ESRS Double Materiality: Step-by-Step Methodology

Why the DMA determines everything else in the CSRD report

We reviewed nine first-cycle DMAs at mid-market entities last year. Four had no documented scoring methodology. Two assessed materiality at the topic level without breaking it down to IROs. One had a scoring methodology but set the threshold so low that every topic was material, which is the same as having no threshold at all. The pattern is clear: companies treat the DMA as a tick-box exercise rather than the analytical process ESRS 1 paragraph 38 requires.

ESRS 1 paragraph 38 requires the entity to identify material IROs through a process that considers both impact materiality and financial materiality. A sustainability matter is reportable if it’s material from either perspective, or both. The outcome is a list of material IROs that map directly to ESRS topical standards (E1 through G1). If a topic isn’t material, the entity doesn’t report against that topical standard (though ESRS E1 climate disclosures require explicit justification for omission).

When you perform CSRD assurance, the DMA is the first thing you evaluate. A poorly executed DMA means the sustainability statement may be missing entire material topics. It may include immaterial topics that waste reporting effort. Or it may have the right topics but no defensible rationale for why they were deemed material. Each of those outcomes creates assurance risk. I’d argue the third is the most common, because it’s the easiest to produce and the hardest to detect without reading the underlying scoring.

EFRAG’s Implementation Guidance (IG 1), published alongside the ESRS, provides a non-prescriptive framework for conducting the DMA. The July 2025 EFRAG draft revisions introduced a top-down approach as an alternative to the bottom-up method most companies used in the first cycle. Under the top-down approach, the entity starts with its business model, strategy, and value chain analysis, then derives relevant topics and IROs from that context. Only IROs expected to lead to material disclosures are examined in depth. EFRAG describes this as an “information materiality filter” designed to prevent resources from being spent on topics that won’t result in reportable information.

Regardless of whether you use a top-down or bottom-up approach, ESRS 1 paragraphs 38–58 define the mandatory concepts. The methodology is flexible. The requirements are not.

Step 1: map the value chain and identify stakeholders

ESRS 1 paragraph 42 requires the entity to consider its own operations, products, services, and business relationships across the upstream and downstream value chain when identifying sustainability matters. Before you can assess materiality, you need to understand where the entity operates and who it affects.

Start with the entity’s business activities. Map the major product and service lines, the geographic locations of operations, the key suppliers (upstream), and the key customers or end users (downstream). EFRAG IG 1 paragraph 175 directs entities to use the list of sustainability matters in ESRS 1 paragraph AR 16 as a starting point, but the value chain map provides the context for interpreting that list. A chemical manufacturer in Rotterdam has different relevant sustainability matters than a software company in Berlin, even if both fall within CSRD scope.

Stakeholder identification runs in parallel. ESRS 2 IRO-1 paragraph 53(b)(iii) requires the entity to describe how it consulted with affected stakeholders as part of the materiality assessment. “Affected stakeholders” means groups who are or could be affected by the entity’s activities: employees, workers in the value chain, local communities, consumers. “Users of sustainability statements” are a separate category, primarily investors and creditors.

The practical challenge is scoping stakeholder engagement proportionately. You don’t need a 200-person survey for every DMA. For a mid-market entity with a well-understood value chain, structured interviews with five or six internal experts (operations, procurement, HR, finance, EHS) combined with a targeted external consultation (key suppliers, employee representatives, a sector expert) can produce sufficient input. The AFM’s July 2024 “Ten waypoints” guidance emphasises that the depth of engagement should be proportionate to the complexity of the entity’s value chain and the nature of its impacts.

Document everything. The assurance practitioner will ask for the value chain map and the records of how stakeholder input was obtained and incorporated. If these records don’t exist, the DMA isn’t assurable.

Step 2: identify potential material sustainability matters and IROs

With the value chain mapped and stakeholders consulted, the next step is building the list of potentially material sustainability matters and their associated IROs. ESRS 1 paragraph AR 16 provides the sector-agnostic list of topics covered by the ESRS topical standards. This list is organised by environment (E1–E5), social (S1–S4), and governance (G1), with sub-topics and sub-sub-topics beneath each.

The entity should work through this list systematically, considering whether each sub-topic is relevant given its business activities, value chain, and geographic context. But the AR 16 list isn’t exhaustive. ESRS 1 paragraph 11 requires the entity to also consider entity-specific matters not covered by the topical standards. An entity with significant tax exposure in multiple jurisdictions, for example, might identify tax transparency as an entity-specific matter even though it doesn’t appear in the ESRS list.

For each relevant sustainability matter, the entity identifies the specific impacts (actual and potential, positive and negative), risks, and opportunities. The process gets granular here. “Climate change” is a topic. “Scope 1 GHG emissions from natural gas combustion at the Rotterdam production facility” is an impact. “Risk of carbon pricing increases under EU ETS Phase 4 affecting production costs by an estimated 8–12% by 2030” is a financial risk. The DMA operates at the IRO level, not the topic level.

The output of this step is a long list. For a manufacturing entity, it’s common to identify 40–80 IROs across all topics before any materiality scoring. That’s normal. The scoring in Step 3 is what reduces this to the material set.

A common shortcut that weakens the DMA is stopping at the topic level. If you assess “climate change” as material or not material without breaking it down into specific IROs, you can’t later map the material determination to specific ESRS disclosure requirements. ESRS 1 paragraph 34 requires the assessment to be performed first at the disclosure requirement level and then at the datapoint level. Starting at the topic level and staying there produces a DMA that looks complete on the surface but doesn’t connect to the reporting obligations underneath.

Why do companies keep making this mistake? Because the consulting market incentivizes it. A topic-level DMA can be delivered in two weeks at a fixed fee. An IRO-level DMA requires four to eight weeks and deeper client engagement. The fee pressure pushes toward the shortcut that ESRS was specifically designed to prevent. The standard demands granularity. The market rewards speed. That tension is the structural reason most first-cycle DMAs underperform, and it won’t resolve itself without assurance practitioners pushing back on the output quality.

Step 3: assess materiality of identified IROs

Most of the analytical work concentrates in this step, and the methodology must be documented precisely enough to survive assurance review. ESRS 1 defines separate criteria for impact materiality and financial materiality.

Impact materiality is assessed based on the severity of actual and potential impacts. ESRS 1 paragraph 44 defines severity using these dimensions for negative impacts: scale (how grave the impact is), scope (how widespread it is), and the irremediable character (how difficult it is to remedy). For positive impacts, only scale and scope apply. For potential impacts, likelihood is an additional factor combined with severity. ESRS 1 paragraph 46 states that for impacts on human rights, severity takes precedence over likelihood. Even a low-likelihood human rights impact can be material if its severity is high.

Financial materiality is assessed based on the likelihood that a sustainability-related risk or opportunity will generate financial effects, and the potential magnitude of those effects on the entity’s financial position, performance, or cash flows. ESRS 1 paragraph 49 requires the entity to consider dependencies on natural, human, and social resources regardless of whether the entity has potential impacts on those resources.

ESRS 1 doesn’t prescribe specific thresholds. This is a judgment call, and it must be documented. A common approach is to use a scoring matrix: a 1–5 scale for severity components and likelihood, with a defined cut-off score above which an IRO is considered material. The thresholds should be calibrated against the entity’s size, its sector, its risk profile, and the nature of its value chain relationships.

Reasonable practitioners disagree on where to set the threshold. Some sustainability consultants advocate a cut-off of 2.5 on a 1–5 scale, which captures more topics and produces a “safe” DMA that’s hard to challenge for omissions. Others argue for 3.5, which keeps the scope manageable but risks excluding a borderline topic that the assurance practitioner later questions. In our experience, 3.0 is the pragmatic middle ground for mid-market entities, but the number matters less than whether the rationale for choosing it is documented and internally consistent.

What actually happens in most first-cycle DMAs: the sustainability coordinator runs a two-hour workshop, asks department heads to score IROs on a spreadsheet, averages the results, and calls everything above 3.0 “material.” Nobody discusses the scoring criteria beforehand. Nobody calibrates across departments. Procurement gives everything a 2. EHS gives everything a 4. The averages look reasonable but the underlying scores are incoherent. The assurance practitioner asks why water discharge scored 3.1 (material) and water consumption scored 2.9 (not material), and nobody can explain the one-point difference.

Internal expert workshops are the most common scoring method for mid-market entities. Each relevant department scores the IROs within their domain (procurement scores supply chain impacts, HR scores workforce impacts, EHS scores environmental impacts). External stakeholder input from Step 1 feeds into the impact materiality scores. Financial materiality scores come from finance and risk management functions, using existing enterprise risk management data where available.

After scoring both dimensions, the results are consolidated. An IRO is material if it exceeds the threshold on either the impact or financial dimension. You don’t need to be material on both sides. This is the “double” in double materiality: the two perspectives are independent gates, not a combined test.

EFRAG’s July 2025 draft revisions introduced additional flexibility here. Companies may now also apply a top-down “information materiality filter” that asks whether an IRO, even if it crosses a scoring threshold, would actually lead to material disclosures in the sustainability statement. This prevents the situation where a technically material IRO generates disclosure requirements with no meaningful content. The practical effect is that the revised approach allows companies to focus resources on the IROs that genuinely drive reporting outcomes.

Step 4: document and connect to ESRS disclosure requirements

The DMA produces a final list of material IROs. That list is the bridge between the assessment and the sustainability statement. For each material IRO, the entity must identify which ESRS topical disclosure requirements apply.

ESRS 2 IRO-1 paragraph 53 specifies what the entity must disclose about the DMA process itself: a description of the process to identify material IROs, the methodologies and assumptions used, how stakeholders were involved, and how the entity determined which sustainability matters are material. ESRS 2 SBM-3 requires disclosure of the material IROs identified and their interaction with the entity’s strategy and business model. ESRS 2 IRO-2 requires disclosure of the entity’s disclosure requirements in ESRS covered by the sustainability statement.

Each material IRO maps to one or more topical ESRS standards. If “Scope 1 GHG emissions from production” is a material impact, the entity reports under ESRS E1 (Climate Change), specifically E1-1 (Transition plan), E1-4 (GHG emission reduction targets), E1-6 (Gross Scope 1, 2, 3 emissions), and other related disclosure requirements as applicable. The mapping isn’t always one-to-one. A single material impact can trigger disclosures across multiple standards if it has environmental, social, and governance dimensions.

For each topical standard, ESRS 1 paragraph 34 requires a second-level assessment: which specific disclosure requirements and datapoints within that standard are material? The entity may omit individual datapoints within a material topic if they aren’t relevant to its specific IROs. But the burden of justification sits with the entity, and ESRS 1 paragraph 35 requires the omission rationale to be documented.

The documentation package for the DMA should include, at minimum: the value chain map, stakeholder identification and engagement records, the full IRO list (including non-material items with rationale), the scoring methodology and threshold definitions, the individual IRO scores, the consolidated materiality determination, and the mapping from material IROs to ESRS disclosure requirements. This is what the assurance practitioner will review. Missing any of these components creates a documentation gap that will slow down the assurance engagement and may result in qualified findings.

One detail that entities frequently overlook: the DMA isn’t a one-time exercise. ESRS 1 paragraph 58 requires the entity to update the materiality assessment when circumstances change. A new regulation, a significant acquisition, a supply chain disruption, or a change in the entity’s strategy can all affect which IROs are material. The documentation should include a section describing how the entity plans to monitor for changes that would trigger a reassessment. The assurance practitioner will expect evidence that the entity considered whether the prior-year DMA remains valid, especially from the second reporting year onward.

Worked example: running a DMA for a mid-market manufacturer

Client scenario: Dekker Chemie B.V., a Dutch specialty chemicals manufacturer based in Dordrecht. 1,200 employees, €480 million annual revenue, with production facilities in the Netherlands and Belgium. Supply chain includes raw material suppliers in Germany and Southeast Asia. Products sold to industrial customers across the EU.

Step 1. Value chain mapping. Dekker’s sustainability coordinator maps the value chain in collaboration with procurement, operations, and sales. Upstream: 42 direct suppliers, of which 8 are in Southeast Asia (palm oil derivatives, mineral additives). Downstream: 180 industrial customers, predominantly in Germany, France, and Benelux. Key affected stakeholders identified: own workforce (1,200 employees across two sites), workers at Southeast Asian suppliers (estimated 3,000+), local communities near the Dordrecht production facility (air quality, water discharge), and industrial customers dependent on product safety data.

Documentation note: Record the value chain map with supplier tiers, geographic breakdown, employee counts, and stakeholder groups. Cite ESRS 1 paragraph 42 and ESRS 2 IRO-1 paragraph 53(b). Attach stakeholder engagement records (internal workshop minutes, supplier questionnaire summary, community consultation notes).

Step 2. IRO identification. Working through ESRS 1 AR 16, the team identifies 52 potential IROs across eight topics. Climate change generates 11 IROs (Scope 1 emissions from chemical processes, Scope 2 from electricity, Scope 3 from raw material transport, transition risk from EU ETS pricing, physical risk from flooding at the Dordrecht site, and others). Pollution generates 9 IROs (air emissions of VOCs, wastewater discharge, product chemical safety). Own workforce generates 8 IROs (health and safety in production, training, pay equity). Workers in the value chain generates 6 IROs focused on labour conditions at Southeast Asian suppliers. The remaining 18 IROs span water, biodiversity, consumers, and business conduct.

Documentation note: Record each IRO with a unique identifier, the ESRS topic and sub-topic it relates to, whether it’s an impact/risk/opportunity, and whether it’s actual or potential. This register is the master document for the DMA.

Step 3. Materiality scoring. Dekker uses a 1–5 severity scale for impact materiality (1 = negligible, 5 = severe/irreversible) and a 1–5 magnitude scale for financial materiality. Threshold for materiality: any IRO scoring 3.0 or above on either dimension. Scores are generated through a two-day internal workshop with department heads, calibrated against external benchmarks (sector peer reports, ECHA chemical safety data, EU ETS price projections of €80–100/tonne by 2030).

Results: 28 of 52 IROs score above the 3.0 threshold. Climate change (7 of 11 IROs material), pollution (6 of 9), own workforce (5 of 8), and workers in the value chain (4 of 6) are the dominant material topics. Water and marine resources (3 of 5 IROs material) also crosses the threshold due to the Dordrecht facility’s water-intensive processes. Biodiversity, consumers, and business conduct each have one or zero material IROs.

Documentation note: Record each score by dimension (scale, scope, remediability for impacts; likelihood and magnitude for financial risks/opportunities). Include the threshold definition and the calibration methodology. Flag any IRO where impact and financial materiality diverge significantly (e.g., a severe environmental impact with low short-term financial consequence). These are the IROs where documentation quality matters most, because they test the independence of the two materiality perspectives.

Step 4. Mapping to ESRS. The 28 material IROs map to five ESRS topical standards: E1 (Climate Change), E2 (Pollution), E3 (Water and Marine Resources), S1 (Own Workforce), and S2 (Workers in the Value Chain). For each standard, the team reviews the specific disclosure requirements and datapoints, identifying which are relevant to Dekker’s material IROs. For example, ESRS E2-4 (Pollutants) requires disclosure of air, water, and soil pollutants. Dekker’s material IROs include VOC air emissions and wastewater discharge, so E2-4 is relevant. ESRS E2-5 (Substances of concern) is also triggered by Dekker’s chemical product portfolio.

Documentation note: Create a mapping table linking each material IRO to specific ESRS disclosure requirements and datapoints. This table becomes the specification for the data collection process and the reporting structure of the sustainability statement. Provide it to the assurance team at the start of the engagement. ESRS 2 IRO-2 requires a version of this mapping to be included in the sustainability statement itself.

The complication

Three weeks after the DMA is finalised, Dekker announces the acquisition of a small adhesives manufacturer in Poland with 180 employees and a different chemical profile. The acquisition adds new supply chain relationships, a new jurisdiction, and potential new environmental impacts (different VOC profiles, different waste streams). Does the DMA need to be redone?

ESRS 1 paragraph 58 requires the entity to update the assessment when circumstances change. An acquisition is a change. But a full redo is disproportionate. The sustainability coordinator reviews the existing IRO register against the acquired entity’s operations. Result: two new IROs added (adhesive-specific chemical safety under E2, and Polish labour regulation compliance under S1), three existing IRO scores increased (Scope 1 emissions now includes the Polish facility, supply chain worker conditions now includes Polish suppliers, and water consumption increases by 15%). No existing material IROs are reclassified as non-material. The net effect is two additional material IROs and updated scores on three existing ones. Total effort: four days, not six weeks.

Documentation note: Record the trigger event (acquisition), the date, the review performed, the IROs added or rescored, and the rationale for not performing a full reassessment. This creates the audit trail ESRS 1 paragraph 58 requires and demonstrates that the entity’s DMA is a living process, not a one-time exercise.

Dekker’s original DMA took approximately six weeks from value chain mapping to final documentation. The most time-intensive step was IRO identification and scoring (approximately 60% of total effort). The resulting documentation package is 34 pages, including the methodology description, stakeholder engagement records, the full IRO register with scores, and the disclosure requirement mapping.

Practical checklist for your next engagement

Common mistakes in double materiality assessments

• EFRAG IG 1 flags that entities frequently assess materiality at the topic level (“climate change is material”) without breaking it down to individual IROs. This produces a DMA that appears complete but doesn’t connect to specific ESRS disclosure requirements at the level ESRS 1 paragraph 34 requires.
• The AFM’s July 2024 “Ten waypoints” guidance found that some entities treated stakeholder engagement as a retrospective validation exercise rather than an input to the assessment. Sending stakeholders a pre-filled materiality matrix for confirmation doesn’t satisfy ESRS 2 IRO-1 paragraph 53(b)(iii), which requires consultation to inform the identification of material IROs.
• Assurance practitioners report that the most common documentation gap is the absence of documented rationale for topics assessed as not material. ESRS 1 paragraph 38 requires the entity to identify material IROs, but the assurance practitioner also needs to understand why the entity concluded that certain topics are not material. If the “not material” determination has no recorded reasoning, it’s effectively unassurable.
• Combining impact and financial materiality scores into a single blended number before determining materiality. We’ve seen this on roughly half the DMAs we’ve reviewed. A sustainability matter that scores 4.5 on impact and 1.2 on financial gets averaged to 2.85, which falls below a 3.0 threshold. But under ESRS 1, that matter is material because it exceeded the threshold on impact alone. The two dimensions are independent gates. Blending them defeats the entire purpose of double materiality.

Related content

• Double materiality. Explains the concept, including the ESRS 1 definitions of impact materiality and financial materiality and how they interact.
• ESRS IRO Compliance Checker. Free tool for mapping your material IROs to the correct ESRS topical disclosure requirements and datapoints, replacing the manual mapping in Step 4.
• CSRD Compliance 2026: What Auditors Need to Know. Companion post covering CSRD assurance requirements after Omnibus I, including what the AFM expects from firms performing their first sustainability assurance engagements.

Related ciferi content

Related guides:

Put audit concepts into practice with these free tools:

Frequently asked questions

Further reading and source references

• ESRS 1, General Requirements: paragraphs 38–58 define the mandatory double materiality concepts.
• EFRAG Implementation Guidance (IG 1), Materiality Assessment: non-prescriptive framework for conducting the DMA.
• ESRS 2, General Disclosures: IRO-1, SBM-3, and IRO-2 specify what entities must disclose about the DMA process and outcomes.
• AFM July 2024, Ten Waypoints for CSRD: Dutch regulator guidance on proportionate stakeholder engagement and DMA quality.