Destinations are increasingly expected to demonstrate sustainability performance. However, the governance systems responsible for delivering such performance are often fragmented, resource-constrained, and dependent on multiple stakeholders. Sustainability benchmarking tools can provide visibility and structure, but when used primarily for reporting, ranking, or reputational signalling, they risk reinforcing compliance rather than sustained organisational learning. This research proposes a recursive-processual model of absorptive capacity for sustainability governance to explain how Destination Management Organisations develop, embed, and sustain learning-led transformations through engagement with sustainability benchmarking tools. Despite the proliferation of sustainability benchmarking indexes such as the GDS-Index, their potential to activate deep organisational learning and governance change remains insufficiently theorised. This research addresses that gap by investigating how such tools function as triggers for recursive, relational, and problem-centred learning within complex institutional settings. Grounded in realist evaluation methodology and absorptive capacity theory, the research reframes sustainability benchmarking instruments not as static interventions but as processes embedded in context, shaped by strategic activation, peer engagement, and reflective governance. It moves beyond tool-centric framings to theorise how learning unfolds through dynamic causal configurations shaped by context and mechanism interaction over time.

The research employed a mixed-method realist evaluation approach, combining quantitative and qualitative strands to investigate how learning unfolds across diverse governance configurations. Data were drawn from two empirical strands: (1) GDS-Index benchmarking data from 2016 to 2022 across 93 cities, and (2) 23 problem-centred realist interviews across nine destinations. These sources were analysed through a context–mechanism–outcome (CMO) logic to model causal pathways and surface generative mechanisms.

Learning outcomes were found to depend less on the formal design of the benchmarking framework and more on how it is interpreted, politicised, and socially activated within specific organisational and governance contexts. Three dominant learning trajectories were identified: symbolic engagement driven by reputational motives; reactive recovery following ‘score shocks’; and transformative learning underpinned by relational, reflective, and politically empowered cultures. These were shaped by scaffolding mechanisms, notably leadership, intermediaries, and resources. These scaffolds operated not as linear enablers, but as ambivalent moderators whose influence varied depending on their activation, credibility, and institutional support. The dynamic interplay between enabling and constraining conditions was especially visible where destinations shifted from passive to activated modes of learning, often triggered by reputational pressure, peer benchmarking, or legitimacy shocks.

Learning was not linear but recursive, characterised by feedback loops in which prior adaptations reshaped future responses and governance strategies. This recursive learning process was visible as DMOs progressed through iterative cycles of disruption, adaptation, and feedback, producing varied outcomes from symbolic compliance to embedded change. Benchmarking functioned not as a neutral metric but as a reflective device through which actors navigated tensions between performance, legitimacy, and long-term transformation. The recursive-processual model of absorptive capacity developed in this research visualises this dynamic as a multi-directional process in which context and mechanism continually shape each other over time. The model captures causal trajectories across potential, transitional, and realised learning states, while also showing how feedback from later stages can disrupt or reconfigure earlier ones.

Theoretically, the thesis contributes a novel recursive model of absorptive capacity that integrates realist causality, organisational learning theory, and sustainability governance. It reframes absorptive capacity not as a linear accumulation of knowledge but as a negotiated, adaptive, and contingent process. Methodologically, it demonstrates the value of realist evaluation in explaining causality across complex governance settings by integrating mixed data to produce mid-range theory. The model also advances empirical insight into how sustainability knowledge is absorbed, transformed, or resisted within networked governance arrangements. Practically, it offers actionable insights for DMO managers, certification bodies, and policy actors seeking to repurpose benchmarking from an evaluative formality or compliance exercise into a basis for reflection, coordination, and adaptive governance.