ESG reporting software gets the budget and attention, but a report is only as trustworthy as the data behind it. Ninety-six percent of finance leaders report problems with the integrity of the sustainability data they receive. Fixing the data layer — collection, integration, quality, lineage — matters more than picking a reporting tool.
Most ESG programs buy a reporting tool first and discover the data problem second. That order is backwards. A dashboard can format a number beautifully and still be wrong, because the number came from a spreadsheet someone updated by hand three months ago, or an estimate nobody documented, or a supplier who never replied to the data request. The tool didn’t cause that. It just made the output look more finished than the underlying data actually is.
The 96% Problem
According to EY’s 2024 Global Corporate Reporting Survey, 96% of finance leaders report some problem with the integrity of the nonfinancial (ESG/sustainability) data they receive for reporting. EY’s Dr. Velislava Ivanova attributes this to “the highly complex nature of sustainability topics and the overwhelming amount of reliable data required” — not a shortage of reporting software, but a shortage of trustworthy inputs. The same survey found 55% of finance leaders worry their industry’s sustainability reporting risks being read as greenwashing, specifically because disclosures aren’t consistently “backed up by the necessary due diligence, data and processes.”
That’s the gap a reporting tool alone cannot close. If facilities data lives in one spreadsheet, cloud billing data lives in a different export, and supplier emissions data lives in whatever a vendor happened to send back last quarter, no dashboard fixes that — it just displays it.
What “The Data Layer” Actually Means
Split any ESG reporting stack into two layers. The reporting layer is what most buyers evaluate first: dashboards, disclosure templates, framework mapping, export formats. The data layer sits underneath it and rarely gets the same scrutiny: the source systems (cloud billing APIs, facilities and energy meters, procurement records, HR systems, supplier questionnaires), the collection mechanism that pulls from them, and the model that ties each raw record to a specific disclosure metric.
A reporting layer can only be as good as the data layer feeding it. If a cloud carbon figure comes from a manual CSV export someone remembers to run quarterly (a common pattern in Fruggr’s own conversations with ESG and sustainability managers), no reporting template turns that into an auditable number. Fixing the data layer usually means automating collection at the source — Fruggr’s own FinOps/GreenOps module, for example, pulls cost and carbon data directly from AWS, Azure, and GCP billing APIs rather than relying on a manual export, which is exactly the kind of source-system connection a data layer needs before a reporting layer can be trusted.
Scope 3 Is the Data Layer’s Hardest Test
If Scope 1 and 2 (direct emissions and purchased energy) are hard to collect well, Scope 3 — everything in the value chain, from purchased goods to business travel to how customers use what a company sells — is where most data layers break entirely. The GHG Protocol’s own Scope 3 standard states plainly that “the majority of total corporate emissions” typically fall under Scope 3 (the standard does not give a single precise percentage, so treat any specific figure you see elsewhere as worth verifying rather than a fixed fact).
The structural reason Scope 3 is harder isn’t ambition — it’s authority. A company has direct control over its Scope 1 and 2 data because the meters and systems are its own. For Scope 3, the GHG Protocol’s Scope 3 guidance is explicit that a company has only “influence,” not control, and that “not all of a company’s relevant suppliers may be able to provide GHG inventory data” at all. A manufacturer’s Scope 3 inventory can span hundreds of suppliers, each with a different reporting maturity, none of them obligated to answer a data request on the reporting company’s timeline.
That’s why treating Scope 3 as a data-collection project with its own owner, timeline, and escalation path — not a line item to fill in during report season — is the single biggest lever most enterprises have to improve their overall data layer.
Five Data-Quality Dimensions That Actually Get Tested
The Enterprise Data Management Council’s data-quality dimensions, as applied to ESG data, give a concrete checklist for where a data layer typically fails:
| Dimension | What breaks without it | A concrete ESG example |
|---|---|---|
| Completeness | Gaps get silently filled with estimates nobody documented | A supplier list where 40% never returned a data request, and the gap isn’t flagged in the final report |
| Accuracy | Numbers look plausible but don’t tie back to a source record | A facilities energy figure copied from last year’s report instead of this year’s meter reading |
| Timeliness | Reported data reflects last year’s operations, not this year’s | A cloud carbon figure from a Q1 export used to represent a full fiscal year |
| Consistency | The same metric is calculated differently across business units | One subsidiary uses location-based grid factors, another uses market-based — and nobody reconciles the difference |
| Traceability / lineage | Nobody can show an auditor where a number came from | A disclosed emissions total that can’t be traced past a spreadsheet cell to the original meter or invoice |
Every one of these is a data-layer problem, not a reporting-layer problem — no dashboard fixes an incomplete supplier list or an undocumented estimate.
Traceability: What “Audit-Ready” Actually Requires
CSRD reporting currently requires limited assurance, with the framework designed to move toward reasonable assurance over time — a higher bar for evidence. EFRAG’s own Value Chain Implementation Guidance confirms that where primary data isn’t available, a company can use estimates, proxies, and sector data — but only if it documents the methodology and resulting accuracy level in its “basis for preparation.” In practice, that means an estimate isn’t disqualifying; an undocumented estimate is.
The exact evidence an assurance provider will expect can vary by engagement and is still evolving as assurance standards like ISSA 5000 mature — treat the general principle above as reliable, but verify the specific documentation your auditor expects before you finalize a reporting process around it.
Building the Data Layer: Where to Start
- Map data owners beyond the sustainability team. Facilities, IT/cloud, procurement, and HR each hold pieces of the picture; a data layer that depends on one team chasing five others every quarter doesn’t scale.
- Automate collection at the source where systems allow it. A direct connection to a cloud billing API or a facilities meter beats a recurring manual export every time — it removes the step where errors get introduced.
- Document every estimate’s methodology, not just the number. Per EFRAG’s guidance above, an estimate with a documented basis survives an audit; an unexplained one doesn’t.
- Build lineage from source record to disclosed metric, not just from spreadsheet to report — the traceability gap is usually the first thing an assurance review finds.
- Treat Scope 3 supplier engagement as its own project, with a named owner and a realistic timeline, rather than a checkbox inside the reporting cycle.

Key takeaways
- Ninety-six percent of finance leaders report integrity problems with their sustainability data (EY, 2024) — a reporting tool alone does not fix a data-collection problem underneath it.
- Scope 3 is structurally the hardest data-layer test because a company has only influence, not control, over supplier data — treat it as its own data-collection project with a dedicated owner.
- The five EDMC data-quality dimensions — completeness, accuracy, timeliness, consistency, and traceability — give a concrete checklist for where most ESG data layers actually break.
Next step: See how Fruggr’s ESG IT reporting module automates data collection directly from source systems — building traceability in from day one instead of bolting it on after the report is due.
FAQ
Reporting software formats and outputs disclosures against a framework like CSRD or ISSB. Data management is the layer underneath it — where the numbers actually come from, how they’re collected, and whether they can be traced back to a source record. A reporting tool can be excellent and still produce unreliable output if the data layer feeding it is weak.
Because a company controls its own facilities and energy contracts (Scope 1 and 2) but only has influence — not authority — over its suppliers and value chain (Scope 3). Not every supplier can or will provide emissions data on request, which is why Scope 3 typically requires a dedicated data-collection effort rather than a standard reporting-cycle task.
Not by itself. Guidance from EFRAG allows estimates and proxies where primary data isn’t available, provided the methodology and resulting accuracy are documented in the disclosure’s basis for preparation. An undocumented estimate is the problem, not an estimate on its own.
Often, yes. The data layer and the reporting layer are separable: a company can improve source-system connections, documentation, and lineage without necessarily switching its reporting tool — though a reporting tool that doesn’t support direct system integrations will eventually limit how far that improvement can go.
This article was drafted with AI assistance and reviewed by our editorial team before publishing.