GRI 102 Scope 3 Reporting: Categories, Data Hierarchy and Supplier Estimates

Technical status: source-grounded publication draft; final human technical review required before release.

Disclosure 102-7 moves Scope 3 from a broad total into a category-by-category reporting discipline. The organisation reports gross Scope 3 emissions in metric tonnes of CO2 equivalent, includes the seven greenhouse gases covered by the Kyoto Protocol, provides a breakdown for each of the 15 upstream and downstream categories, and explains its consolidation approach, methodologies, assumptions, calculation tools and emission-factor sources. The disclosure also contains base-year and recalculation information. (GRI 102, Disclosure 102-7.)

The practical consequence is important: the reporting team needs a controlled register for all categories, not merely a single workbook total. Screening remains essential, but its purpose is to identify the calculation depth, data-acquisition priority and quality-improvement plan for each category. It is not a licence to make inconvenient categories disappear from the disclosure.

A defensible Scope 3 project has four linked outputs: a complete category register, a calculation file, a data-quality and controls record, and a disclosure pack. The register is the bridge between GRI 102 and the GHG Protocol. It shows what the category covers for this organisation, what was included or excluded, which method was used, who owns the data, how the result was reviewed and what will improve next year.

Define the organisational and value-chain boundary. Record the consolidation approach selected for Scope 1, Scope 2 and Scope 3 - equity share, financial control or operational control - and apply it consistently. Then map the activities that sit upstream and downstream of the reporting organisation.

Screen every category. Use rough estimates and relevance criteria to identify likely magnitude, influence, transition exposure, stakeholder interest, outsourced activities and sector-specific significance.

Choose the calculation route. Select supplier-specific, hybrid, activity-based average, spend-based or other defensible methods according to the category and data available.

Rate data quality and uncertainty. Assess technological, temporal and geographical representativeness, completeness and reliability. Record boundary and allocation weaknesses separately from numerical uncertainty.

Calculate and challenge. Reconcile activity data to procurement, logistics, HR, asset and financial systems; check units and emission-factor years; investigate movements and outliers.

Disclose the category result and limitations. Explain methods and factor sources, show category-level totals, and use a reason for omission only where information cannot be reported as required.

Approve an improvement plan. Prioritise categories that combine high emissions, low data quality and strong ability to influence the value chain.

Figure 1. Scope 3 category map: screen all 15 categories, then tailor calculation depth and improvement activity.

The matrix below is an implementation aid, not a replacement for the minimum boundaries and detailed calculation guidance in the GHG Protocol. “Typical data” identifies practical starting points. The final method must reflect the organisation’s facts, category boundary and intended use of the information.

The GHG Protocol recommends screening Scope 3 activities using rough estimates and considering more than size. Relevant criteria include the expected magnitude of emissions, the organisation’s influence, climate-related risks and opportunities, stakeholder concerns, activities previously performed in-house but now outsourced, sector guidance and other organisation-specific criteria. A category can be strategically important even when its first rough estimate is not the largest.

For GRI 102, keep three decisions separate. First, is climate change a material topic under the GRI material-topic process? Second, what falls within each Scope 3 category and how will it be calculated? Third, where should the organisation invest in better data and action? A category screening score should not be presented as if it were the organisation’s GRI materiality determination.

A data hierarchy should guide judgement, not replace it. IFRS S2 and GHG Protocol concepts both point towards more direct, activity-specific, timely, geographically and technologically representative data where practicable. GRI 102 guidance recommends reporting the percentage of emissions obtained through primary data for each category. Yet a supplier number is not automatically high quality: it can use the wrong period, omit upstream stages, apply an inconsistent consolidation approach, allocate emissions arbitrarily or lack review.

Figure 2. A practical Scope 3 data hierarchy and first-year roadmap.

Do not begin with a generic “send us your carbon footprint” request. Ask for data that correspond to the category, product and reporting boundary. A useful supplier request identifies the product or service purchased, reporting period, quantity or allocation base, organisational and product boundary, gases and GWP values, Scope 1/2/3 components included, emission-factor sources, verification status and known exclusions.

Secondary data are not a reporting failure. They are a normal part of Scope 3 measurement, especially in a first year. The quality question is whether the proxy is appropriate, consistently applied and transparently disclosed. A physical proxy - tonnes of steel, tonne-kilometres, kilowatt-hours, litres of fuel, waste mass or product use - usually preserves more operational meaning than spend. Spend-based factors can nevertheless be efficient for screening and for service categories where physical activity is genuinely unavailable.

Maintain an emission-factor register with the factor name, publisher, database version, geography, technology, unit, GWP basis, applicable category, date downloaded and owner. Lock the factor set for the reporting cycle. Changes after calculation should be controlled, documented and, where material, trigger a recalculation or restatement assessment.

GRI 1 permits the reason for omission “information unavailable/incomplete” for disclosures and requirements where reasons for omission are allowed. The organisation must specify the information that is missing, explain why it is unavailable or incomplete, and describe the steps being taken and the expected timeframe for obtaining it. Under GRI 102 guidance, this route is available where a Scope 3 category cannot be reported as required. It should be a controlled exception, not the default calculation method.

Before using an omission, test whether a reasonable estimate is possible. If a category can be estimated using activity data, a representative factor, sampling or a sector proxy, a transparent estimate will often provide more useful information than a blank cell. The decision should be documented and approved, especially when the missing category could be significant.

The company does not wait for supplier footprints. It uses mass-based product factors for steel, aluminium and packaging; a spend-based model for residual purchased services; distance and mode data for the known logistics population plus a tested extrapolation for the remainder; and an engineering lifetime-use model for sold equipment, segmented by product type and sales region. It records the percentage of each category based on primary activity data, documents the factors and assumptions, and launches supplier requests for the ten highest-emission material groups.

The result is not “perfect”, but it is complete enough to reveal where action and data improvement matter. The company’s Year 2 plan is to replace spend estimates for priority materials, improve logistics coverage from 60% to 90%, collect verified use-phase assumptions from product engineering and add automated reconciliations to procurement and sales systems.

Treat the first inventory as the beginning of a reporting system. Lock the category register, assign data owners, connect improvement actions to supplier engagement and product decisions, and schedule a post-publication review while the evidence is still fresh. The strongest Year 2 improvement plan is based on observed data-quality weaknesses, not a generic promise to “collect more primary data”.