How to run a double materiality assessment

What a double materiality assessment is

A double materiality assessment (DMA) is the structured process that determines which sustainability matters a company must report under the EU’s CSRD3 and the ESRS2. Every candidate topic is tested from two perspectives — impact materiality (the company’s actual and potential effects on people and the environment) and financial materiality (the sustainability-related risks and opportunities affecting the company) — and a matter is material, and therefore reportable, if it is material from either perspective2.

The official implementation guidance is EFRAG IG 1: Materiality Assessment Implementation Guidance, finalised in May 20241.

It describes the recommended process, the scoring dimensions for each perspective, and how the results feed the disclosures required by ESRS 212.

This page is the methodology companion to our double materiality concept explainer — read that first if you need the definitions and the UK/EU policy divide.

Here we cover the process itself: steps, scoring, stakeholder engagement, value-chain coverage, and disclosure.

When UK companies need one

No UK regime requires a double materiality assessment.

UK SRS S1 and S2, published by the Department for Business and Trade on 25 February 20265, use financial (enterprise-value) materiality only — the UK SRS S1 materiality judgement is a single-perspective exercise5.

Three groups of UK companies still need, or benefit from, a DMA:

• CSRD-caught UK groups — those meeting the post-Omnibus Article 40a test (EU net turnover above €450m for each of the last two consecutive financial years, plus a large EU subsidiary or an EU branch above €200m turnover), reporting at group level from FY2028 [4]
• UK subsidiaries of in-scope EU parents — the parent’s ESRS report consolidates the UK entity, so its data flows through the group DMA [2] [3]
• Voluntary GRI reporters — the GRI Standards are built on impact materiality [6], which is one half of the double materiality test; a structured DMA strengthens and future-proofs that work

Which UK groups are actually caught by CSRD after Directive (EU) 2026/4704 is a threshold question in its own right — our guide to CSRD for UK companies works through the scope tests.

Even out-of-scope companies sometimes run a lightweight DMA voluntarily: it produces a defensible topic list for strategy and stakeholder reporting beyond what single materiality yields1.

The EFRAG IG 1 process — four steps

EFRAG IG 11 describes the recommended assessment as a sequence of steps running from context to conclusions. In practice it is iterative — scoring often sends teams back to refine the IRO list — but the architecture is consistent:

1. Understand the context. Map the business model, activities, products and services, geographies, and business relationships across the upstream and downstream value chain, and identify affected stakeholders1. This grounds the assessment in what the company actually does and who it touches.
2. Identify actual and potential impacts, risks and opportunities (IROs). Build a long-list of the company’s positive and negative, actual and potential impacts on people and the environment, and of the sustainability-related risks and opportunities facing the company — using the ESRS topic list as a completeness check12.
3. Assess and score materiality on both dimensions. Score impacts by severity — scale, scope, and irremediable character — with likelihood added for potential impacts; score risks and opportunities by the magnitude of the anticipated financial effects and their likelihood1. Apply documented thresholds to separate material from non-material.
4. Determine material matters and material information. Consolidate scored IROs into material sustainability matters, decide which ESRS topics and disclosure requirements apply, and record the process and outcomes for disclosure under ESRS 2 IRO-1 and IRO-212.

The output is twofold: a list of material matters driving which topical standards the company reports against, and an audit trail — methodology, thresholds, stakeholder input, conclusions — that the assessment disclosures and any assurance provider will draw on1.

Scoring the two dimensions

EFRAG IG 1 keeps the scoring criteria deliberately principle-based — companies design their own scales and thresholds — but the dimensions are fixed1.

Scale asks how grave (or beneficial) the impact is; scope asks how widespread — how many people, or how large an area; irremediable character asks how hard the harm is to put right1.

On the financial side, magnitude and likelihood are assessed over short, medium and long-term horizons, recognising that impacts and dependencies today often become financial effects later1.

Stakeholders and the value chain

The assessment is not a desktop-only exercise.

IG 1 positions stakeholder engagement as a core input: affected stakeholders — workers, communities, consumers, others touched by the company’s activities — and users of the report inform both the identification and the assessment of IROs1.

Engagement can range from direct consultation to using credible proxies (surveys, grievance data, sector research) where direct dialogue is impractical1.

Coverage is equally explicit: the assessment spans own operations and the upstream and downstream value chain1.

A retailer’s most severe impacts may sit with suppliers’ workers; a lender’s may sit in its financed portfolio.

Excluding the value chain is the fastest way to fail the impact half of the test1.

For UK subsidiaries feeding a group DMA, this is where most of the work lands — supplying site-level operational data and local stakeholder perspectives the EU parent cannot see from headquarters2.

Impact materiality vs financial materiality

The two perspectives ask opposite questions about the same topic.

Impact materiality is inside-out: does the company significantly affect the world?

Financial materiality is outside-in: does the matter significantly affect the company12?

The perspectives frequently diverge.

A company’s water abstraction in a stressed basin can be highly impact-material while barely registering financially; a carbon price on the horizon can be highly financially material while the company’s own emissions are modest1.

Because ESRS applies the union of the two tests2, both examples produce reportable matters.

They are also connected over time.

IG 1 notes that impacts and dependencies are frequently the origin of financial risks and opportunities — a severe negative impact today invites regulation, litigation, or reputational cost tomorrow1.

A well-run assessment scores the perspectives separately but reads them together.

How GRI, ISSB and ESRS define materiality

The three major framework families each take a different slice of the materiality question — which is why “are we already doing this?” has a different answer depending on what you report today.

The practical consequence: a GRI reporter entering CSRD scope keeps its impact work and adds financial scoring16; an ISSB or UK SRS reporter keeps its financial analysis and adds the impact half57; only ESRS requires the integrated assessment2.

See the ESRS guide for how the standards use the results.

Disclosing the assessment — ESRS 2 IRO-1 and IRO-2

The assessment itself is disclosed.

Under ESRS 2, IRO-1 requires a description of the process to identify and assess material impacts, risks and opportunities — methodology, thresholds, stakeholder input, value-chain coverage — and IRO-2 requires the outcomes: which disclosure requirements the company is reporting against as a result12.

This is why documentation discipline matters throughout the four steps: the scoring scales, threshold rationale, and engagement records produced in steps 1–4 become the substance of the IRO-1 narrative1.

Assessments run informally are hard to disclose and harder to assure.

Note the standards themselves are moving: Directive (EU) 2026/470 mandated simplification of the ESRS, and a revised, shortened set is under development by EFRAG48.

The double materiality architecture is retained4, so assessment work done now carries forward even as the disclosure requirements slim down.

Common pitfalls

Recurring failure modes we see in first-time assessments — each traceable to a requirement in the guidance:

• Running a single-materiality exercise twice — scoring financial materiality and relabelling stakeholder interest as “impact”, instead of assessing severity of real-world effects (scale, scope, irremediability) [1]
• Skipping the value chain — assessing own operations only, when IG 1 requires upstream and downstream coverage for both perspectives [1]
• Applying the intersection instead of the union — treating matters as material only when both lenses agree, when ESRS 1 requires either [2]
• Undocumented thresholds — scoring without pre-agreed scales and cut-offs, leaving nothing defensible to disclose under IRO-1 [1] [2]
• Ignoring likelihood rules — averaging likelihood into severity for negative human-rights impacts, where severity takes precedence [1]
• Treating the DMA as one-and-done — the assessment should be revisited as the business, value chain, and standards change [1], particularly with revised ESRS in development [4] [8]

Frequently asked questions