Short answer
The answer, before the reasoning
IFRS S1 and IFRS S2 do not require perfect data before useful sustainability-related financial information can be reported. Reasonable estimates, proxies, modelled data and ranges can be necessary and appropriate when direct measurement is unavailable, provided the resulting information is faithfully represented, material methods and assumptions are explained, and significant measurement uncertainty is transparent.
Reporting teams should distinguish uncertainty about future outcomes from uncertainty in measuring an amount; document value-chain gaps and data hierarchy; control models and factors; and maintain an estimate register showing sources, assumptions, sensitivity, ownership, review and an improvement plan. A data gap is usually a reason to estimate and disclose limitations—not automatically a reason to omit information.
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Why estimates are normal—and still need strong controls
Many sustainability-related metrics cannot be measured directly. Scope 3 greenhouse gas emissions rely on supplier information, activity data, emission factors and secondary data. Physical-risk exposure may depend on hazard models and geospatial assumptions. Anticipated financial effects often combine scenarios, financial planning and uncertain management responses. Even apparently simple workforce or operational metrics can depend on boundary, cut-off and source-system judgements.
IFRS S1 explicitly recognises that reasonable estimates are an essential part of reporting and do not undermine usefulness when accurately described and explained. The reporting challenge is therefore not to eliminate estimation. It is to select a method that faithfully represents the disclosure objective, identify the most significant uncertainty, avoid false precision, and leave an evidence trail that supports review and year-on-year improvement.
Quick orientation
- Applies to
- Reported amounts, ranges and modelled information under IFRS S1 and IFRS S2, especially value-chain and forward-looking disclosures.
- Primary decision
- Whether an estimate is usable, how it should be measured and controlled, and what uncertainty and limitations must be disclosed.
- Key sources
- IFRS S1 paragraphs 74–82 and B50–B54; IFRS S2 18–22, 29(a), B19 and B38–B57.
- Common confusion
- Outcome uncertainty concerns what may happen; measurement uncertainty concerns how uncertain the reported amount is. One disclosure can contain both.
1. Outcome uncertainty and measurement uncertainty are not the same
The issued Standards use “measurement uncertainty” for uncertainty affecting amounts reported. Official ISSB effects-analysis and educational materials also help explain a separate practical idea: uncertainty about the outcome of a possible future event. The distinction is useful because it changes the evidence, model and disclosure needed, but it should not be presented as two formal accounting categories with identical requirements.
Figure 1. Outcome uncertainty and measurement uncertainty can overlap, but they call for different questions and explanations.
In practice
| Dimension | Outcome uncertainty | Measurement uncertainty |
|---|---|---|
| Core question | Will a future event occur, when, and with what consequences? | How uncertain is the amount because it cannot be measured directly? |
| Examples | Future carbon price; severity and timing of a flood; technology adoption; customer demand shift; policy response; management adaptation. | Incomplete supplier emissions; modelled hazard exposure; an emission factor; an estimated affected-asset percentage; a range of anticipated financial effects. |
| Typical inputs | Scenarios, forecasts, probabilities, time horizons, strategic responses and alternative pathways. | Source data, proxies, measurement techniques, models, factors, assumptions, approximations and validation evidence. |
| Disclosure emphasis | Possible outcomes, resilience, sensitivity, assumptions and implications for strategy and prospects. | Amounts with high uncertainty; sources of uncertainty; assumptions, approximations and judgements; sensitivity, range and changes. |
| Can they overlap? | Yes. A future climate event can be uncertain and the amount assigned to its potential financial effect can also be uncertain. | Yes. The model may depend on uncertain future events and uncertain data or methods. |
2. What IFRS S1 requires about measurement uncertainty
IFRS S1 requires information enabling users to understand the most significant uncertainties affecting amounts reported. The entity identifies amounts subject to a high level of measurement uncertainty and explains the sources of that uncertainty and the assumptions, approximations and judgements used in measuring each amount. The focus is on estimates requiring the most difficult, subjective or complex judgements—not on attaching boilerplate to every estimated datapoint.
In practice
| Requirement | Practical implementation | Evidence |
|---|---|---|
| Identify high-uncertainty amounts | Use a consistent screening criterion based on materiality, complexity, sensitivity, data quality and degree of judgement. | Estimate register, materiality assessment and reviewer challenge. |
| Explain sources of uncertainty | Identify dependence on future outcomes, measurement technique, value-chain data quality, model structure, factors and boundary. | Data lineage, model documentation and data-gap log. |
| Explain assumptions, approximations and judgements | Name the important assumptions and why they were selected, rather than stating only that estimates were used. | Assumption register, methodology note and approval. |
| Consider sensitivity and range | Explain how the amount changes under reasonably possible alternatives where useful; ranges can communicate uncertainty better than a false point estimate. | Sensitivity results, scenario runs and range-selection rationale. |
| Explain changes | Where uncertainty remains unresolved, describe changes to past assumptions and their effect. | Change log, comparative analysis and method-control record. |
3. A data hierarchy for estimates and proxies
A practical hierarchy begins with the most direct and specific information reasonably available, but does not treat one data type as universally superior. IFRS S2’s Scope 3 framework lists characteristics to prioritise without placing them in a rigid order: direct measurement, data from specific value-chain activities, timely data, information representative of the relevant jurisdiction and technology, and verified data. Trade-offs should be documented.
In practice
| Data level | Illustrative source | Strength — Limitation to disclose or control |
|---|---|---|
| 1. Direct measurement | Metered fuel use; directly measured emissions; measured asset condition. | Close to the underlying activity and potentially less model dependent. — Calibration, coverage, cut-off, missing locations and organisational boundary. |
| 2. Primary activity data | Supplier quantities; kilometres travelled; tonnes purchased; product use; site-level exposure. | Specific to the entity or value-chain activity. — Supplier method, period mismatch, completeness, estimation within the supplier data and verification status. |
| 3. Entity-specific proxy | Spend or production proxy derived from a comparable internal operation or supplier segment. | Can preserve business-model specificity when direct data are incomplete. — Representativeness, scaling basis and sensitivity. |
| 4. Secondary factor or industry average | Published emission factor, industry intensity, public hazard or economic dataset. | Available and comparable; often necessary for broad value-chain coverage. — Jurisdiction, technology, age, sector match, uncertainty and source authority. |
| 5. Modelled or extrapolated value | Sample-to-population extrapolation; scenario model; econometric or geospatial model. | Can provide a complete estimate and range where measurement is impossible. — Model risk, parameter selection, validation, excluded variables and sensitivity. |
4. Scope 3: value-chain gaps do not remove the need to estimate
IFRS S2 expects Scope 3 measurement to include estimation rather than direct measurement alone. The entity uses a measurement approach, inputs and assumptions that faithfully represent the measure, and applies all reasonable and supportable information available without undue cost or effort. Secondary data, industry averages and proxy data are therefore normal parts of the process when primary data are incomplete.
The Standard contains a presumption that Scope 3 emissions can be estimated reliably using secondary data and industry averages. Only in rare cases, after every reasonable effort, may estimation be impracticable. In that case the entity discloses how it manages its Scope 3 emissions. This is a much higher threshold than “supplier data are missing”.
In practice
| Scope 3 gap | Usable estimate approach | Control and disclosure |
|---|---|---|
| Some suppliers have no emissions data | Use purchase quantities or spend with representative factors; prioritise high-emitting suppliers for primary data. | Disclose primary-versus-secondary data extent, factor source, representativeness and supplier-data roadmap. |
| Supplier reporting period differs | Use the most recent data available where the IFRS S2 conditions are met. | Confirm equal period length and disclose significant events or changes between reporting dates. |
| Product-use assumptions vary | Use a documented expected-use profile, scenario or range based on product and market evidence. | Sensitivity to useful life, energy mix, utilisation and geography; avoid unsupported single-value precision. |
| Category population is very large | Use statistically or operationally supportable sampling and extrapolation. | Sample design, coverage, exclusions, scaling method and reasonableness test. |
| Financial institution lacks counterparty data | Use asset-class, sector or jurisdiction proxies consistent with the selected methodology. | Coverage, proxy hierarchy, attribution method, data quality score and improvement priorities. |
5. Model risk in sustainability reporting
A model is not only sophisticated software. Any structured method that transforms inputs into a reported amount can create model risk. This includes emission calculations, extrapolation spreadsheets, scenario models, geospatial overlays, financed-emissions attribution, target pathways and financial-effect sensitivities. A reporting team should apply controls proportionate to the model’s materiality and complexity.
In practice
| Model-risk source | Failure mode | Control |
|---|---|---|
| Purpose and scope | Model answers a different question from the disclosure requirement or uses the wrong boundary. | Document intended use, disclosure objective, population, period and exclusions. |
| Input data | Incomplete, stale, duplicated or unrepresentative inputs drive the result. | Data-quality tests, source hierarchy, cut-off controls and coverage reconciliation. |
| Method structure | Formula, attribution or scenario logic omits important relationships. | Technical review, independent re-performance and model limitation note. |
| Parameters and factors | Emission factors, hazard thresholds or economic assumptions are outdated or mismatched. | Approved parameter library, source date, jurisdiction/technology match and version control. |
| Sensitivity and uncertainty | Point estimate hides a wide range of reasonably possible outcomes. | Sensitivity analysis, ranges, alternative scenarios and uncertainty disclosure. |
| Change management | Method changes are made without assessing comparatives or explaining the effect. | Model change approval, impact assessment, comparative treatment and audit trail. |
6. Estimate register: the core control
An estimate register connects the public amount to its source, method, uncertainty and improvement plan. It supports drafting, technical review, assurance readiness and comparative consistency. The register should focus on material or high-uncertainty estimates, not become an unmanageable inventory of every formula in the organisation.
Figure 2. An estimate register controls the reported metric, source hierarchy, model, uncertainty, limitation and remediation.
In practice
| Register field | Minimum content |
|---|---|
| Estimate ID and disclosure link | Stable identifier, Standard paragraph, metric/disclosure, draft location and materiality rationale. |
| Boundary and unit | Reporting entity, value-chain segment, geography, period, unit and aggregation level. |
| Source data | System or external source, owner, extraction date, coverage, period alignment and verification status. |
| Proxy and hierarchy rationale | Why the selected source is sufficiently specific, timely and representative; alternatives considered. |
| Method and model version | Calculation, factor, model, scenario, code/spreadsheet version and change history. |
| Assumptions and approximations | Key assumptions, ranges, extrapolation, judgement, sensitivity and interaction with outcome uncertainty. |
| Measurement uncertainty | High/medium/low internal assessment, principal sources, range or sensitivity and effect on usefulness. |
| Controls and approvals | Preparer, reviewer, validation, reconciliation, governance approval and assurance status. |
| Published limitation wording | Specific data gaps, exclusions, use of secondary data, period mismatch and method constraints. |
| Improvement plan | Action, owner, target date, expected quality improvement and trigger for method reassessment. |
In practice
7. Decision criteria: use an estimate, a range, qualitative information or no amount?
| Decision | Use when | Required caution |
|---|---|---|
| Point estimate | The method and inputs support a central amount that is decision-useful and not misleadingly precise. | Explain material assumptions and significant measurement uncertainty; consider rounding. |
| Range | Reasonably possible outcomes or measurement alternatives are too wide for a single amount to communicate faithfully, or the relevant IFRS requirement permits a range. | Explain how endpoints were selected, whether probabilities differ and what could move the result within or beyond the range. |
| Scenario or sensitivity set | Users need to understand how the amount changes under different assumptions or pathways. | Do not imply probabilities unless assessed; connect to strategy, resilience and financial planning. |
| Qualitative information | The specific financial-effects criteria permit omission of quantitative information, or narrative is needed to explain an amount. | Apply the exact criteria; identify likely affected financial-statement line items and combined quantitative effects where required. |
| No amount because impracticable | Only when the relevant requirement contains an impracticability exception and the entity cannot apply it after every reasonable effort. | Disclose the fact and required information; retain evidence of every reasonable effort. |
In practice
9. Illustrative disclosure wording
| Feature | Why it works | Evidence required |
|---|---|---|
| Method and coverage | Users can see how much is supplier specific and how the remainder was estimated. | Supplier dataset, coverage reconciliation and factor mapping. |
| Specific uncertainty source | The limitation is linked to method consistency and factor representativeness, not a generic “data challenge”. | Data-quality assessment and factor-source review. |
| Sensitivity | Shows how an important assumption affects the amount. | Re-performable sensitivity analysis. |
| Improvement plan | Makes the limitation time-bound and owned. | Supplier programme, milestones and responsible owner. |
| Comparative treatment | Explains why prior information was revised and links to the relevant note. | Comparative-change assessment and approval. |
In practice
10. Weak versus stronger uncertainty disclosure
| Weak wording | Why it is weak | Stronger pattern |
|---|---|---|
| “Some data were estimated.” | Does not identify amounts, method, source, assumptions or significance. | Identify the material estimated amount, source hierarchy, method, key assumptions and why the estimate remains useful. |
| “Scope 3 data quality is poor.” | Provides no category, coverage, proxy, factor or improvement information. | Explain which categories or populations use primary and secondary data, representative limitations, sensitivity and improvement action. |
| “Results may change due to uncertainty.” | Does not distinguish future outcome uncertainty from measurement uncertainty. | Describe the uncertain future event and separately explain data, model and assumption uncertainty affecting the reported amount. |
| “The model is industry standard.” | Authority label does not demonstrate fitness for the entity’s disclosure. | Explain model purpose, boundary, version, inputs, key limitations, validation and how it supports the disclosure objective. |
In practice
11. Common mistakes
| MISTAKE 1 | Equating a data gap with permission to omit the metric. |
|---|---|
| Why it happens | Teams assume direct measurement is required. |
| Why it matters | Material information is omitted even though a reasonable estimate or proxy could provide useful information. |
| Correction | Apply the relevant data hierarchy and test faithful estimation before considering any specific impracticability provision. |
| Evidence of correction | Estimate decision record, alternatives considered and documented efforts. |
In practice
| MISTAKE 2 | Reporting a precise number from a highly uncertain model. |
|---|---|
| Why it happens | A point estimate appears easier to publish and reconcile. |
| Why it matters | False precision obscures sensitivity and can mislead users about confidence. |
| Correction | Use appropriate rounding, range or sensitivity and explain the principal uncertainty sources. |
| Evidence of correction | Sensitivity results and approved presentation rationale. |
In practice
| MISTAKE 3 | Listing assumptions without identifying which materially affect the result. |
|---|---|
| Why it happens | Methodology notes become comprehensive but not decision-useful. |
| Why it matters | Users cannot understand the most significant uncertainty or how the number might change. |
| Correction | Prioritise difficult, subjective or complex assumptions and quantify sensitivity where useful. |
| Evidence of correction | Material-assumption assessment linked to the estimate register. |
In practice
| MISTAKE 4 | Changing factors or proxies without comparative analysis. |
|---|---|
| Why it happens | Updates are treated as routine data maintenance. |
| Why it matters | Trends reflect methodology change rather than performance, and prior-period treatment may be wrong. |
| Correction | Classify the change, assess whether prior circumstances are evidenced, and apply IFRS S1 comparative guidance. |
| Evidence of correction | Method-change log, comparative decision and disclosure. |
In practice
| MISTAKE 5 | Using supplier data without understanding the supplier method. |
|---|---|
| Why it happens | Primary data are assumed to be inherently superior. |
| Why it matters | Inconsistent boundary, period or methodology can reduce comparability and faithful representation. |
| Correction | Assess specificity, timeliness, jurisdiction, technology, verification and method consistency—not only whether data came from the supplier. |
| Evidence of correction | Supplier-data quality score and review record. |
In practice
12. Myth versus reality
| MYTH | IFRS S1 and IFRS S2 require audited, directly measured data for every sustainabi |
|---|---|
| REALITY | The Standards recognise estimation, proxies, secondary data and modelled information. Reasonable estimates can be essential and useful when accurately described and explained. The entity must apply the relevant measurement requirements, prioritise appropriate inputs, disclose significant uncertainty and meet any local assurance obligation separately. |
| Why the confusion arises | Financial-reporting discipline is sometimes simplified into a belief that only directly measured or externally assured data are reportable. |
| Practical consequence | Reporting teams should invest in methodology, data lineage, model controls and transparent limitation disclosure rather than delaying all reporting until perfect data exist. |
In practice
15. Related standards and learning path
| Link role | Instrument or article | Use |
|---|---|---|
| Primary | IFRS S1 paragraphs 74–82 | Significant judgements and measurement uncertainty disclosures. |
| Scope 3 | IFRS S2 B38–B57 | Estimation, data hierarchy, method disclosure and rare impracticability. |
| Proportionality | IFRS S1 and S2 Proportionality: 'Undue Cost or Effort' | Separate the information-search boundary from estimation and uncertainty. |
| Comparatives | IFRS S1 and S2 Comparatives, Restatements and Errors | Classify factor, model and estimate changes and correct prior periods. |
| GHG implementation | IFRS S2 Scope 1, Scope 2 and Scope 3 Emissions: Complete Measurement Guide | Apply detailed GHG boundary, factor and quality controls. |
Outcome uncertainty concerns what may happen, whereas measurement uncertainty concerns how uncertain a reported amount is. A disclosure can contain both, but the distinction affects the assumptions, evidence, model and explanation needed and should not be presented as two formal categories with identical requirements.
Questions
Questions people ask
Are estimates acceptable under IFRS S1 and S2?
IFRS S1 and IFRS S2 do not require perfect data before useful sustainability-related financial information can be reported. Reasonable estimates, proxies, modelled data and ranges can be necessary and appropriate when direct measurement is unavailable, provided the resulting information is faithfully represented, material methods and assumptions are explained, and significant measurement uncertainty is transparent.
What is measurement uncertainty?
The issued Standards use “measurement uncertainty” for uncertainty affecting amounts reported. Official ISSB effects-analysis and educational materials also help explain a separate practical idea: uncertainty about the outcome of a possible future event.
Is outcome uncertainty the same thing?
Outcome uncertainty concerns what may happen, whereas measurement uncertainty concerns how uncertain a reported amount is. A disclosure can contain both, but the distinction affects the assumptions, evidence, model and explanation needed and should not be presented as two formal categories with identical requirements.
Can Scope 3 use secondary data?
Many sustainability-related metrics cannot be measured directly. Scope 3 greenhouse gas emissions rely on supplier information, activity data, emission factors and secondary data.
When can Scope 3 be omitted as impracticable?
The Standard contains a presumption that Scope 3 emissions can be estimated reliably using secondary data and industry averages. Only in rare cases, after every reasonable effort, may estimation be impracticable. In that case the entity discloses how it manages its Scope 3 emissions.
What belongs in an estimate register?
An estimate register connects the public amount to its source, method, uncertainty and improvement plan. It supports drafting, technical review, assurance readiness and comparative consistency.
Sources
Primary sources
Take it with you
The checklists as a working spreadsheet
Every checklist and table on this page, with empty status, owner and evidence columns for your team to fill in and keep.
✓ LRA AI Assistant · Human-in-the-loop
Ask about this guide
It answers from this page, and reaches into the linked disclosure cards when your question is about the standard itself. Your first two answers are free without signing in.
Go deeper · IFRS S1 / S2
IFRS S1 and S2 training
Financial materiality, scenario analysis and the S2 climate disclosures, applied to your own reporting.
Available as Guided Flex, Live Cohort, 1:1 Expert Mentorship or Corporate Programme.
