In the previous article, The ESRS Reset: What Really Changed and What Didn’t, we examined how the November 2025 ESRS update simplified the reporting architecture without dismantling its underlying logic. Nowhere is that simplification more debated than in the context of double materiality.
Across sustainability teams, one question continues to resurface: Do we still need to identify and score every single impact, risk and opportunity?
For many organizations, double materiality became synonymous with exhaustive Impacts, Risks and Opportunities (IRO) inventories, scoring matrices and extensive bottom-up workshops. The process often expands into hundreds of line items, each assessed numerically to demonstrate completeness.
This now common interpretation is rarely challenged.
The revised ESRS clarifies something important.
The Misconception: DMA as Mandatory Exhaustive Scoring
In practice, many teams approached Double Materiality Assessment (DMA) as a mechanistic exercise. Their logic was straightforward as they identified every conceivable IRO, scored them individually, aggregated results and let the numbers determine materiality.
While structured, this approach often led to methodological over-engineering. Workshops became longer. Matrices grew more complex. Numerical scales multiplied. The emphasis shifted from demonstrating coverage to explaining relevance.
The November 2025 clarification does not prohibit bottom-up scoring. But it does not require it to be a default starting point either. That distinction matters.
What the Revised ESRS Clarifies
The updated standards explicitly allow materiality conclusions to be reached at the topic level. A company may determine that a topic is material or not material based on structured analysis, without necessarily scoring every individual IRO beneath it.
Top-down analysis is a legitimate primary approach that examines the business model, strategy, sector characteristics, geographic exposure, and value chain structure. If materiality is evident based on this structured screening, further bottom-up scoring is not automatically required.
Conversely, if materiality is not clear from that top-down assessment, deeper IRO-level analysis becomes necessary. In other words, bottom-up work is conditional, not automatic.
The same applies to quantification. Numerical scoring models are a methodological choice. They can be helpful, but they are not a regulatory obligation. Qualitative analysis may be sufficient if based on reasonable and supportable information and grounded in coherent reasoning.
What This Means in Practice
For practitioners, the implication is significant but nuanced. Double materiality doesn’t have to start with an exhaustive catalogue of impacts, risks and opportunities. It can begin with a structured, evidence-based analysis of how the company operates, where it creates value and where it is exposed to sustainability-related effects. In most cases, material topics can be identified at this level. Only when ambiguity remains is a deeper IRO-level assessment required. While this reduces artificial complexity, it does not eliminate accountability.
The goal is not exhaustive scoring, but proportionate judgment.
The Real Risk: Oversimplification Without Explanation
There is, however, an important balance to strike.
The risk under the revised ESRS is not simplification. Instead, it’s the risk of asserting materiality or non-materiality without explaining why.
A defensible DMA now depends more on clear reasoning and less on numerical precision. Companies must articulate how they considered impacts, risks and opportunities, what information they used, and how they reached their conclusions. The narrative must align with disclosures and reflect the company’s actual risk profile and business model.
Judgment is allowed. But it must also be documented.
What Comes Next
Once a topic is assessed as material, another common question emerges: Does a material topic automatically require reporting every datapoint associated with it?
In the next article, we will examine how materiality continues to operate at the datapoint level and why a material topic does not necessarily mean reporting everything.
