GRI Impact Inventory: How to Build a Complete Long List of Impacts

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

An impact inventory is the controlled long list of the organisation’s actual and potential impacts on the economy, environment and people, including human rights, across its activities and business relationships. It is built after understanding the organisation’s context and before assessing significance.

GRI 3 does not require a file with the title “impact inventory”. It requires the organisation to understand its context, identify its actual and potential impacts, assess their significance and prioritise the most significant impacts for reporting. A structured inventory is the practical evidence trail that shows these stages were not collapsed into one management vote or a generic ESG topic list.

GRI 3 asks the organisation to consider activities, products and services, sectors, locations, employees and other workers, business relationships, sustainability context and stakeholders. A six-lens map is a practical way to stop the exercise from becoming a head-office policy review.

Figure 1. The impact universe should cover activities, products, sites, workers, communities and business relationships.

Lock the organisational context. Create a current map of entities, sectors, activities, products, locations, workers, markets and business relationships, including minority interests and relationships beyond Tier 1 where relevant.

Map potentially affected stakeholders. Identify people and groups whose interests or rights are or could be affected, including those without a direct relationship, vulnerable groups and those unable to articulate their views.

Harvest internal evidence. Extract impact leads from due diligence, incidents, grievances, audits, inspections, compliance systems, HR, product, procurement, environmental, legal and project records.

Add external and stakeholder evidence. Review regulators, courts, civil society, media, experts, worker representatives, communities, sector sources and location datasets.

Write impact statements, not topics. Describe the affected object, the change, the activity/product/relationship causing the pathway, and the relevant location or value-chain stage.

Classify the impact. Record actual/potential, positive/negative, intended/unintended, short/long term, reversible/irreversible, and cause/contribute/directly linked where relevant.

Separate distinct impacts. Do not offset positive against negative or merge different affected groups merely because they fit the same topic label.

Assign evidence and ownership. Link every statement to sources, date, confidence, fact owner, subject-matter reviewer, open questions and update trigger.

Challenge completeness and duplicates. Review the list by site, product, affected group, geography and value-chain stage; consolidate true duplicates without losing distinctions.

Freeze the inventory for scoring. Approve a versioned long list before Step 3 significance assessment begins; later additions require a documented change.

Figure 2. Impact-inventory workflow: map context, harvest evidence, write and classify impacts, challenge completeness, then freeze the list for scoring.

GRI 3 explicitly points to diverse evidence: impact assessments, legal reviews, anti-corruption systems, financial audits, occupational health and safety inspections, shareholder filings, business-relationship assessments, grievance mechanisms, outward-impact information in enterprise risk management, news, civil society, stakeholders, experts and Sector Standards. No single source is complete.

External tools can reveal context that internal systems miss. They are most useful when linked to specific sites, commodities, sectors or affected groups and then tested against local evidence. Record the dataset, version or extraction date, geographic resolution, limitations and the decision it informed.

GRI 3 explains that negative impacts cannot be offset by positive impacts. A renewable-energy project can reduce emissions and expand access to electricity while also creating land, biodiversity or human-rights impacts. Record these as separate inventory rows, with separate evidence and significance assessments.

For negative impacts, the cause/contribute/directly linked distinction helps the organisation understand the pathway and later management response. “Directly linked” is not limited to the first contractual tier. A product can be directly linked to an impact through several business relationships even when the organisation did not cause or contribute to it.

The following register is an LRA implementation template. It is designed to preserve evidence and judgement before scoring; it is not a mandatory GRI form.

Taxonomy should help retrieve and compare impacts, not replace them with generic topics. Use controlled tags for affected object, right or environmental component, activity, location, value-chain stage, stakeholder group, actual/potential status and involvement. Keep the full statement as the primary record.

One impact, multiple tags. A supplier overtime impact may be tagged labour rights, health and safety, procurement practice, Tier 1 manufacturing and a specific country.

One topic, multiple impacts. “Water” may contain community access, ecosystem degradation, worker sanitation and product-use impacts. Do not assume they have the same severity or owner.

Cluster after identification. Proposed topic labels can be added for navigation, but the frozen scoring population should retain each distinct impact ID.

Use supersession, not deletion. When two entries are true duplicates, preserve the old IDs and record the surviving master record and rationale.

The sustainability team should own the process, not every fact. Site managers, HR, procurement, product, legal, compliance, finance and community-relations teams should own source evidence. Subject-matter specialists challenge technical conclusions, and a cross-functional reviewer approves the frozen list.

The quality gate asks whether the list is ready to be assessed - not whether every impact will become material. A complete long list can contain uncertain, contested or low-significance impacts, provided their status is transparent.

The team maps sites, products, commodities, workers and business relationships. It harvests water permits, injury and near-miss data, seasonal-worker complaints, supplier audits, deforestation alerts, customer nutrition feedback, packaging life-cycle studies, community interviews and labour/climate context datasets.

Instead of retaining four generic topics, it writes separate impact statements: dry-season water pressure on a local ecosystem; potential excessive working hours for seasonal labour; child-labour and deforestation impacts directly linked through cocoa tiers; positive livelihood effects from stable farmer purchasing; potential worker and community impacts from the plant closure; consumer health effects linked to product composition; and packaging waste across customer markets.

Each statement has its own location, affected group, involvement, evidence and owner. Only then does the organisation assess severity, likelihood, scale and scope and cluster the prioritised impacts into material topics. The final topic list is shorter than the inventory, but every topic can be traced back to concrete impacts.

Once the inventory is frozen, design the significance assessment so it preserves the distinctions already captured. Negative actual impacts are assessed by severity; potential negative impacts by severity and likelihood, with severity taking precedence over likelihood for potential human-rights impacts. Positive impacts use scale and scope, and likelihood for potential impacts. The quality of that assessment depends on the completeness and specificity of the long list built here.

An ESG topic list can be used as one prompt, but not as the impact inventory or the materiality result. The organisation should map its context and affected stakeholders, harvest internal, external and stakeholder evidence, write concrete actual and potential impact statements, check completeness and duplicates, and only then assess significance and cluster prioritised impacts into material topics.