The Law of Alignment

Written by

Sergio Caredda

Posted on

May 13, 2026

Key Implications of the Law of Alignment

1. Alignment Is Always Present — the Question Is to What

Every organisation is aligned to something. A bureaucracy is aligned to procedural compliance. A startup is aligned to founder intuition. A matrix organisation is aligned to whatever the strongest informal network decides, regardless of what the reporting lines say. The point is not whether alignment exists but whether it is intentional.

This is why alignment audits that measure "how aligned are we to our strategy" often miss the target. The more revealing question is: what are we actually aligned to, and how does that differ from what we declared? I have seen this in practice more than once: an organisation that insists it is aligned to customer centricity while every incentive, every budget cycle, every promotion criterion points toward product volume. The alignment is real. It is just not to the thing on the strategy slide.

2. Conformity Is the Most Common Failure Mode

The distinction between Alignment and conformity is this law's load-bearing argument — the way the complication-versus-variety distinction runs through the Law of Requisite Variety.

Conformity is what happens when organisations pursue alignment by enforcing uniformity across structural levels. The same KPIs cascade unchanged from the executive committee to the frontline. The same process templates get deployed across regions with wildly different market conditions. The same values statement appears on every wall, in every town hall, while the actual reward system incentivises something quite different. Rosa et al. (2025) found that coherence gaps — between corporate communication and actual practice — reduce employee commitment more than the absence of a values statement altogether. The appearance of alignment without operational coherence is worse than acknowledged misalignment.

This is the trap. Conformity feels like alignment because it produces visible consistency. Everyone says the same things. The dashboards all point in the same direction. But coherence — the state where what you say and what you actually do are the same thing — requires something harder: it requires that the design of work, decisions and rewards at each level genuinely serves the stated direction, even when the form varies.

3. Vertical and Horizontal Alignment Are Different Problems

Most organisations manage vertical alignment reasonably well. Goals cascade from strategy through business units to teams. The challenge is horizontal alignment — ensuring that functions at the same level are congruent with one another, not just individually linked to the level above. A 2025 bibliometric review confirms what practitioners have long suspected: the field still underweights horizontal congruence.

Consider a short-term incentive design — the annual bonus structure that most listed companies use to translate strategy into individual targets. A design that perfectly aligns each function's bonus metrics to the corporate strategy. Sales is measured on revenue. Operations on cost efficiency. HR on engagement scores. Each metric, taken alone, is vertically aligned. But horizontally, they pull in different directions: sales pushes volume that operations cannot deliver efficiently, while HR measures sentiment in a workforce squeezed between the two. The vertical alignment is real; the horizontal incoherence is also real. The design produces conformity to the strategy cascade while undermining the coherent action the strategy actually needs.

This is not a hypothetical. I have encountered versions of it in every organisation design engagement I have led. The symptom is familiar: every function can demonstrate strategic alignment on paper, yet the organisation cannot execute cross-functional priorities. The alignment is there. The coherence is not.

4. Alignment Must Deliberately Stop to Preserve Variety

Here is where Alignment intersects Requisite Variety. If you enforce perfect alignment across every level and every unit, you destroy internal variety — the very thing the organisation needs to respond to a complex environment. A federal constitution aligns member states on common defence and trade but deliberately preserves local legislative variation. An immune system aligns diverse antibody responses to the same threat without requiring every cell to behave identically. The design principle is the same: alignment sets the vector; variety determines the range of responses within that vector.

Kathuria, Joshi and Porth (2007) found that the pathway from alignment to performance runs through coherence of action, not top-down enforcement. High alignment metrics combined with low local autonomy did not outperform. The implication for design is clear: alignment is a constraint, not a maximisation target. The question is not "how aligned can we get?" but "where must alignment hold, and where must it deliberately stop?"

Alignment as a Discipline, Not a Slogan

The management conversation about alignment has been dominated by measurement frameworks — Kaplan and Norton's balanced scorecards, Henderson and Venkatraman's strategic alignment model, and the cascading OKR systems that followed. These tools are useful for diagnosing misalignment. They are less useful for producing coherence, because coherence is not a measurement problem — it is a design problem.

Galbraith's Star Model gets closer. The model holds that strategy, structure, processes, rewards and people must be mutually reinforcing — and that performance degrades when any point of the star drifts out of alignment with the others. But even Galbraith frames alignment as fit between components. What the Law of Alignment adds is the failure mode: the most common way alignment fails is not through drift but through conformity — through designing every component to look consistent while the actual operating behaviour diverges.

The discipline, then, is not alignment itself. It is the continuous work of testing whether declared alignment matches operational reality — and being willing to accept that the answer is often no. I have written elsewhere about consistency as a design test: the degree to which the elements of an organisation reinforce one another rather than pull apart. Alignment is the law. Consistency is the test. The two are related but not identical, and confusing them is a design error worth naming.

The AI Angle: Alignment Automated

There is a contemporary application worth calling out. Organisations are deploying AI to automate processes that were previously held together by human judgment — and in doing so, they are surfacing alignment failures that human workers had been silently absorbing for years.

When a process was "misaligned" but a skilled employee compensated by interpreting the intent behind the procedure rather than following it literally, the misalignment was invisible. Automate that process, and the machine follows the procedure as written — which is aligned to the declared design, not to the actual intent. The result is that AI deployment is becoming an unintentional alignment audit: it reveals every gap between what the organisation says it does and what it actually does. The gaps were always there. The automation just made them visible.

This mirrors the argument in the Requisite Variety piece about AI as a variety amplifier being misused as a variety reducer. Here the pattern is parallel: AI has the potential to surface and correct alignment failures at scale. Most organisations are instead using it to enforce conformity faster.

Applications: Where the Law of Alignment Matters in Practice

The Law of Alignment is not a metaphor. It is a design constraint that should sit alongside cost, governance and scalability when decisions about how organisations work are made. Four areas deserve more deliberate application:

Operating Model Design. When designing or reviewing an operating model, test for coherence, not just consistency. Ask whether the processes, governance and technology stack are aligned to the actual strategy or to the strategy-as-declared. The gap between the two is the alignment debt.
Reward Architecture. Incentive design is where alignment is most visibly tested and most frequently faked. Vertical alignment of metrics is necessary but not sufficient. Horizontal coherence — ensuring that what Sales is rewarded for does not contradict what Operations needs to deliver — is where the design discipline sits.
Change and Transformation. Major change programmes often assume they are "aligning the organisation to the new strategy." In practice, they are frequently enforcing conformity to a new set of declared values while leaving operational alignment untouched. The Law of Alignment suggests testing the change programme's own coherence: does the transformation roadmap change actual work and decisions, or only the language?
Leadership Practice. Alignment is initiated at the leadership level, and this is where the conformity trap is most consequential. A leadership team that demands alignment while rewarding conformity — punishing dissent, discouraging local adaptation, celebrating visible consistency over operational coherence — is the single most common source of the failure mode this law names.

The “Laws of Organisation Design” Collection. This article belongs to a series focused on the different principles and laws that govern organisational behaviours, decision-making and group work, and that should be taken into account in organisation design projects.

The Laws of Organisation Design

Coming Soon

Law of Self-Similarity
Law of Synergy
Law of Minimizing Interface Complexity

Conclusion

The Law of Alignment is deceptively simple: an organisation is always aligned to something, and the design job is to ensure that the alignment is intentional and coherent rather than accidental and conformist. The hidden danger — and I keep coming back to this across the series — is that the most common failure does not look like failure. It looks like everyone agreeing, every dashboard green, every function "strategically aligned." The conformity is visible. The incoherence is not.

This law does not claim that alignment is always desirable, or that more alignment is always better. There are contexts — early-stage ventures, creative teams, loosely coupled ecosystems — where enforcing alignment would destroy the very thing that makes the system work. The law claims something narrower: that when alignment is present, its quality matters more than its quantity, and that conformity is the mode in which alignment most commonly fails.

Treating alignment as a design law — not a scorecard metric, not a leadership mantra — means accepting that coherence requires continuous testing, that conformity is the trap, and that the discipline is in knowing where alignment must hold and where it must deliberately stop.

Ignoring this does not make the tension disappear. It only makes the failures harder to see.