adapt early, leading to solutions that may be technically sophisticated but socially unworkable.

4.  Institutional Inertia and Resistance:

Big systems disrupt entrenched routines, resource flows, and power structures. Sudden, sweeping change provokes defensive resistance from actors who feel threatened, further undermining implementation.

5. Irreversibility and High Sunk Costs:

Once large systems are deployed, sunk costs and reputational stakes make it politically costly to acknowledge failure or pivot course. This creates “lock-in” to dysfunctional architectures.

The history of ambitious system reforms is littered with such failures—from top-down agricultural collectivization schemes, to massive enterprise IT overhauls that were abandoned midstream, to overly centralized public service delivery reforms that crumbled under operational realities.

Why Small Systems Succeed

By contrast, small, simple systems are more likely to succeed and scale because they embody several crucial advantages:

1. Reduced Complexity:

Fewer components and relationships make it easier to understand, test, and refine the system. Problems are visible and manageable.

2.  Faster Feedback:

Small-scale pilots generate rapid feedback loops, enabling learning and adaptation before flaws become entrenched.

3. Lower Stakes, Lower Resistance:

Limited scope reduces political and institutional resistance. Stakeholders perceive less risk, making them more open to experimentation.

4.  Modular Scalability:

Working subsystems can be combined and recombined to create larger systems. This modular evolution mirrors natural and technological systems alike.

This pattern is evident in domains ranging from open-source software ecosystems  (which grow from simple, functioning modules) to microfinance models (which began as local experiments before scaling globally) and public policy innovations like conditional cash transfers (initially piloted in small regions before national rollout).

Implications for Systems Design and Reform

Gall’s Law implies a fundamental reframing of how we approach systemic change. Rather than attempting to design fully-fledged complex systems from scratch, reformers should focus on building small, working systems and allowing complexity to emerge through iterative scaling. Practically, this entails:

1. Start with Minimum Viable Systems:

Identify the smallest coherent set of functions needed to demonstrate value. This “minimum viable system” approach parallels the “minimum viable product” concept in innovation. The priority is not completeness but viability.

2. Build for Evolution, Not Perfection:

Design initial systems with modularity and adaptability in mind. Expect future modification as new information emerges. Avoid over-specification that locks in early assumptions.

3.  Embrace Experimental Governance:

Use pilots, prototypes, and regulatory sandboxes to test new approaches under real-world conditions. Treat failures as information, not setbacks.

4.  Scale Horizontally Before Vertically:

Encourage replication of working local models across multiple contexts before attempting vertical integration into national systems. This allows designs to be stress-tested and diversified.

5.  Invest in Feedback Infrastructure:

Build mechanisms for ongoing monitoring, user feedback, and system learning. Without continuous feedback, systems cannot evolve effectively.

Addressing Common Objections

Critics sometimes argue that small-scale approaches are too slow to address urgent large-scale challenges. However, attempting to “leapfrog” evolution often leads to wasted resources and lost time when big systems collapse. Small systems can be rapidly deployed, iterated, and scaled in parallel, accelerating rather than delaying systemic change.

Others worry that fragmentation of small experiments undermines coordination. This can be mitigated by designing for interoperability and curating knowledge-sharing platforms where learning flows across projects. Coordination should emerge from connecting functioning units, not imposing integration prematurely.

Case Example: The UK Government Digital Service (GDS)

An illustrative example of Gall’s Law in action is the creation of the UK’s Government Digital Service (GDS). Instead of attempting to overhaul all government digital services at once, GDS began by rebuilding a single platform, Gov.uk, around user needs.

Once this worked, it gradually developed reusable components, design standards, and agile delivery practices that were adopted across departments. By evolving from a small working subsystem, GDS avoided the fate of earlier “big bang” IT reforms that had repeatedly failed at massive cost.

Conclusion

Gall’s Law captures a vital truth for anyone seeking to design or reform complex systems: working complexity emerges, it is not engineered whole. Attempts to create large, intricate systems from scratch are prone to collapse, not because of incompetence or bad intent, but because complexity by its nature resists centralized design.

To succeed, system builders must start small—crafting simple systems that work, then letting complexity grow through iterative adaptation, modular expansion, and continuous learning. This evolutionary path is not a sign of timidity; it is the only reliably successful route through complexity’s inherent uncertainty.

In a world of increasingly complex and interdependent challenges, the future will belong to those who master the art of starting small to go big.

In a collaborative meeting, experts participate in strategic conversations to promote business growth and innovation.