CSRD data point inventory How should teams prepare ESRS disclosure requirements for reporting?

Start from ESRS disclosure requirements and data points, then add company-specific reporting controls. EFRAG IG 3 describes a data point as a clearly separable and specific piece of information required by ESRS Disclosure Requirements, generally at paragraph, subparagraph, or sub-subparagraph level.

The inventory should therefore track the ESRS standard, Disclosure Requirement, paragraph reference, short data point description, data type, materiality status, phase-in or conditional status where applicable, owner, source system, evidence location, consolidation boundary, value-chain dependency, review status, and XBRL readiness.

Apply the correct materiality status to each ESRS row. EFRAG IG 3 distinguishes data points that are mandatory irrespective of materiality from data points that depend on the materiality assessment, and EFRAG IG 1 explains that materiality of information applies at the Disclosure Requirement or data point level.

A useful inventory has an explicit status for each row: always reported, reported because the related matter or information is material, omitted because the metric or information is not material, voluntary, conditional or not applicable, entity-specific, phased in, estimated, or unavailable with an explanation. The status should point back to the double materiality assessment, thresholds, and impact, risk and opportunity (IRO) conclusions. Track data availability separately: missing data does not itself permit omission of a required disclosure.

Evidence ownership should follow the data, not the reporting calendar. A data point about workforce headcount, greenhouse gas emissions, supplier policies, remediation channels, or financial effects usually needs a different process owner, source system, control, and approval trail.

For each data point, assign one accountable evidence owner and name the contributors who supply inputs. The owner should be able to explain the method, source, calculation, estimate, approval, and change history for the reported value or narrative. If a data point depends on value-chain information, record the supplier, customer, investee, public source, proxy, or sector-data basis used. Under Directive (EU) 2026/470, undertakings with no more than 1,000 average employees may qualify as protected undertakings and decline CSRD-purpose requests that exceed the information in the voluntary standards.

EFRAG IG 3 can reduce rework by helping teams structure human-readable reports for later digitalisation. It cannot serve as the basis for machine-readable reporting because it is not the ESRS XBRL Taxonomy.

Add XBRL-readiness fields during preparation: expected XBRL element or tag family, data type, unit, period type, dimensions or breakdowns, table structure, narrative location, continuation risk, entity-specific extension need, validation status, and reviewer comments. For narrative disclosures, keep the text granular enough that a tag can be applied without spreading the same data point across unrelated sections. Directive (EU) 2026/470 states that undertakings are not required to mark up sustainability reporting until the relevant marking-up rules are adopted.

An assurance-ready CSRD data point inventory is traceable, versioned, and reviewable. It should show the reporting framework applied, the disclosure period, the ESRS requirement, the materiality conclusion, the evidence owner, the control performed, the evidence retained, and any unresolved limitation or estimate.

CSRD introduced assurance of sustainability reporting, and the assurance process needs a record that can be tested. Resolve undocumented overrides, calculations without source links, unsupported narrative claims, and source records that cannot be tied to the final statement. Escalate unresolved evidence gaps before sign-off; a missing evidence link does not itself permit omission of a required disclosure.