Conduct a DMA in four steps

The short version

A DMA has four steps: context mapping (what does the undertaking do, who does it affect), IRO identification (what sustainability matters arise — actual and potential, across the value chain), scoring (impacts on severity and likelihood; risks and opportunities on magnitude and likelihood), and reporting (IRO-1 narrative, IRO-2 structured table). The four steps are the same for Wave 1, Wave 2, and voluntary filers. The depth applied at each step varies.

Step 1: Context mapping

EFRAG IG 1 identifies three context dimensions: the undertaking itself (activities, business model, strategy, geographic footprint), the value chain (upstream suppliers, downstream customers, end use), and affected stakeholders (workers, communities, consumers, investors, nature). IG 1 §2

Document the business model first — not a corporate-communications version, an operational one. What does the undertaking make, sell, or enable? Where are the physical facilities, the workforce, the critical inputs, the principal customers? A useful test: could an incoming auditor, reading only your context map, understand what a material matter would look like for this organisation? If not, the mapping is too thin.

Segment the value chain into own operations, upstream, and downstream. Each segment has its own material profile. A consumer electronics firm may have limited direct environmental impact but significant upstream (mining, component manufacturing) and downstream (product use energy, end-of-life waste) footprints. Value-chain segmentation must survive to the IRO-1 disclosure — an IRO can be material upstream and immaterial in own operations, and that distinction matters.

Identify stakeholders in two groups. Affected stakeholders — those whose interests are positively or negatively affected by the undertaking’s activities (workers, community residents, value-chain workers, nature as a silent stakeholder). Users of sustainability information — those making economic decisions on the basis of the report (investors, lenders, credit rating agencies, business partners). Both groups inform materiality, but differently: affected stakeholders inform impact materiality; users of information inform financial materiality.

Step 2: IRO identification

IRO identification is a prompted exercise, not an exhaustive inventory. Use AR 16 as a starting frame, then extend to entity-specific and sector-specific matters. The taxonomy covers ten topical areas (climate, pollution, water, biodiversity, resource use, own workforce, value-chain workers, affected communities, consumers, business conduct) broken into sub-topics.

Three structural disciplines govern this step:

Separate impacts from risks from opportunities. The same sustainability matter can generate all three, but they assess differently. A climate transition exposure generates a potential negative impact on communities near fossil fuel operations, a risk of stranded assets, and an opportunity from renewable investment. Each thread is identified and assessed separately.

Distinguish actual from potential. IG 1 requires both. Actual impacts are those currently occurring; potential impacts are those that could plausibly occur given the undertaking’s activities. Potential impacts are scored on their likelihood as well as their severity. Actual negative impacts weight more heavily than equivalently severe potential impacts.

Attach a value-chain position to each IRO. An upstream water-stress risk and a downstream product-use emissions impact are separate IROs even if both relate to the same topic. The assessment cannot collapse them.

Step 3: Scoring

Scoring applies the materiality test to each IRO. IG 1 sets out the mechanics. IG 1 §4

For impacts:

• Scale — how severe would the impact be if it occurred? A fatality is scale 5; a minor temporary inconvenience is scale 1.
• Scope — how many people or how much environment affected? Not the same as scale. An impact can be severe for a few (high scale, narrow scope) or moderate for many (medium scale, wide scope).
• Irremediability — can the impact be undone? Permanent loss is irremediable; temporary disruption is largely remediable. IG 1 treats irremediability as a severity amplifier.
• Likelihood — for potential impacts, how probable is occurrence? Actual impacts have likelihood 1 by definition.

For risks and opportunities:

• Magnitude — the financial effect, defined in terms of the undertaking’s financial performance, position, and cash flows. Thresholds set against specific metrics (revenue, EBITDA, capex) rather than absolute euros.
• Likelihood — assessed over a defined time horizon. Short, medium, and long term are each assessed independently for each IRO. Averaging across horizons is non-compliant.

The threshold is where judgement gets most contested. IG 1 does not prescribe a single threshold value — it requires the undertaking to set and document a threshold that produces a defensible conclusion. Typical practitioner approaches: the top quartile of scored IROs; IROs scoring above a fixed point on the chosen scale; any IRO scored above the threshold on either dimension. All are acceptable if documented.

Management override is permitted and expected in specific cases — for instance, when an IRO scores just below threshold but addresses a stakeholder concern that has dominated recent engagement. The override must be documented; it is not a licence for arbitrary inclusion or exclusion.

Step 4: Reporting under ESRS 2 IRO-1 and IRO-2

IRO-1 is titled “Description of the process to identify and assess material impacts, risks and opportunities.” It asks the undertaking to disclose how the assessment was conducted — inputs used, stakeholder engagement, methodological choices, thresholds applied.

The disclosure is narrative, not tabular. Expected length is typically 800 to 2,500 words depending on undertaking size. Structure should cover: overview of the process and the four-step framework followed; sources of evidence used; how actual and potential impacts were distinguished; how severity and likelihood were assessed; how stakeholder input was incorporated; how thresholds were set and applied; and what oversight or review was conducted.

The IRO-1 narrative should be readable to a non-specialist audience. Auditors use it to assess the defensibility of the methodology; investors use it to understand the rigour of the process. It is the single most-read element of the sustainability statement for sophisticated users.

IRO-2 is titled “Disclosure Requirements in ESRS covered by the undertaking’s sustainability statement.” It is a structured list showing which sustainability matters were identified as material and which disclosure requirements consequently apply. For each ESRS topical standard (E1 through G1), the undertaking states whether the topic is material, and if so, which sub-topics and datapoints are disclosed. Matters assessed and determined non-material are noted with brief rationale.

IRO-2 is where the cycle converges with the rest of the sustainability statement. It is the index that tells users what to find where. Its completeness is what an auditor checks first.

Common failure modes across the cycle

Across first-cycle assessments, three recurring issues account for most audit findings:

• Thin context mapping. IROs scored without adequate documentation of the business model, value chain, or stakeholder context. The scoring becomes defensible in isolation but the methodology cannot be evaluated by a third party.
• Value-chain collapse. Upstream and downstream IROs merged with own-operations IROs into a single aggregate score, losing the distinctions IG 1 requires.
• Threshold opacity. A material/non-material cut-off is applied without documenting where the line was drawn or why. An auditor asking “why is this IRO below the line and that one above?” receives an evasive answer.

All three are recoverable with discipline. None can be recovered after the fact — the evidence either exists in the methodology or it does not.

Foire aux questions