The question does not entirely match our standard answer

The question is narrower or broader than the standard answer

Usually you can still answer, with a brief explanation alongside the answer. A question rarely aligns 100% with a standard answer from the chain dossier — portals formulate questions in their own words, with their own scope. The question "do you have an ISO 27001 certificate or equivalent framework?" is broader than the answer "ISO 27001: yes/no": if there is no certificate but there is an internal policy covering the control measures in Annex A, that is relevant to this question, even if it does not trigger the standard answer about the ISO 9001 certificate or about ISO 27001 itself. The other way around: sometimes the portal asks something much more specifically than the dossier records, for example per location rather than for the entire company.

The system recognizes the question and proposes the closest answer, but the confirmation per answer is there precisely for this moment: you see the question text alongside the proposed answer and assess whether it fits. If it does not fit completely, then "no" or manual adjustment is the right choice — not bending the dossier to fit the question. For questions that differ slightly from scope 3 emissions in the EFRAG VSME basic module, or that fall just outside the ten measures in Article 21 NIS2, the same applies: the dossier provides the fact, the client assesses the match with the question asked.

Why the confirmation step catches this

The platform recognizes questions based on patterns in previously completed lists, but recognition is not a guarantee of an exact match — which is why each proposed answer is confirmed separately before it is taken over. That applies equally to identity questions that are formulated slightly differently, such as a question for the registration number which specifically requests a branch number instead of the number from the Business Register, as for substantive questions on sustainability or cybersecurity. We do not build an interpretation of the question — that remains with the client, who reads and assesses the question in the portal of his own client.