This study examines the practice of social impact assessment (SIA) in the global mining industry. We begin with the observation that SIA is undertaken within a specific set of commercial and structural (organisational and institutional) conditions: it is typically commissioned by mining companies and conducted by consultants. Whether the SIA is commissioned and managed by corporate ‘study teams’ or site level ‘social performance teams’, it is primarily a function of social performance management (broadly defined). Our aim is to understand how these and other ‘structural-commercial conditions’ shape SIA practices and outcomes and where changes are needed. This study responds to a persistent concern within parts of the mining industry, and among SIA practitioners, regulators and academics, that SIA is not achieving its intended purpose: that SIAs ‘fail to make an impact’. In doing so, we have surfaced honest, insider perspectives on the compromises and perverse incentives found in the dominant company-commissioned consultant-led mode of practice that shape SIA outcomes. We have identified how the potential of SIA is constrained by a structural conflict of interest in the relationship between mining companies and consulting firms, whereby a form of applied research that is ostensibly a ‘public good’ is undertaken primarily for private gain. Mining companies and consultants exist in a co-dependent relationship marked by significant power asymmetries that influence what types of impacts are identified and which ones are formally recognised and acted upon. Consultants rely upon mining companies for work and income, and mining companies depend upon consultants because they have not built up the internal capability to conduct SIAs. Social performance managers or study teams may argue that they need to use consultants to ensure that SIAs have a degree of ‘independence’, but a lack of internal technical capability remains an issue and this has a huge influence on the production and use of SIAs. Ultimately, limited social performance capability drives a feedback loop that perpetuates poor quality SIA practices and limits uptake and use. The recent scandals surrounding the global consulting firm PWC highlight this issue. In this case, the Australian Government has outsourced critical capability to consulting firms to such an extent that it has been accused of creating a ‘shadow’ public service that has ‘captured’ the government. Renowned journalist Judith Brett remarks: While external advice to government can be useful, the downsides of the practice are many as the public service loses and fails to replenish skill and institutional memory. Her observations about government integrity and overreliance on consultancies also hold for the mining industry. When companies fail to develop internal social performance capability this results in a lack of institutional memory on complex social issues around their assets which in turn compounds harm to communities. The objective of consulting firms (large and small) is to make a profit. They are, after all, businesses. Individual SIA consultants might be motivated by a desire to use their technical skills to do good, but they operate within a set of structural-commercial constraints that can compromise their ability to achieve these ends. This means that consultants and the firms they work for are not simply disinterested technocrats contributing to the public good, but part of a wider system of resource governance that upholds the power of the mining industry. Overall, the findings of this research confirm that if the mining industry wants to realise greater potential in SIA, then it should not outsource its core responsibilities and must invest in building social performance capability, especially in technical fields that support impact assessment processes and data-driven decision-making. This does not mean bringing all SIA processes ‘in-house’ (as there is always a role for specialist advisors) but developing technical competence across the range social performance domains (including but not limited to SIA). Minimally, this means developing capability in applied social science; study design and management; and capability to translate, apply and integrate SIA outcomes into operational decision-making processes. Without in-house technical capability, it will not be possible to reduce reliance on consultants or maximise their inputs. The remainder of this introductory section sets out the aims and methods of this study. Section 2 situates SIA in the mining social performance context and characterises the features of ‘good quality’ SIA in mining. Section 3 provides a synthesis of major themes and findings, and Section 4 presents key opportunities for change.