RSE and CSRD reporting in food service: how CalcMenu auto-generates GRI 305 data
What CSRD concretely requires in collective catering, why the recipe database is already the sustainability data source, and what CalcMenu's GRI 305 export contains.
The Corporate Sustainability Reporting Directive (CSRD) is the EU’s new framework for non-financial reporting. It replaces the NFRD (Non-Financial Reporting Directive) with a significantly broader scope: where the NFRD covered around 11,000 companies in Europe, the CSRD was originally expected to cover more than 50,000 — before the Omnibus I directive (Directive (EU) 2026/470, published in February 2026) narrowed its scope to companies with more than 1,000 employees and over EUR 450 million in net turnover, removing the vast majority of companies originally in scope. Large listed companies published their first CSRD reports covering the 2024 financial year. Following the April 2025 “stop-the-clock” directive and the February 2026 Omnibus revision, listed SMEs were removed from mandatory scope entirely; the remaining in-scope companies will report from financial year 2027, publishing their first reports in 2028.
For collective catering operators and hotel site managers belonging to groups subject to the CSRD — large healthcare establishments, chain hotels, concession catering companies — the question is no longer whether CSR reporting is mandatory. It is where the data will come from.
What CSRD requires for food emissions
The CSRD draws on the European Sustainability Reporting Standards (ESRS), notably ESRS E1 for climate change. The relevant indicator for food service is Scope 3, Category 1: purchased goods and services — in this case, the ingredients bought to produce meals.
The GRI 305: Emissions 2016 standard, whose greenhouse-gas disclosures (305-1 to 305-5) are mapped to ESRS E1 in the official GRI–ESRS Interoperability Index published by GRI and EFRAG, requires disclosure of emissions in CO₂ equivalent, with the data source, the calculation method used and the scope covered. For food service, this means concretely: for each ingredient purchased, an emission factor in kg CO₂e per kg of ingredient, multiplied by the quantities consumed over the reporting period.
Collecting this data manually for a kitchen producing 1,000 meals per day with 200 to 500 ingredient references is not feasible without automation.
Why the recipe database is already the data source
The good news for operators using CalcMenu is that the data needed for the GRI 305 report is already in their system.
Every ingredient in the CalcMenu database is linked to its Agribalyse v3.1 equivalent — the ADEME database that compiles life-cycle assessments (LCA) for thousands of food products. The Agribalyse emission factor is already attached to each ingredient. When a recipe is used in a menu, CalcMenu records the quantities consumed. The quantity × emission factor multiplication is performed automatically for each ingredient, each recipe, each service.
There is no additional data entry required to produce a CSR report. The data is generated continuously through normal use of the software.
What the GRI 305 export contains
CalcMenu’s CSR export produces a structured file covering the desired period (quarter, half-year, year) and including the following elements.
Total Scope 3 emissions in kg CO₂e, broken down by ingredient category (meats, dairy, cereals, vegetables, fish, etc.) and by site if you operate across multiple units.
Carbon intensity expressed in kg CO₂e per meal served — a performance indicator that is comparable across periods and between sites.
Percentage of LCA data available — that is, the proportion of ingredients for which an Agribalyse factor is available as opposed to estimated ingredients. This metric supports GRI 305-3’s requirement to disclose the standards, methodologies, assumptions and sources of emission factors used, helping qualify data reliability.
Additional CSR indicators: share of local sourcing (based on configured distances), share of certified organic products, percentage of ingredients carrying sustainability labels (MSC, Fairtrade, IP-SUISSE, etc.).
This file can be used directly by a CSR auditor or reporting firm. It does not need to be reconstructed from scattered records.
Who this affects in food service
The CSRD targets large companies first. But the reporting chain extends further: a catering company supplying a group subject to the CSRD may be asked to provide emissions data to its institutional client. Concession caterers in airports, hospitals, universities or large listed corporations are already in this situation.
In Switzerland, analogous obligations arise from Articles 964a–c of the Code of Obligations (non-financial reporting by large public-interest companies, first applied to the 2023 financial year), from the Ordinance on Climate Disclosures in force since 1 January 2024, and from binding FINMA disclosure requirements for the largest supervised banks and insurers. Cantonal hospitals and large care-home groups incorporating CSR criteria into their public procurement contracts are also affected.
And in the United States? The picture is far less settled. The SEC adopted a federal climate disclosure rule in 2024, but it was never enforced, and in 2026 the SEC proposed rescinding it. California is pushing ahead instead: SB 253 requires companies with more than USD 1 billion in revenue doing business in California to report their Scope 1 and 2 emissions — first reports deferred to November 10, 2026, with Scope 3 following from 2027 — while the climate-risk reports under the companion law SB 261 are currently blocked in court. The net effect: US sustainability reporting remains voluntary except for the largest companies touching California, and the rules are still moving. Large hospitality groups with California operations would nonetheless do well to start preparing their emissions data pipelines now.
The common thread: in all these cases, the source of food emissions data is the kitchen. And the kitchen runs on CalcMenu.
To see what a GRI 305 export generated from your own recipe database looks like, request a demonstration.
Comments
Comments coming soon.