Methodologies

Selection of topics

The topics covered in this Report were selected on the basis of our materiality analysis, which assessed the relevance and impact of selected topics for both our company and various stakeholders. The sustainability data is qualitative and quantitative; the qualitative information can also contain quantitative elements. The materiality matrix and the process by which it is created is described in General information.

Scope

The sustainability data in this Report covers all entities that belong to the scope of the Consolidated Financial Statements. In the value chain, sustainability data relates primarily to Tier 1 suppliers. Unless otherwise explicitly stated, the sustainability data in this Report have not been verified externally.

Divestments

Environmental data is reported until the moment control of the divested entity is transferred, and social data until the end of the month in which control of the divested entity is transferred. The date on which control of the company is transferred generally coincides with the date a divestment is closed, and control of the shares is transferred to the new owner.

Methodology descriptions

In the following pages, we describe how we have calculated the metrics included throughout the Sustainability Statements. This includes mandatory disclosures in line with ESRS, voluntary (partial) disclosures that are consistent with ESRS and entity-specific metrics that have been prepared following the Transitional provision related to entity-specific disclosures in ESRS 1.

The methodology descriptions include information under the following sections:

• Disclosure requirements

• Definitions

• Calculation methodologies

• Judgements and uncertainties

Definitions

• Energy consumption: Total energy used on site, including electricity, fuels, and heat, disaggregated by source and type

• Scope 1 emissions: Direct GHG emissions from owned or controlled operations, including combustion of fuels and process emissions

• Scope 2 emissions: Indirect GHG emissions from purchased electricity and heat, reported using both market-based and location-based methods

• Water withdrawal: Total water drawn into the site from all sources

• Water discharge: Total water leaving the site boundary

• Water consumption: Water withdrawal minus water discharge

• Resource outflows: All waste streams leaving the site, categorized by type and treatment

• Renewable electricity: We consider renewable electricity, as electricity from renewable sources such as wind, solar, geothermal, sustainably sourced biomass (including biogas and biomethane) and sustainable hydropower. Renewable electricity is sourced using various instruments such as self-generation, direct procurement from renewable generators, retail contracts and unbundled Energy Attribute Certificates (EACs). In certain markets a double counting risk exists. To prevent this risk, instruments such as contracts are used to ensure credibility and verifiability.

Calculation methodologies

Targets related to climate change mitigation and adaptation: Scope 1 & 2, and renewable electricity: Near-term and net-zero emissions reductions targets have been set by dsm-firmenich and validated by the Science Based Targets Initiative (SBTi). The base year is calendar year 2021; January 1, 2021, until December 31, 2021. All Scope 1 & 2 (market-based) emissions are in scope. An additional target on the purchasing of renewable electricity was also set.

Energy: Energy consumption is reported as the total energy used on site, including electricity, fuels, and heat, whether purchased or self-generated, and is disaggregated by renewable and non-renewable sources. Data is collected from invoices, meter readings, and validated calculations, with conversion to standard units (low heating value for fuels) and application of appropriate emission factors. The percentage of renewable energy is determined by the share of renewable electricity, heat, and fuels in total consumption, following IEA and RE100 guidelines.

Scope 1 & 2 emissions: Scope 1 emissions are calculated based on fuel consumption and process activity data collected from site-level meters, invoices and operational logs. Emissions factors are applied to calculate the corresponding emissions.

Scope 2 emissions are calculated based on the amount of purchased electricity, steam, heat or cooling collected from invoices, meter readings or site records. Emissions are calculated for both location-based and market-based approaches. Location-based uses the average grid emission factor for the country or region, market-based uses supplier-specific emission factors. Renewable electricity (on the basis of credible certificates or contracts), the supplier-specific emission factor is zero or near zero.

All calculations follow the GHG Protocol, with biogenic CO₂ reported separately.

Targets related to water and marine resources: water efficiency improvement in water-stressed areas: The water efficiency target encompasses all dsm-firmenich sites located in a water-stressed area, defined as sites scoring “high” or “very high” for water stress for baseline or 2030 business as usual scenario, according to WRI Aqueduct database. It accounts for the quantity of water withdrawal from freshwater (excluding once-through-cooling), groundwater or third-party sources per ton of product. Efficiency improvement is calculated by comparing real water withdrawal of the reported year to the expected water withdrawal if water intensity would have been the same as in baseline year 2023.

Water consumption: Water consumption is defined as the difference between total water withdrawal and total water discharge, representing water not returned to the environment or third parties. Data is obtained from metering, validated calculations, and process engineering estimates, with all sources and assumptions documented in site SOPs. For material sites, water balances are maintained and meters are calibrated according to a defined plan.

Voluntary indicators: E2-4 Pollution of air, water and soil: Emissions to air (e.g., VOCs, SO₂, NOₓ), and water (e.g., COD, N, P) are reported for all material sources, using a combination of direct measurements, mass balances, and calculations based on production and process data. Methodologies are aligned with local regulatory requirements and international standards.

On Phosphorous (P) and Nitrogen (N) emissions, the P+N efficiency ambition encompasses all sites located in areas with high water pollution index, defined as sites scoring 4 or more in Science Based Targets for Nature (SBTN) tool “State of Nature Water Layers”. It accounts for the net quantity of P+N released to fresh or sea water per ton of product. Efficiency improvement is calculated by comparing real net P+N of the reported year to the expected net P+N using the intensity of baseline year 2023.

E5-5 Resource outflows: Waste generation and disposal are reported by category and by treatment method. Data is sourced from site weighing systems, waste registers, and contractor invoices, with estimates used only where direct measurement is not feasible.

Judgements and assumptions

• Materiality: Reporting focuses on material contributors, as determined by annual assessments

• Data quality: Preference is given to direct measurement; estimates are used only when necessary and are documented

• Emission factors: Updated annually and aligned with international standards

• Reporting boundaries: All sites under financial control are in scope for reporting. Sites may be excluded for material reasons

• Uncertainty: Where estimates or proxies are used, uncertainty is documented and efforts are made to improve data quality over time

Definitions

• Scope 3 emissions: Indirect emissions (CO₂, CH₄, N₂O, HFCs, PFCs, SF₆, NF₃) that occur in both upstream and downstream value chain, outside our direct control

• Emission factor: The amount of GHG emitted per unit of material or activity (e.g., kg CO₂e/kg, kg CO₂e/€). Emission factors are reviewed and updated annually

• Environmentally Extended Input-Output (EEIO): A method using economic data and sectoral emission factors to estimate emissions for indirect spend

Calculation methodologies

Scope 3 target: Our near-term and net-zero reduction targets have been SBT-validated. The base year is calendar year 2021; January 1 until December 31. The target covers 4 categories:

• Category 1: Purchased goods and services 90%

• Category 3: Fuel and energy related activities 100%

• Category 4: Upstream transportation and distribution 100%

• Category 5: Waste generated in operations 100%

Category 1: Purchased goods and services direct spend (production-related): Emissions are calculated via supplier-specific cradle-to-gate emission factors if available. Where not available, industry-average emission factors from LCA databases (ecoinvent, WFLDB, Agri-footprint) are used. For remaining spend, we extrapolate using weighted averages.

Indirect spend (non-production-related): Calculated using EEIO emission factors (USEEIO, DEFRA) applied to spend data for facility management, digital and technology, and technical services. Exclusions: capital goods, water/waste management, mobility/hotels.

Category 2: Capital goods: Emissions from capital goods (mechanical equipment, civil services, process controls, investments) are calculated using spend-based methods with industry-average EEIO emission factors.

Category 3: Fuel- and energy-related activities (not included in Scope 1 or 2): Emissions from extraction, production, and transport of fuels and purchased energy. Calculated using activity data (fuel/energy purchased) and region or country-specific emission factors. Includes transmission and distribution (T&D) losses.

Category 4: Upstream transportation and distribution: Emissions from transportation we purchase (road, marine, ground, air, warehouse storage). Shipment data is processed in Transvoyant using the distance-based method and remaining emissions are extrapolated based on shipment volumes or spend data.

Category 5: Waste generated in operations: Emissions from third-party treatment/disposal of waste generated at our sites, such as process and non-process hazardous or non-hazardous waste. Calculated using waste-type specific methods and emission factors for each treatment (e.g., landfill, incineration, recycling).

Category 6: Business travel: Emissions from air, rail, car (taxi, rental, employee-owned), and public transport. Major travel service providers supply primary emissions data (covering ~73–75% of travel), with extrapolation for full coverage. Emission factors are from ecoinvent; hotels emissions are disclosed under optional.

Category 7: Employee commuting: Calculated via HR data on employee numbers, assumptions on transport mode split (70% car, 30% public transport), working days, and average commute distance. Emission factors are from ecoinvent.

Category 8: Upstream leased assets: Emissions from leased cars (supplier data) and offices (from employee numbers, region, and average office space per employee). Emission factors are region-specific and based on ecoinvent.

Category 9: Downstream transportation and distribution: Emissions from storage of sold products in warehouses and distribution centers (not owned/controlled by dsm-firmenich). Calculated using sales volumes and emission factors for storage.

Category 10: Processing of sold products: Emissions from the first downstream processing step at the first off-taker of dsm-firmenich products. 100% of emissions from mixing are conservatively allocated to dsm-firmenich products, using sales volumes and energy mixing emission factors.

Category 11: Use of sold products: Emissions are based on the metabolized fraction via product-specific carbon content and molecular composition. Emissions are reported under optional being indirect use phase emissions Remaining emissions are in Category 12.

Category 12: End-of-life treatment of sold products: Emissions from waste disposal or treatment of all products and packaging we sell at end-of-life. Calculated via product volumes, carbon content, packaging data, treatment shares, and emission factors for each method.

Category 15: Investments: Emissions from investments not already covered in Categories 1–14. Entities are evaluated for relevance; emissions are calculated using net sales, dsm-firmenich sales share, and sector-specific emission factors (from EEIO databases).

Non-relevant categories: Categories 13 (Downstream leased assets) and 14 (franchises) as we do not have activities in these areas.

Judgements and assumptions

• Data quality: Supplier-specific emission factors are prioritized and industry averages used where necessary. Reviews are annual or quarterly for supplier-specific factors

• Extrapolation: Where data is incomplete, emissions are extrapolated based on purchased volume averages or spend.

• Emission factor hierarchy: Used emission factors are prioritized as follows: supplier-specific > weighted average > industry average > in-house LCA model > proxy > extrapolation. Uncertainty is highest where extrapolation or proxies are used

• Biogenic emissions: Reported separately if relevant, in line with GHG Protocol and ESRS

• Carbon credits/offsets: Only reductions within the value chain are counted; external offsets are not used for SBTi/ESRS compliance except for net-zero residuals

Definitions

• Substance of Concern (SoC): A chemical element or compound meeting the criteria in Article 57 of REACH and/or classified in Annex VI of CLP Regulation (EC No. 1272/2008) with relevant hazard statements

• Substance of very High Concern (SvHC): Substances identified on the ECHA Candidate List under Article 59(1) of REACH

• Main hazard class: Grouping of substances by hazard statements, as described in ESRS Annex II Table 2 on Terms defined in the ESRS – Substances of Concern.

Calculation methodologies

Total mass volumes of SoC and SvHC are determined by integrating product sales data with detailed composition information from our two main product data platforms. Substances present in products are identified and quantified using standardized chemical identifiers, i.e., CAS-number, and are matched against current regulatory lists to establish their classification as SoC or SvHC.

A materiality threshold is applied, whereby substances present at concentrations below 0.1% in any product are excluded from the analysis. Substances lacking standardized identifiers are also excluded.

For the purposes of reporting and given the nature of our business, it is assumed the mass volume of substances purchased or produced is equivalent to the mass volume present in products sold.

Judgements and assumptions

• Sales volume equivalence: It is assumed that the mass volume of substances purchased/produced equals the mass volume in products sold, given the nature of dsm-firmenich’s business

• Exclusion of samples: Samples are excluded from reporting. The impact of this exclusion has been judged to be low

• Cut-off threshold: A 0.1% cut-off is applied to focus on substances that materially contribute to SoC/SvHC volume. Quantitative assessment shows low impact on overall results.

• CAS number requirement: Only substances with CAS numbers are included. The impact of excluding substances without CAS numbers is assessed and documented internally.

• Use of estimates: No estimates are used in the reporting of SoC/SvHC

• Identification of substance: Substances of Concern are identified based on the hazard statements included in Annex VI. The source lists are updated throughout the year. For the analysis the following lists are used: SVHC candidate list downloaded October 31, 2025 from https://echa.europa.eu/candidate-list-table Hazardous substances classification (SoC list) ATP21 from https://echa.europa.eu/information-on-chemicals/annex-vi-to-clp

Definitions

• Biodegradation: the breakdown of organic substances into simpler components by microorganisms, which can be measured through various tests with corresponding OECD test protocols.

• Ultimately biodegradable: the complete biodegradation of a substance into its basic elements, such as carbon dioxide, water, and inorganic minerals, by microorganisms under specified test conditions. To be considered ultimately biodegradable, a substance typically needs to break down by at least 60% according to international testing guidelines such as OECD 301.

• Readily biodegradable: If 60% of degradation is reached within a 10-day window in a 28-day OECD 301 test period, the material is considered readily biodegradable.

• Washable products: applications where products are rinsed-off at the use phase or end-of-life, leading to these products entering the aquatic systems through drainage.

Calculation methodologies

The scope consists of the washable products in the Perfumery & Beauty (P&B) Business Unit, consisting of Fine Fragrance, Ingredients and Beauty & Care business segments, and the different application and subsegments of these businesses where the products could be rinsed off after usage at end-of-life.

The calculation and disclosure of the percentage of ultimately biodegradable products within the washable universe leverages the OECD 301/302/310 standards. Sales data for the in-scope products, and product composition data are utilized. Biodegradability is established at ingredient level. If an ingredient is a mixture, biodegradability is based on the percentage (by weight) of biodegradable components in the ingredient. Biodegradability is calculated by multiplying each product’s sales volume by its total composition of ingredients considered ultimately biodegradable. The percentage is calculated by summing these values and dividing them by the total volume of the washable universe.

Judgements and assumptions

• Water (also called eau de ville, eau déminéralisée): excluded from the scope as it is an inorganic compound.

• Items without test results: Items without biodegradability test results are considered not biodegradable.

Definitions

• Own Workforce: Employees who are in an employment relationship with the undertaking (‘employees’) and non-employees who are either individual contractors supplying labor to the undertaking (‘self-employed people’) or people provided by undertakings primarily engaged in ‘employment activities’ (NACE Code N78).

• Non-Employees: Non-employees in an undertaking’s own workforce include both individual contractors supplying labor to the undertaking (“self-employed people”) and people provided by undertakings primarily engaged in “employment activities” (NACE Code N78).

• Permanent Employees: A permanent employee is hired by a company with no predetermined end date and typically receives a full benefits package.

• Temporary employees (temporary contracts): Temporary contracts are used within the company to backfill absences and temporary leave, cover peak workload periods, and address country-specific practices for probation periods. It also includes all students, trainees and internships.

Total training hours: hours trained on learning platforms consumed by the own workforce.

Calculation methodologies

Characteristics calculated and reported include the employees:

• By gender distribution (male, female and not disclosed)

• By geographical distribution, by significant country

• By employment type which includes contract type namely: permanent, temporary, non‑guaranteed hours and non-integrated‑ acquisitions

• Diversity indicators (executives, management, other and non-integrated acquisitions)

• By age bracket (<30 years, 30–50 years, >50 years and unknown)

Judgements and assumptions

• Significant countries: Countries where the company has at least 50 employees which is at least 10% of the total employee count.

• Management levels: Executives are Vice Presidents, Presidents, and C-suite officers, with some minor exceptions. Management are Directors, Managers, and Experts, with some minor exceptions. Other refers to all other employees.

Definitions

• Employee Engagement Survey (EES): People are prioritized and encouraged to voice their views through various platforms, including the annual Employee Engagement Survey and regular feedback mechanisms. These efforts provide insights on engagement, inclusion, and integration, helping address issues proactively. The EES includes specific questions which are grouped to analyze the results in specific topics such as employee engagement, intention to stay, inclusion, unity and people management which are also disclosed. The number of participants in the EES is measured to calculate the participation rate.

• Global Management Team (GMT) List: Includes members of the Executive Committee, Job Bands B-E, TTH Creators in Job Band F, LTI-eligible employees in Job Band F.

• GMT Diversity: The process of identifying and tagging employees who are within the GMT list.

• Employee Engagement and Inclusion: Employee Engagement and Inclusion are measured annually as part of the EES, including scores for employees self-identifying as minorities. Regular surveys, meetings, and feedback sessions support focused action planning across Business Units. Results and target levels are reported post factum in the Compensation report to drive a workplace where everyone feels valued and heard.

Calculation methodologies

The method of processing and reporting Engagement & GMT diversity metrics involves the collaboration of several key roles within the organization.

Engagement surveys: Employee Engagement survey results are shared across all layers of the organization. Effectiveness of surveys is assessed by comparing survey results with previous years. Additional surveys capture employee and manager feedback at key lifecycle stages (namely recruitment, onboarding, anniversaries, and exit) providing evidence-based decisions to enhance engagement and retention.

The Employee Central report containing the population of employees in the company is the basis for the GMT listing, with the exclusion of certain countries like Belgium, Denmark, and Spain, as per the local regulations.

GMT Diversity (LTI Calculation): Ethnicity is mapped into three groups – ‘western European,’ ‘non‑western‑European,’ and ‘chose not to respond’ – in the EES.

Inclusion Index: The Inclusion Index involves three core questions namely, a sense of belonging, value of perspectives, and fair treatment at dsm-firmenich. The percentage of favorable responses is determined for each question. The overall Inclusion Index is calculated and compared against a target. Results are reported to relevant senior leaders.

Judgements and assumptions

• Exclusions GMT list: Specific regions/employees are excluded from the GMT Listing population due to local country law restrictions, these include Belgium, Denmark and Spain.

• Exclusions EES: Certain employees were excluded from the EES due to specific regulations or ongoing integration activities. The exclusions represent less than 2% of the workforce.

• Inclusion index: With regards to the Inclusion index, the rating scale is verified and if missing data is less than 5%, the Inclusion score can be taken as accurate. If missing data is more than 5%, manual calculations are performed.

Definitions

• TRIR-all: Total Recordable Incident Rate for all employees and contractors, per 100 FTE (200,000 hours).

• Health-rate-all: Frequency of all recordable health incidents (own + contractors) per 100 FTE.

• PSI rate: Process Safety Incident rate per 100 FTE.

• Contractor: Non-supervised contractors performing work on dsm-firmenich premises.

• Recordable incident: As defined by OSHA and ICCA, including injuries, occupational illnesses, and process safety events.

Calculation methodologies

SHE incidents are reported in internal systems, using a classification matrix based on OSHA for safety and health, and ICCA for process safety. Incidents are reviewed for completeness and accuracy.

FTE numbers for employees are reported according to ESRS S1-6. FTE numbers for contractors are based on actual numbers from gate logs, manual registration or invoices, or, if not available, reasonable estimates.

The company target on safety as well as the incident rates are calculated on a 100-man-day basis and exclude recent acquisitions. These are described as ‘entity-specific metrics’ following the approach defined in ESRS. These metrics have similar definitions to the ESRS-aligned metrics reported on the basis of 1,000,000 hours.

All incident rates are reported on a monthly basis internally on a rolling 12-month basis, and externally for the full calendar year.

Judgements and assumptions

• Work-related incidents: Incidents are judged to be work-related if an event or exposure in the work environment contributes to the condition. This requires judgement in borderline cases (e.g., travel status, home office injuries).

• Incident rate calculations: Rates are normalized to 100 FTEs = 200,000 hours based on 8 hours/day and 20.83 working days/month. This can be multiplied by five to align with the ESRS basis of 1,000,000 hours.

• Classification of cases: Cases are classified based on expert judgement. This is particularly relevant where cases may be near the threshold, such as first aid or medical aid cases, or repeated exposure (health) or single acute events (safety).

Definitions

Collective bargaining agreements: As per the Annex II of the Delegated Regulation, collective bargaining agreements refer to all negotiations conducted between:

• An employer, a group of employers, or one or more employers’ organizations; and

• One or more trade unions, or—where trade unions are absent—duly elected and authorized workers’ representatives, in accordance with applicable national laws and regulations, for the purpose of determining working conditions and terms of employment; and/or regulating relations between employers and workers; and/or regulating relations between employers or their organizations and a workers’ organization or workers’ organizations.

Calculation methodologies

Collective bargaining agreement coverage by region is determined based on country-level coverage reported through the regional human resources organization.

As no European Economic Area (EEA) countries at dsm-firmenich meet the ESRS thresholds, the total EEA percentage is reported.

Judgements and assumptions

• Coverage rate: Reported percentages are based on reporting by countries covering 89% of total headcount. For the purposes of this calculation, countries that have not reported a percentage are assumed to have no collective bargaining agreement in place.

Definitions

• Adequate wages: a wage that provides for the satisfaction of the needs of the worker and his/her family in the light of national economic and social conditions.

• Living wage: the wage required to purchase the goods and services needed to meet a minimum acceptable living standard for workers and their families.

• Unadjusted gender pay gap: the difference between what men typically earn overall in an organization compared to women, irrespective of their role or seniority.

• Adjusted gender pay gap: the difference between what men typically earn overall in an organization compared to women, taking into account their role or seniority.

• Annual total remuneration ratio: the ratio of the annual total compensation for the organization’s highest-paid individual (the CEO) to the median annual total compensation for all permanent dsm-firmenich employees (excluding the highest-paid individual).

Calculation methodologies

Adequate wages: Wage data, consisting of base salary and if applicable other guaranteed elements, is extracted and compared against the WageIndicator minimum wage benchmarks, confirming that no employee is paid below the benchmark level.

Gender pay gap: Gender pay gap is calculated using the remuneration components: Annual base salary, STI, Prime, LTI, social security, pension and benefits in kind.

Adjusted gender pay gap: Weighted average of the gender pay gap taking into account the employee categories based on job bands.

Annual total remuneration ratio: The remuneration ratio is calculated on the Reference date (i.e., 31 December) of the relevant financial year (‘RFY’). Only permanent employees of a consolidated dsm-firmenich subsidiary on the Reference date are considered; there are no corrections for hires or terminations during the year, nor for the impact of any M&A activities closed during the year.

The total renumeration of the highest paid person includes:

• The actual paid base salary in the RFY

• The STI paid in the RFY year related to the RFY-1

• The LTI PSU grant times the fair value IFRS price per PSU

• The actual employer social charges (Pension + social security cost) and benefits paid in RFY

To calculate the denominator of the remuneration ratio, the total annual remuneration (in EUR) per employee, excluding the highest paid employee (the CEO) is calculated. This total includes Annual base salary, STI, Prime, LTI, social security, pension and benefits in kind. Once the total remuneration is calculated, the median value from the dataset is determined. This value is the denominator of the remuneration ratio.

Judgements and assumptions

• Employee headcount: calculations for these metrics is based on the year-end workforce. Employees who have joined or left the company during the period are not included in the calculations due to incomplete data. Non-integrated acquisitions (not integrated into central HR systems) and support staff such as trainees, students, interns, apprentices are not taken into account.

• Living wage benchmark: The living wage is based on the typical family size of 2 adults and number of dependents adjusted to national fertility rate and employment rate statistics. Price data at the 25th percentile.

• Currency conversion: Currency conversions used are the average exchange rates as used in the preparation of the Consolidated Financial Statements.

• Remuneration ratio: Social security contributions (Employer part), Employer contributions to company pension plans and benefits in kind are included for the CEO on actual basis and for other employees based on a multiplier..

Definitions

• Code of Business Ethics: Outlines the company’s values and commitments to each other, communities, and the planet. It is both a reference document and a day-to-day manual, guiding the company’s people on how to do and do not conduct business. The company’s Code is universal. It applies globally to everyone in the company.

Calculation methodologies

Incidents complaints and severe human rights impacts: Incidents that breach human rights standards can be reported directly to local HR contacts, line managers, legal partners or the business ethics team. Reports can also be submitted anonymously through the Speak Up platform, which is operated by an external provider and is accessible 24/7.

All investigations into allegations reported via Speak Up are conducted impartially and are reported to the Group Ethics Committee at least once a quarter.

In addition to the Speak Up platform, the company runs an Employee Engagement Survey every year in which employees can provide their anonymous feedback.

Judgements and assumptions

• Incident data: Data is based on self-reporting and therefore may not capture all incidents.

Definitions and Calculation methodologies

Spend coverage with transparency platforms:

The assessment and audit of procurement direct and indirect suppliers for the spend coverage was performed using EcoVadis ratings, Sedex SMETA, EcoVadis Vitals, TfS audits, and B Corp certifications. All suppliers are considered with the exclusion of NGOs, government agencies/offices, customs authorities, research institutes, schools, universities, individuals as these certifications are not relevant for these suppliers.

Suppliers average EcoVadis score: The calculation method is a weighted average based on supplier spend and the overall score on EcoVadis rating.

Spend coverage by suppliers with SBTi – validated targets: Covers direct and indirect procurement suppliers spend with SBTi status “target set”. Calculated as % of rolling 12-month total group procurement spend realized with a supplier with a “target set” SBTi status.

Share of suppliers with human rights risks trained: Tracks which suppliers have completed human rights training through EcoVadis or TfS Academies, or those who have demonstrated minimum 10 points improvement in their EcoVadis Human Rights pillar score. The KPI specifically targets suppliers which were underperforming (<60pts) as of January 1st, 2025, on the EcoVadis Human Rights pillar, considering both formal training or 10pts score improvement as valid evidence of enhanced human rights awareness and compliance. The performance includes those trained in 2024 (baseline) and 2025.

Share of Procurement staff members trained on Responsible Sourcing stakes: Measures the percentage of the procurement staff community (hires included until October 2025) that has completed Responsible Sourcing training since January 2024. These training sessions were developed internally based on external platforms such as TfS, WBSCD or UNCG and customized to the company’s internal objectives. It covers key topics such as supplier sustainability standards, human rights, ethical sourcing, risk identification, and the use of the Responsible Sourcing dashboard, ensuring that procurement professionals are equipped with the knowledge and tools to drive responsible sourcing practices across the organization.

Certified raw materials at source: The number of raw materials codes fully certified at source. It relies on two separate lists that are updated annually. First, the raw materials to be certified list, which is defined each year based on business requests or requirements arising from due diligence remediation plans. Second, the list of accepted standards, which is reviewed annually and built on internationally recognized schemes, aligned with client expectations and informed by relevant sector benchmarks and IUCN biodiversity standards work. Only raw materials codes appearing on the list and certified according to the accepted standards list are counted in this indicator.

Judgements and assumptions

• Reliance on third party data: It is assumed that the data collected from third-party sources, supplier self-assessments, and on-site evaluations are accurate and reliable.

• Incident data: Data is based on self-reporting and therefore may not capture all incidents. If employees fear retaliation, it may lead to underreporting of violations.

Definitions

• Key natural ingredients: Ingredients that are assessed periodically for a combination of:

  • Risk level

  • Either found in or produced by nature

  • Considered a priority by the Business Units and customers

• Naturals sourcing program completion rate: completion of due diligence framework for a value chain on key natural ingredients responsibly sourced

Calculation methodologies

Naturals sourcing program completion rate: calculated as value chains (ingredient × country) completed the due diligence framework (risk screening, assessment, action plans, monitoring, report) divided by the total number of value chains in the 2030 roadmap.

An annual review of in-scope key natural ingredients is performed with the Business Units, based on risk and business criticality, to define the denominator.

Desktop due diligence applies to all categories (including Naturals), leveraging external platforms (EcoVadis, SMETA, UEBT, etc.). On-site due diligence is focused on Natural ingredients.

On-site due diligence is performed on key ingredients with due diligence reports disclosed. Salient risks are addressed by, among others, supplier-level action plans, coalitions, and projects at source. A key requirement, in line with international guidelines, is to ensure risks are acknowledged and action plan defined. Partnerships leveraged include UN Global Compact, UEBT and sourcing coalitions with NGOs.

The stages of the due diligence framework:

1. Desktop risk screening & assessment

2. Risk mitigation & prioritization

3. Pre & on-site assessment

4. Prevention, mitigation & remediation

5. Monitoring, handover & ongoing learning

6. External communications & reporting

Judgements and assumptions

• Reliance on platforms: It is assumed that the data collected from third-party platforms is accurate and reliable. Platforms have been reviewed and assessed as fit for purpose.

Definitions

• Premix price: Average annual price of the Premix (vitamin-mineral blend).

• Premix fortification/inclusion rate: Percentage of premix added to the food product by weight.

• Intake per person: Annual amount of product consumed per individual as per guidelines provided by United Nations Children’s Fund (UNICEF), Organisation for Economic Co-operation and Development (OECD), United Nations International Multiple Micronutrient Antenatal Preparation Multiple Micronutrient Supplements (UNIMMAP MMS), World Health Organization (WHO).

• Compliance rate: Measures the extent to which individuals comply with prescribed product consumption guidelines and protocols established by recognized authorities, including the National Library of Medicine (NLM), World Food Programme (WFP), OECD, South African 2003 Regulations on Fortification of Foodstuffs, and the Global Fortification Data Exchange (GFDx).

• Cost per unit of intervention: Annual cost of delivery of the fortified ingredient per person per product category.

Calculation methodologies

The reporting of number of lives reached (‘person’ or people) by dsm-firmenich solutions addressing micronutrient deficiencies are based on sales data of the specific solutions in scope. These solutions are classified under 7 product sub-categories

• Emergency & Therapeutics (E&T)

• Fortified Rice Kernels (FRK)

• Staple Food Fortification (SFF)

• Oil Fortification (OF)

• Multiple Micronutrient Supplements (MMS)

• Multiple Micronutrient Powders (MNP)

• Vitamin A Capsules.

Annual Nutrition Improvement (NI) sales are extracted from the financial reporting system and split by product subcategory based on NI customer classification.

Financial assumptions for key variables are validated annually. In these reviews, each variable’s values are checked each year. The cost of intervention per person is then calculated by applying the premix price to the assumed intake per person.

The calculation of people is obtained by dividing the sales per subcategory by the cost of intervention. The total number of people is obtained by summing up the people through each subcategory.

Judgements and assumptions

• Subcategory-specific variables: Each product subcategory uses annually validated financial and non-financial variables — such as premix price, inclusion rate, annual intake, and compliance rate — to calculate cost per intervention, which determines the total number of lives reached.

• Overlap: Each sub-category, or intervention, targets a specific group of beneficiaries (people) in specific geographies and with specific needs. It is considered unlikely that beneficiaries will receive products from multiple sub-categories. Based on this, we assume no overlap between the sub-categories.

• Calculation assumptions: Each subcategory utilizes assumptions to arrive at the cost of intervention for each subcategory and leads to the number of lives reached annually. Each sub-category relies on a combination of assumptions and judgements based on third-party information sources and/or internal expert judgement. Every assumption is subject to review.

• Reach: reach is based on annual performance. It is not cumulated year-on-year.

• Methodology update: In 2025, the methodology was updated to include sales to smaller customers (<€100,000), and to calculate Oil Fortification separately from Staple Food Fortification

Calculation methodologies

Confirmed incidents of corruption and bribery: Incidents of corruption and bribery can be reported directly to local HR contacts, line managers, legal partners or the business ethics team. Reports can also be submitted through our Speak Up platform, which is operated by an external provider, allows for anonymous reporting and is accessible 24/7. All investigations into allegations reported via Speak Up are conducted impartially and are reported to the Group Ethics Committee at least once a quarter.

Judgements and assumptions

• Incident data: Data is based on self-reporting and therefore may not capture all incidents.

• Gifts and entertainment: These must be, among others, reasonable, proportionate and appropriate in the circumstances. They can never be provided in cash or cash equivalents (such as gift certificates or vouchers) and must comply with internal standards.

• Sponsorships and donations: Donations on behalf of dsm-firmenich are allowed and should aim to generate a positive impact on the communities in which we operate. Receiving donations and sponsorships is only permitted under certain circumstances. In any case, when giving or receiving donations or sponsorships, internal standards must be complied with. The company’s business activities and public positions are strictly non-political in all circumstances. The company does not publicly endorse, financially support, or donate to any political party, candidate, or any religious organizations. dsm-firmenich respects its employees’ right to engage in the political process – but only on an individual basis, not as representatives of the company.

• Relationships with third parties: Are permitted if they comply with the internal standards.