Stakeholder analysis is the part of the double materiality assessment (DMA) that companies most often over-engineer or under-document. Some first-wave reporters sent long surveys to hundreds of contacts and ended up with data they could not use. Others skipped stakeholders entirely and struggled to explain their impact scores to the auditor. The revised ESRS, adopted as Delegated Regulation (EU) 2026/1563, now make clear where the sensible middle ground lies.
In short: Engagement with affected stakeholders remains a key input to the DMA (revised ESRS 1 §42), but engagement you already carry out as part of your due diligence is sufficient: no separate stakeholder process for the DMA is required. NGOs, civil society organisations and trade unions can act as proxies, nature counts as a "silent" stakeholder, and the value chain can be assessed without direct supplier input using sector or regional data (§33). ESRS 2 IRO-1 only asks you to describe stakeholder consultation if you consult. A practical stakeholder analysis therefore has six steps: list your groups, map them, choose proportionate methods, ask the right questions, feed the answers into your IRO scoring and document everything for the auditor.
This guide walks you through those six steps. If you need the full DMA process first, start with our step-by-step guide to the CSRD double materiality assessment.
What the revised ESRS say about stakeholder engagement
The 2023 ESRS expected stakeholder engagement as part of the DMA, and many companies read that as "run a big survey". The revised ESRS 1 is more precise. Our article on the double materiality assessment under the revised ESRS covers all changes; these are the ones that matter for stakeholders:
| Rule | What it means in practice |
|---|---|
| Engagement with affected stakeholders is "a key input" (ESRS 1 §42) | Stakeholder perspectives must inform your impact assessment. Ignoring them is not an option. |
| Engagement from due diligence is sufficient (application requirements) | Reuse what you already do: grievance mechanisms, supplier risk analysis, works council dialogue, community meetings. No separate DMA engagement process is needed. |
| CSOs, NGOs and trade unions can act as proxies | You do not have to reach every affected person directly. Representatives can speak for groups that are hard to engage. |
| Nature may be considered a "silent" affected stakeholder | Use scientific data, environmental NGOs or expert input to represent nature. |
| Value chain without direct input from value chain actors (§33) | Average regional data, sector data or generally available information are acceptable. |
| Information available without undue cost or effort (§32) | You do not have to assess every possible IRO, and you do not have to ask every possible stakeholder. |
| ESRS 2 IRO-1: describe consultation only if you consult | If you consulted stakeholders, explain how. If you relied on other sources, describe those instead. |
| Update only on significant change (§34) | No need to repeat the full stakeholder exercise every year unless your business, value chain or methodology changes significantly. |
The direction is clear: less ritual, more relevance. Stakeholder input is expected, but it can come from existing processes and representatives, as long as you can show where it came from and how it shaped your conclusions.
Affected stakeholders vs. users of the sustainability statement
The ESRS distinguish two groups of stakeholders. Mixing them up is one of the most common mistakes in stakeholder analysis.
| Affected stakeholders | Users of the sustainability statement | |
|---|---|---|
| Who | People or groups whose interests are or could be affected by your operations and value chain | Primary users (investors, lenders, other creditors) and other users (business partners, trade unions, civil society, analysts, public authorities) |
| Examples | Own employees, value chain workers, local communities, consumers and end-users, nature (as a silent stakeholder) | Banks, investors, key customers requesting ESG data, insurers |
| Main input to | Impact materiality: severity and likelihood of your impacts | Financial materiality and the relevance of information: which risks and opportunities matter and what users need to know |
| Typical method | Due diligence dialogue, workshops, proxies, interviews | Interviews, review of ESG questionnaires and rating requests |
The groups can overlap. A trade union can represent affected employees and also be a user of your report. What matters is that you know in which role you are asking someone and which part of the DMA their input informs.
Step 1: List your stakeholder groups
Start from your value chain, not from your address book. Walk through upstream activities, your own operations and downstream activities and ask: whose interests are affected here, and who relies on information about it?
A typical list for a manufacturing company:
- Own workforce: employees, works council, temporary workers
- Value chain workers: workers at key suppliers, logistics providers
- Affected communities: neighbours of production sites, communities around raw material extraction
- Consumers and end-users: customers using your products
- Nature: represented through environmental data, NGOs or experts
- Users of the report: banks, investors, major customers, insurers
Group stakeholders rather than listing individuals. You want five to ten meaningful groups, not 80 names.
Step 2: Map stakeholders on an influence/impact grid
Not every group needs the same depth of engagement. A stakeholder map helps you decide where to invest effort. The classic version is a grid with two axes:
- Impact: how strongly are this group's interests affected by your business (or how strongly does it affect you)?
- Influence: how much influence does the group have on your decisions, and how much do you have on them?
A third dimension is useful in the DMA: knowledge. How well do you already understand this group's concerns? A group with high impact and low knowledge is where direct engagement pays off most. A group you already know well from regular dialogue may only need a validation step.
Stakeholder analysis and stakeholder relevance matrix in the Materiality Master software
The result tells you how to engage each group:
| Position on the map | Suggested approach |
|---|---|
| High impact, low knowledge | Direct engagement: interviews, workshops, or proxies if the group is hard to reach |
| High impact, high knowledge | Reuse existing dialogue (due diligence, works council, grievance channels) and validate your IRO scores |
| Low impact, high influence (often users) | Short interview or review of their ESG data requests |
| Low impact, low influence | Desk research is usually sufficient |
Step 3: Choose methods proportionately
The revised ESRS do not prescribe any method. Choose the lightest method that gives you reliable input for the IROs that are not evident.
| Method | Best for | Effort | Watch out for |
|---|---|---|---|
| Existing sources (employee survey, grievance mechanism, supplier audits, complaints) | Groups you already engage through due diligence | Low | Check that the data is recent and covers the relevant topics |
| Proxies (NGOs, trade unions, industry initiatives) | Value chain workers, communities, nature | Low to medium | Document why the proxy is credible for that group |
| Desk research (sector studies, NGO reports, scientific data) | Nature, distant value chain, top-down conclusions | Low | Cite sources so the auditor can follow |
| Short online survey | Large, reachable groups (employees, customers) | Medium | Keep it short and tied to your IRO list; long surveys produce noise |
| Interviews | Key users (bank, investors), works council, key suppliers | Medium | Prepare questions in advance and keep notes |
| Workshops | Validating IRO scores with internal experts and representatives | Medium to high | Invite the right mix; record results directly against IROs |
Two principles help:
- Top-down topics need little or no engagement. If a topic is evidently material or not material (ESRS 1 §27), a documented rationale is enough. Concentrate stakeholder input on the borderline topics, where it actually changes the result.
- Reuse before you ask. Your annual employee survey, your supplier risk analysis or the results of your grievance mechanism often already contain what you need. Asking the same people again creates fatigue without adding value.
Step 4: Ask the right questions
Whatever method you choose, the questions should be tied to your list of impacts, risks and opportunities (IROs), not to generic sustainability topics. Good questions include:
- Which of our activities affect you most, positively or negatively?
- Is our list of potential impacts complete, or are we missing something?
- How severe do you consider this impact (scale, scope, how hard to remediate)?
- How likely is it that this potential impact occurs?
- Which sustainability information do you need from us, and for which decisions? (for users)
For ready-made question sets by stakeholder group, including survey and interview templates, see our materiality assessment questionnaire.
Step 5: Feed stakeholder input into your IRO scoring
Stakeholder input is not a separate result. It is evidence for your scores. For every IRO where you engaged stakeholders, record:
- Who gave input (group, proxy, or source)
- What they said about severity, likelihood or financial relevance
- How it changed or confirmed your score
If stakeholders rate an impact as much more severe than your internal team did, that is not automatically decisive, but you need to explain why you followed or did not follow their view. Our double materiality assessment example shows how a mid-sized manufacturer combined workshops, a works council interview, NGO reports as proxies and a bank interview, and how that input fed into the scoring.
Step 6: Document for the auditor
Under limited assurance, the auditor reviews your process, not just your results. For stakeholders, a complete audit file contains:
- The stakeholder list with groups, roles (affected or user) and the reasoning behind your map
- The engagement log: which method, when, with whom, or which existing source or proxy you used
- Questions asked and a summary of the answers
- The link between input and IRO scores, including where you deviated from stakeholder views and why
- For proxies and desk research: why the source is credible for the group it represents
- The date of the last update and whether significant changes occurred since (§34)
This is also the basis for your ESRS 2 IRO-1 disclosure. If you consulted stakeholders, you describe how. If you relied on due diligence and proxies, you say so. Either way, the documentation has to support what you write. Our DMA audit checklist lists what else auditors look for.
Is there software that manages the stakeholder input process?
Yes. Spreadsheets and survey tools can collect answers, but they break the link between stakeholder input and the IRO scores it should inform. That link is exactly what the auditor asks about.
In Materiality Master, stakeholder analysis is part of the DMA itself:
- Stakeholder register and matrix: add stakeholder groups, rate interest, influence and knowledge with a justification, and the stakeholder matrix is generated automatically.
- Input directly in the assessment: invite stakeholders or their representatives into the assessment with access only to the IROs they should rate or comment on, so their input lands in the scoring instead of in a separate file.
- Link stakeholders to IROs: for every IRO, record which stakeholder was involved and how (survey, interview, workshop, proxy).
- Audit trail: every change to a score is logged with a reason, so you can show how stakeholder input shaped the result.
The help article How to conduct a stakeholder assessment in your DMA shows each step with a video, and user access management explains how to give stakeholders restricted access.
Collect stakeholder input directly in your DMA
Map your stakeholders, invite them to rate the IROs that concern them and keep every input linked to your scores, with a full audit trail. Aligned with the revised ESRS.
Common mistakes to avoid
- One big survey for everyone. Generic surveys produce generic answers. Tie questions to your IROs and the role of each group.
- Confusing affected stakeholders with users. Investors tell you about financial relevance, not about the severity of your impacts on supplier workers.
- Engaging where the result is evident. Spend stakeholder effort on the borderline topics.
- Forgetting silent stakeholders. Nature and hard-to-reach value chain workers need proxies, not omission.
- No link between input and scores. If you cannot show how input influenced a score, the engagement will not count for much in the audit.
- Repeating everything every year. Under §34, you update only when something significant has changed.
Key takeaways
- Stakeholder engagement remains a key input to the DMA (ESRS 1 §42), but due diligence engagement is sufficient; a separate DMA process is not required.
- Distinguish affected stakeholders (input on impacts) from users of the statement (input on financial relevance and information needs).
- Use an influence/impact grid to decide where direct engagement adds value; use proxies and existing sources elsewhere.
- NGOs, CSOs and trade unions can act as proxies, and nature can be treated as a silent stakeholder.
- Document who was involved, how, and how their input changed your IRO scores. Describe consultation in ESRS 2 IRO-1 if you consulted.
Ready to run your stakeholder analysis and DMA in one place? Start for free in Materiality Master or compare the plans and pricing.
