Abstract
Digital transformation research overwhelmingly assumes that automation represents the dominant and desirable trajectory of technological change. Yet many organisations operate in contexts where core competencies cannot be automated without undermining legal mandates, economic value propositions, or temporally embedded expertise. This paper addresses the question: How do organisations pursue digital transformation when core competencies cannot be automated? Drawing on a qualitative multiple case study of five Scottish heritage textile manufacturers, we examine how firms navigate digital transformation under legal, economic, geographic, and temporal constraints. The study is based on 19 semi-structured interviews, extensive secondary data, and cross-case comparative analysis using an abductive, theory-building approach. Our findings reveal a distinctive pattern of constrained affordance actualisation. Rather than resisting digitalisation or pursuing automation, organisations actualise what we term preservation affordances, action possibilities oriented toward capturing, maintaining, and transferring human expertise while keeping humans central to value creation. This pathway unfolds through three stages: irreplaceability recognition, augmentation orientation, and preservation infrastructure development. We identify five mechanisms underpinning this process, including tacit knowledge externalisation, distributed cognition enablement, inter-generational translation, identity-preserving augmentation, and place-based enhancement. The paper contributes to affordance theory by conceptualising constraints as perceptual filters that redirect rather than suppress affordance actualisation. We extend digital transformation scholarship by identifying an expertise-preserving pathway that challenges automation-centric assumptions. By foregrounding contexts where human capability remains irreducible, we offer a more nuanced understanding of how organisations engage technology when replacement is neither feasible nor desirable.