Constraints and Sequencing

AI transformation draws on five connected domains: Strategy, Leadership, Governance, Transformation and Change Management. A weakness in one rarely stays in one. It limits what the others can achieve and changes the order in which conditions need attention. This page sets out how the SLGTC Framework describes those relationships, and how ORAVORN uses them to name a primary constraint.

Each ORAVORN result identifies one primary constraint: the organisational condition to address first. The domains depend on one another, so a weakness in one can reduce the effectiveness of the others, and the lowest score does not always show where to begin.

A constraint, in this framework, is a condition whose weakness, in the context of the current state of the other four domains, materially reduces their collective effectiveness and limits the organisation’s capacity to progress AI transformation. The priority is therefore not always the lowest score. Where direction is undefined, it comes first, as the sequencing principles describe. Once AI activity has begun, a domain failing across all three of its variables may take priority.

ORAVORN names one primary constraint as the starting point for action, without implying that other weaknesses matter less. Any other domain that is also failing is identified alongside it. The constraint is not fixed. As the organisation acts and circumstances change, it may shift, and attention should be reassessed accordingly.

The framework applies four sequencing principles to explain how weaknesses in one domain can limit the effectiveness of others and influence the order in which organisational conditions should be addressed. These principles establish priorities for action. They do not rank the five SLGTC domains by overall importance.

A defined direction comes first. Where strategic clarity is weak, efforts across the other domains risk being misdirected because the organisation has not established what it wants AI to achieve or how success will be recognised. Governance, leadership alignment and change initiatives may be well developed, yet still serve priorities that have not been clearly defined. Competitive positioning and capacity relative to ambition influence which direction the organisation can pursue and how it pursues it, but they do not substitute for strategic clarity.

Leadership alignment multiplies execution. Alignment among senior leaders influences how effectively the organisation can mobilise its existing capabilities and investments. Where leaders share an understanding of AI priorities and direction, decisions, resources and implementation efforts are more likely to reinforce one another. Where alignment is weak, investment across the other domains may become fragmented, delayed or directed towards competing priorities, and weak alignment suppresses the effect of investment elsewhere.

Governance establishes a threshold for scaling. Governance does not need to be fully developed before AI activity begins. It does need to reach a level appropriate to the risks before AI is extended into core processes and consequential decisions. As deployment expands, so do the consequences of errors and inadequate oversight, and arrangements sufficient for a limited pilot may be inadequate at scale. Where governance falls below the threshold required for the intended deployment, scaling is not advisable until the deficiencies are addressed, regardless of strength in other domains.

Transformation is gated by organisational readiness. Transformation involves embedding AI into operating models, workflows, data systems and workforce capabilities, often requiring substantial investment and exposing the organisation to wider operational consequences. Its effectiveness depends on the interaction of strategic direction, leadership alignment, proportionate governance and the conditions needed to support adoption. Where these foundations are insufficient, accelerating transformation may amplify existing weaknesses and risks rather than produce sustainable progress. The extent and pace of transformation should therefore reflect the organisation’s readiness, with foundational weaknesses addressed before they undermine implementation at scale.

Beyond the four sequencing principles, the framework identifies specific relationships through which weaknesses in one domain can intensify weaknesses in another, limiting the organisation’s overall capacity to progress AI transformation. These interactions help explain why a weakness may be more consequential in combination with another than when considered in isolation.

Strategy and leadership. Weak strategic clarity can intensify disagreement among senior leaders about AI priorities, while poor leadership alignment can prevent an otherwise clear strategy from being translated into coordinated action. The two domains should therefore be interpreted together: strategic direction shapes what leaders need to align around, while leadership alignment influences whether that direction can be pursued consistently.

Ambition and governance. The governance required before transformation proceeds depends partly on the organisation’s AI ambitions, the intended scope of deployment and the consequences of potential failure. An organisation seeking to create new market opportunities through AI may face different uncertainties and risks from one seeking to keep pace with established practice. Greater ambition can increase the need for safeguards proportionate to the novelty, reach and consequences of the proposed activity.

Leadership and transformation. Leadership alignment influences whether investment in operating models, data and workforce capabilities translates into coordinated transformation. Where senior leaders disagree about the purpose, priorities or expected outcomes of AI, implementation efforts can become fragmented, even when the necessary resources and capabilities are available.

Transformation and change. Transformation can outpace an organisation’s capacity to prepare, support and sustain adoption. When new systems and workflows are introduced faster than employees can understand, integrate and use them effectively, adoption gaps can emerge. These gaps may reduce the value of the investment and require additional resources to address, particularly once less effective ways of working have become established.

Psychological safety and governance. Governance depends not only on formal controls but also on whether people can raise concerns, report failures and challenge AI-related decisions without fear of adverse consequences. Where psychological safety is weak, important information may not reach those responsible for oversight, allowing risks and problems to remain undetected. Effective governance therefore requires both appropriate formal arrangements and an organisational environment in which concerns can be surfaced and acted upon.