How to Design an ESG Operating Model That Actually Works
Many organisations treat environmental, social and governance work as a rolling sequence of initiatives: a disclosure here, a carbon inventory there, a supplier survey when a customer demands one. This project-based habit produces motion without momentum. It also explains why so many firms find themselves under-prepared when regulation arrives. Designing an ESG operating model that actually works means shifting from episodic effort to a governed, repeatable function with defined roles, clear ownership, disciplined data flows and decision rights that reach into the business. The question is not whether ESG matters to the enterprise, but whether the enterprise is structured to deliver on it reliably.
The distinction is more than semantic. A project has a start date, an end date and a temporary team. A function has standing accountability, a budget line, service level expectations and a place in the organisation's governance architecture. ESG has crossed decisively into the second category. Consider the European Union's Carbon Border Adjustment Mechanism, now moving into a period of active scrutiny, with a World Trade Organization panel set to review the scheme and dozens of trade partners watching the dispute triggered by Russia. Steel suppliers are being advised to prepare early for verification requirements, because the embedded emissions data that CBAM demands cannot be assembled retrospectively in a single sprint. It requires continuous measurement, verifiable records and a chain of ownership from procurement to disclosure. That is an operating model problem, not a reporting deadline.
Effective ESG operating model design begins with roles and ownership, and this is where most organisations falter. Sustainability teams are frequently accountable for outcomes they do not control. A central team may own the carbon target, but the emissions sit in operations, procurement and logistics. Assigning ownership therefore means placing accountability where the data and the decisions actually live, then giving the central function the mandate to set standards, challenge performance and consolidate results. A practical structure distinguishes three layers: a small central team that sets policy, methodology and reporting standards; business and functional owners who hold accountability for performance in their domains; and a governance body, typically anchored at board or executive committee level, that reviews progress and resolves trade-offs. Ambiguity at any of these layers is where ESG commitments quietly erode.
Data flows are the second pillar, and arguably the one that determines whether the model survives contact with regulators and auditors. ESG data is notoriously scattered across spreadsheets, third-party platforms, utility bills and supplier attestations of varying quality. A working operating model treats ESG data with the same rigour applied to financial data: defined sources, documented methodologies, controls over collection and transformation, and an audit trail that a verifier can follow. The CBAM experience is instructive here. Businesses that can trace embedded emissions from primary data through to declaration will hold a genuine competitive advantage, while those relying on default values and last-minute estimates will pay for the gap in both cost and credibility. Data architecture is not a technical afterthought to the operating model. It is central to it.
Governance is the third pillar, and recent developments in artificial intelligence make its importance impossible to ignore. As firms bring AI into ESG processes, from emissions estimation to disclosure drafting, governance discipline becomes a condition of trust rather than a constraint on speed. An EY study found that almost half of US companies skip their AI governance policies to accelerate deployment. That shortcut is precisely the failure mode a mature operating model is built to prevent. When AI tools feed into regulated disclosures, the absence of governance is not an efficiency gain but a liability waiting to surface. The same principles that govern ESG data quality apply to the AI systems that increasingly process it: clear ownership, documented controls, human review at defined points, and accountability that does not evaporate when the technology produces an answer. The wider debate, including calls from developing nations at the United Nations for a greater say in AI governance, signals that expectations around accountable, auditable systems are hardening, not softening.
Bringing these pillars together requires a deliberate design sequence rather than an organisational chart drawn in isolation. The starting point is a materiality-informed view of what the organisation must manage, which then determines what must be measured, who must own it and how it must be governed. From there, the operating model specifies the cadence of the function: monthly performance reviews, quarterly governance sessions, an annual assurance cycle. It defines the interfaces between ESG and finance, risk, procurement and operations, so that sustainability data enters the same decision forums as commercial data. It also sets out escalation paths for when targets are missed or new regulation, such as an amended CBAM regime, changes the requirements. A model that cannot absorb regulatory change without a reorganisation is not a model. It is a snapshot.
There is a strategic dividend to this discipline that senior leaders should recognise. Regions pursuing manufacturing decarbonisation, as analysis of the Western Balkans has shown, will reward firms whose emissions performance is demonstrable and whose data withstands verification. The same holds across supply chains subject to border adjustment mechanisms and expanding disclosure regimes. An operating model that produces reliable, verifiable ESG information becomes a commercial asset: it shortens due diligence, supports access to capital, and positions the organisation to respond to customer and regulatory demands without disruption. The firms that treat ESG as a function are the ones able to answer difficult questions on demand, while those running it as a project are still commissioning the study.
Designing an ESG operating model that actually works is ultimately an exercise in institutional design: placing accountability where performance is created, treating ESG data with financial-grade rigour, and building governance that extends to the AI systems now embedded in the process. CorpStage works with boards and executive teams to design and implement operating models of this kind, mapping roles and ownership, structuring data flows for assurance, and establishing governance that holds across both ESG and AI. The objective is not a heavier compliance burden but a function that runs, adapts and earns trust. Organisations that make that transition now will be the ones prepared for what regulation and markets are already demanding.