What is the environmental impact of cosmetics?

The gap between what we think we know and what the figures show runs throughout this guide. For shampoo, the bottle accounts for 15 to 30% of the total carbon footprint, the formulation 10 to 25%, and the use phase, a few minutes of hot water in the shower, 40 to 70%. For a face cream, the reverse is true: the use phase is marginal and packaging dominates. Two products from the same brand, two radically different hierarchies of impacts. There is only one reliable tool for making the choice: life cycle assessment (LCA).

This guide is written for product managers, CSR managers and brand founders who want to scope a proper ecodesign process, rather than simply tick a box. It covers the method, the sources that matter in an LCA (and those that matter less than we think), the regulatory framework tightening in 2026-2027, the levers that work, costs and practical first steps.

1What ecodesign means for a cosmetic

Ecodesign means the documented reduction of a product’s environmental impact over its entire life cycle, from growing the raw materials to the bottle’s end of life. For a cosmetic, it covers six steps:

• Growing and extracting ingredients
• Processing into active ingredients
• Manufacturing the finished product in the factory
• Packaging (bottle, carton, seal, pump)
• Transport and distribution
• Consumer use phase and packaging end of life

The reference standard is ISO 14006, based on life cycle assessment (LCA) as defined by ISO 14040 and 14044.

A few common misunderstandings are best cleared up at the outset. Ecodesign is not a label (Cosmos, Ecocert and NATRUE certify composition, not environmental impact). Nor is it a "natural" or "clean" claim (ISO 16128 describes origin, not footprint). And it is not an EcoBeautyScore score produced by a consortium tool: a score is an output, not a process. Ecodesign is quantified, iterative and substantiated by LCA, and makes concrete changes to the product to reduce its actual impact.

Watch out for the most common confusion in practice: product LCA and Bilan Carbone® are different things. Bilan Carbone® measures an organisation’s emissions, while LCA measures a product’s impacts across 16 indicators. A serious cosmetics manufacturer needs both. Our dedicated comparison explains the details.

2A cosmetic’s four impact sources and their orders of magnitude

All the cosmetics LCAs we have led or read point to the same finding. Four sources concentrate most of the impact, with important differences between a rinse-off product and a leave-on. Ranking them correctly is the first step towards avoiding a process that invests time and money in the wrong lever.

Use phase: 40 to 70% for rinse-off products

For shampoo, shower gel, conditioner or rinse-off body care, the use phase represents 40 to 70% of the carbon footprint of the full life cycle. The user heats water for 3 to 8 minutes, and the energy involved (electricity or gas depending on the national mix) exceeds everything the brand has spent on manufacturing, transporting and packaging the product.

For shampoo, the Cosmetics Europe study that established the product’s PEFCR rules using the Commission’s method (published in 2018) identifies the use phase as the largest contributor across almost all impact categories, because of the water heated in the shower. For leave-on products (face cream, serum, lipstick), the reverse is true: the use phase is marginal (less than 5%), and formulation and packaging dominate. The distinction between rinse-off and leave-on is not cosmetic: it shapes the assessment.

Packaging: 15 to 50% depending on the category

Packaging concentrates 15 to 30% of the carbon footprint of an average rinse-off cosmetic, and 40 to 50% for a leave-on that is not used in the shower. Glass, instinctively seen as "premium", is often the worst environmental choice: a 50 ml glass perfume bottle generates 3 to 5 times more CO₂e than an equivalent plastic bottle, because its melting furnace reaches 1,500 °C.

• Recycled PET: a good carbon choice, provided the supplier documents the actual recycled content
• Sugar cane bioplastic: good for carbon, less good for eutrophication
• Recycled aluminium: excellent for aerosols
• Refillable glass: defensible only with more than 3 actual use cycles

The key point: ecodesigning packaging does not mean reflexively "switching to glass". It means comparing quantified options and making an informed choice.

Formulation: 10 to 25% of the footprint

Formulation typically accounts for 10 to 25% of a cosmetic’s carbon footprint. Less than packaging, and much less than the use phase for a rinse-off product. But it accounts for a significant share of other indicators: biodiversity, land use and water consumption. A plant ingredient grown in intensive monoculture can have a greater ecosystem footprint than a synthetic equivalent.

Typical examples: uncertified palm oil weighs heavily on deforestation, but its carbon footprint is lower than coconut oil’s. Plant-based glycerine performs better on biodiversity but often worse on carbon than the petrochemical version. These choices are made using emission factors from ADEME’s Base Empreinte or Ecoinvent within an LCA, rather than intuition.

Production site and transport: 5 to 20%

Factory manufacturing accounts for 5 to 15% of the carbon footprint, and transport 2 to 5%, even from Asia. But these ranges depend directly on the site’s electricity mix and the process heat used. The same formulation made in a coal-powered or hydroelectric-powered factory sees its manufacturing footprint vary by a factor of 1.3 to 2.

A particular case is overseas sites or those with a substantial sea freight component (Martinique, Guadeloupe, Réunion): the transport module often weighs twice as much as that of an equivalent mainland site, and the local electricity mix must be used as it is, rather than mainland France’s, to avoid completely distorting the result.

An accurate product LCA therefore starts with real knowledge of the manufacturing site: actual electricity consumption (kWh/kg of product), electricity origin (grid mix or green PPA), process heat, water consumption and effluent treatment. These data are rarely available in secondary databases; they are collected site by site and shift the result.

3LCA, PEF, EcoBeautyScore: the methods and how they fit together

The methodological landscape has taken shape over three or four years. Knowing which framework you are entering avoids paying twice for the same data. Here is the map, in the order a brand generally encounters it.

ISO 14040/14044 LCA: the foundation of every process

ISO 14040 (principles) and ISO 14044 (requirements) are the international reference standards for every life cycle assessment. They are the basic building block underpinning everything else: European PEF is a more prescriptive derivative, and industry scores (including EcoBeautyScore) refer to them. An ISO LCA supports internal ecodesign, communication (with critical review) and future regulatory requirements.

This is the starting point we recommend because it produces a reusable asset: data collected for an ISO LCA can then feed into a PEF, an EcoBeautyScore or CSRD scope 3 without redoing 80% of the work. One investment, four possible uses. The full process is in our first product LCA guide.

Cosmetics PEF: the European method

PEF (Product Environmental Footprint) is the European Commission’s harmonised method for calculating a product’s footprint across 16 indicators. PEFCR (Product Environmental Footprint Category Rules) are the sector-specific rules applying the method. For cosmetics, no PEFCR has yet been adopted with the Commission and the JRC: the reference work remains Cosmetics Europe’s work on shampoo, published in 2018.

Cosmetics Europe’s work sets the functional unit (1 gram of product applied or rinsed, depending on the category), the boundary (mandatory cradle-to-grave), allocation rules and preferred databases. This framework will serve cosmetics environmental labelling if France or the EU decide to extend it to the sector. No timetable has been announced to date.

Industry scores (EcoBeautyScore)

EcoBeautyScore (a consortium led notably by L’Oréal, with around fifty brands), Beauty Impact Index and a few other private initiatives offer scores that can be displayed in shops or online. They are based on the cosmetics PEFCR and produce a grade from A to E. Their real benefit is harmonising consumer labelling and avoiding a battle of scores.

There are two limitations. First, these are outputs, not inputs: a score does not tell you how to ecodesign; it tells you where you stand. Second, these scores are not regulatory standards: if the ECGT directive or a future French decree requires another framework, adjustments will be needed. The right order is an ISO LCA or PEF as the foundation, then connecting an industry score to it for labelling. Not the reverse.

4The framework taking shape in 2026-2027

Three European texts will gradually regulate environmental communication in the cosmetics sector. There is no need for a legal reading: what matters is that unquantified "green" claims will disappear, and it will be necessary to produce the calculations behind every promise. That is exactly what an LCA is for.

ECGT directive: an end to vague claims, upon transposition

Directive (EU) 2024/825 (known as ECGT, for Empowering Consumers for the Green Transition) amends the European Unfair Commercial Practices Directive. It was due to apply across the Union on 27 September 2026, but France has not yet transposed it.

Once the directive is transposed in France, these statements will no longer be usable without acceptable substantiation:

• "eco-responsible", "environmentally friendly", "green"
• "natural", "climate-neutral", "zero emissions"
• Any unsourced reduction percentage ("-30% footprint")

A reduction percentage is substantiated by an LCA covering the entire life cycle; a generic term requires an officially recognised ecolabel, which an LCA does not replace, and neutrality based on offsetting is prohibited in all circumstances. "Natural" and "organic" are not generic environmental claims within the meaning of ECGT: they are composition claims governed by general rules on misleading practices and substantiated by an ISO 16128 calculation or certification (Cosmos, Ecocert) attesting to composition, not impact. But a brand displaying "carbon footprint reduced by 30%" today without a verifiable LCA is racing against time.

PPWR regulation: mandatory recycled content quotas

Regulation (EU) 2025/40 (known as PPWR, for Packaging and Packaging Waste Regulation) has been in force since February 2025, directly applicable without transposition. It requires:

• Minimum recycled content in plastic packaging: 10% in 2030, 25 to 30% in 2040 depending on cosmetics categories
• Actual recyclability assessed in classes A, B or C, with class C the minimum
• A ban on certain formats (particularly hotel miniatures for rinse-off products)

Combined with ECGT once transposed, selling a cosmetic labelled "ecological packaging" without an officially recognised ecolabel will be impossible: the actual recycled content and recyclability will have to be displayed instead, with supporting evidence. Brands that have not started mapping their packaging suppliers will need to accelerate.

Green Claims Directive: third-party evidence before publication

The Green Claims Directive, proposed in 2023 to supplement ECGT, envisaged external validation before a brand could publish an explicit environmental claim: a compliant LCA, an independent critical review and a technical dossier accessible to the regulator. Negotiations have been suspended since June 2025, when the Commission announced its intention to withdraw the text, without a formal withdrawal to date. If it goes through, communicating "-20% emissions per bottle" will require submitting a dossier to an approved third-party body.

5Ecodesign levers that shift the LCA

Once the LCA has identified the hotspots, we know where to act. Ecodesign work then focuses on four main families of levers. None works everywhere: their combination depends on the product category and the LCA result. Here are those we have seen actually shift scores in practice.

Formulation: avoiding the "natural = better" trap

Three concrete formulation levers:

• Reduce the concentration of high-impact ingredients (exotic oils, heavy preservatives, complex fragrances) where performance allows.
• Replace petrochemical emulsifiers with sourced plant-based alternatives that do not come from deforestation.
• Use upcycling where relevant: agricultural co-products (grape pomace, olive stones, citrus peel) that reduce the allocated burden in the LCA, a lever widely used by tropical bio-based sectors (banana, cane, cocoa).

The trap to avoid is switching to "100% natural" for perception without checking the impact through LCA. A sulphated surfactant based on imported coconut oil can have a footprint 2 to 3 times higher than a synthetic surfactant produced in Europe. The test is always the same: replace nothing before comparing the two emission factors, weighted by the dose in the formulation.

Packaging: choosing with figures

Four packaging levers, ranked by decreasing impact:

• Lightweighting: -15 to -25% material in a bottle = a direct impact reduction, with no trade-off.
• Post-consumer recycled material: PET-rPET at 50-100% divides the bottle’s carbon footprint by 2 to 3, provided the supplier documents the actual content.
• Single-material packaging (without a complex pump): does not immediately change the LCA, but prepares for PPWR compliance.
• Refills: effective only if actual refilling exceeds 3 cycles. Below that, they are counterproductive.

Each of these levers is quantified in a preliminary LCA before deciding. This is exactly what screening is for: testing 3-4 variants, quantifying their comparative impact and choosing before committing to moulds and suppliers.

Use phase: the hidden lever for rinse-off products

For rinse-off products, this is the number one lever. Three R&D directions:

• More concentrated or water-soluble textures: they can reduce rinsing time by 20 to 30%, and the use-phase impact by the same amount.
• Low-temperature formulation: communicating "effective at 25°C", just as detergents have learnt to do.
• Solid formats (shampoo bars, syndet soaps): they eliminate intensive rinsing and often divide the total impact by 2 to 3.

This lever is often underfunded because it requires formulation R&D rather than a change of supplier. Yet it is the one that will shift labelling scores when they arrive.

Manufacturing site: the foundation that pays off across all subsequent LCAs

The final lever, often the blind spot of processes conducted remotely, is the production site. A "site" LCA at factory level reveals optimisations invisible at product level: heat recovery from autoclaves, a green PPA, reduced process water consumption and effluent recovery. Once documented, these optimisations feed into the LCAs of products manufactured at the site, using actual emission factors rather than penalising generic averages.

The sequence that works is factory LCA first, then product LCAs calibrated to the actual factors. It is the most robust long-term approach because the site LCA becomes a reusable strategic asset: every new product launched benefits from the documented factory foundation, and any extension or new production line project connects to it without starting from scratch!

6Where to start

Four steps, in order. The whole process takes 4 to 8 months for a first complete process for one product, less if starting with screening. This is not a theoretical protocol: it is the sequence we recommend on projects and follow with our clients.

Choose the right pilot product

Target a product that ticks at least one of these three boxes:

• The best seller: results will represent your business.
• The most exposed to current environmental claims: the one risking ECGT non-compliance as soon as it is transposed.
• The one a distributor is already asking you to document.

Avoid starting with a niche marketing product. The aim is for the learning to serve the whole catalogue afterwards.

Check funding

Before contacting a provider, check your eligibility for Bpifrance’s Diag Éco-conception (fewer than 250 employees, turnover below €50 million) and Diag Décarbon'Action. The form takes 30 minutes, and validation 2 to 4 weeks. If eligible, the remaining cost falls to €5,400 to €7,200 excluding VAT for a complete assignment. Do not budget anything before getting that answer.

Map the site and critical suppliers

The longest step, and by far the most rewarding. For the pilot’s production site (owned or subcontracted), gather four data blocks: annual electricity consumption and supply mix, process heat, water and effluent treatment. In parallel, list your 5 most critical raw material suppliers (volume and presumed footprint) and send them a short questionnaire: geographical origin, process and certifications. These data will serve whichever method is chosen (ISO LCA, PEF, EcoBeautyScore) and whichever requirement materialises first.

If you are building or extending a factory, this is the ideal time to start a site LCA at the design stage. Structural, envelope, energy system and water decisions determine 70 to 80% of the building’s final environmental impact. Once plans are approved, room for manoeuvre shrinks drastically. The advantage of starting early: material and equipment choices are still open, and you leave with a foundation that will feed all future product LCAs.

Choose between screening and a full LCA

Two starting points, depending on your situation:

• Screening: €5,000 to €12,000 excluding VAT, 4 to 8 weeks. The recommended starting point if you have never carried out an LCA. Enough to identify hotspots and steer the process.
• Full ISO 14040 LCA: €12,000 to €25,000 excluding VAT, 3 to 6 months. Essential if you have a near-term regulatory deadline (ECGT transposition, a claim to substantiate).

In both cases, the right internal contact is the formulation R&D manager or industrial manager, not the marketing director. They know the processes, suppliers and data.

7Key takeaways

Cosmetics ecodesign is no longer a communication exercise. Between the ECGT directive, applicable since September 2026 in countries that have transposed it, PPWR tightening packaging requirements, PEFCR finalising the European method and the Green Claims Directive, suspended but not withdrawn, which would require external evidence, the framework is closing in over 18 to 24 months. Brands preparing today will be ready. Those waiting for an inspection or delisting will have six months to do eighteen months’ work.

• The use phase accounts for 40 to 70% of the footprint of a rinse-off product. More than packaging, more than formulation. It is the number one lever and the most underfunded.
• Packaging often weighs more than formulation in carbon terms. Glass = no by default, recycled PET = often yes, refills = yes with more than 3 actual cycles.
• ISO 14040 LCA = the right starting point. The data collected then serve PEFCR, CSRD scope 3 and consumer scores. One investment, four uses.
• Bpifrance’s Diag Éco-conception: 60 to 70% funded for SMEs. Remaining cost €5,400 to €7,200 excluding VAT. A funding lever to activate first.
• Upstream site LCA + downstream product LCAs: the most robust approach. Starting with the factory where possible makes every later product LCA pay off.

The path is mapped out. An €8,000 screening identifies your hotspots in two months. A full LCA at €15,000-25,000 gives you an asset reusable across your catalogue, co-funded at 60-70% by Bpifrance for SMEs. And the second LCA costs half as much as the first. You might as well start now, while the timetable allows!

Main sources: ISO 14040:2006 and ISO 14044:2006 (LCA methodology), ISO 14006 (integrating ecodesign), Directive (EU) 2024/825 ECGT, Regulation (EU) 2025/40 PPWR, shampoo PEFCR (Cosmetics Europe, Integrated Environmental Assessment and Management, 2018), proposed Green Claims Directive (European Commission, negotiations suspended since June 2025), ADEME’s Base Empreinte, FEBEA 2024 report on the French cosmetics market, Bpifrance Diag Éco-conception (2025-2026 specification).

Further resources

Frequently asked questions

1Is a "natural" cosmetic less polluting?

Not necessarily. Natural certification (Cosmos, Ecocert, ISO 16128) describes the origin of ingredients, not their environmental impact. An imported tropical plant oil can have a greater carbon or biodiversity footprint than a European synthetic equivalent. Only an LCA compares ingredients on a quantified basis.

2How much does a cosmetics product LCA cost?

Between €5,000 and €25,000 excluding VAT, depending on depth. Simplified screening costs €5,000 to €12,000 for 4 to 8 weeks. A full ISO 14040 LCA costs €12,000 to €25,000 for 3 to 6 months. Bpifrance’s Diag Éco-conception funds 60 to 70% of the cost for SMEs, leaving €5,400 to €7,200 excluding VAT to pay.

3What is the difference between EcoBeautyScore and an LCA?

EcoBeautyScore is a consumer score (A to E) based on a sector-specific cosmetics PEF. It is a labelling tool. LCA is the underlying scientific method that produces the data. A brand needs the LCA to ecodesign and feed a possible score; it can then use EcoBeautyScore to communicate. Neither replaces the other.

4Does the ECGT directive apply to cosmetics SMEs?

Yes, with no size threshold. Directive (EU) 2024/825 applies to all companies marketing products bearing environmental claims in Europe, whatever their size. It has applied since 27 September 2026 in countries that have transposed it; in France, the transposing law has not yet been passed. Once in force, a cosmetics SME communicating "eco-responsible" or "carbon-neutral" will be non-compliant, even with an LCA: a generic claim requires an officially recognised ecolabel, and neutrality through offsetting is prohibited.

5Why would shampoo’s footprint be greater during use than production?

Because the consumer heats several litres of water for 3 to 8 minutes in the shower each time it is used. Across all uses of the bottle, the energy (electricity or gas) to heat this water exceeds what the brand has spent on manufacturing, packaging and transport combined. For rinse-off products, this phase often represents 40 to 70% of the total carbon footprint.

6Should LCA come before ecodesign or the other way round?

LCA first, always. Ecodesigning without LCA means optimising by intuition, with a high risk of choosing the wrong lever (replacing plastic with glass, for example, which often worsens the carbon footprint). LCA identifies the 2-3 sources concentrating 70-80% of the impact, and these are the ones to work on. Ecodesign without LCA is a marketing process, not a technical one.

7Can my factory LCA serve my product LCAs?

Yes, and that is the right order where possible. A production site LCA documents actual emission factors (electricity mix, heat, water) that then feed into the LCAs of products manufactured there. Without these site data, product LCAs use generic averages that systematically penalise them. The site investment pays off from the second or third product LCA.

8Which sources weigh most in a cosmetics product LCA?

For a rinse-off product such as shampoo or shower gel, the use phase weighs most: 40 to 70% of the carbon footprint because of water heated in the shower, ahead of packaging (15 to 30%) and formulation (10 to 25%). For a leave-on product (face cream, serum), use accounts for less than 5% and packaging reaches 40 to 50%. Factory manufacturing represents 5 to 15% and transport 2 to 5%, even from Asia. Formulation remains decisive for other indicators: biodiversity, land use and water consumption.

9What ecodesign levers are available for cosmetics formulation?

The first is reducing the concentration of high-impact ingredients (exotic oils, heavy preservatives, complex fragrances) where performance allows. Next come replacing petrochemical emulsifiers with plant-based alternatives not originating from deforestation, and upcycling agricultural co-products (grape pomace, olive stones, citrus peel). For rinse-off products, formulation also affects the use phase: textures reducing rinsing time by 20 to 30%, low-temperature effectiveness and solid formats. Every substitution is checked through LCA, as a surfactant from imported coconut oil can emit 2 to 3 times more than a European synthetic one.

10What levers are available for cosmetics packaging?

In order of impact: lightweighting the bottle (-15 to -25% material directly reduces the footprint), switching to 50 to 100% post-consumer recycled PET, which divides the bottle’s carbon footprint by 2 to 3 if the supplier documents the actual content, adopting single-material packaging to prepare for PPWR compliance, and offering refills, effective only beyond 3 actual cycles. Switching to glass is rarely a good choice: a 50 ml glass perfume bottle emits 3 to 5 times more CO2e than an equivalent plastic bottle. Each option is quantified in a preliminary LCA before launching the moulds.

Ask Guillaume your questionA personal reply within 24 working hours

Send my questionor: [email protected]