Short read
The packaging industry spent decades learning that contamination kills recycling at scale. Not all plastics are the same. Not all paper grades are the same. Now the textile industry is learning the same lesson, but faster, under more regulatory pressure, and with billions of euros of new recycling infrastructure depending on the answer. A cotton recycler needs feedstock at 95% purity. Syre needs polyester. Infinited Fiber needs a minimum of 88% cotton. Fall below those thresholds and the batch fails. But purity alone is no longer enough. The material also has to be verified, and most brands cannot verify what their own products are made of.
Some can. H&M traces 200 million garments from fibre to shelf. adidas tracks recycled polyester in real time. But when their verified products enter the same collection stream as unverified ones from brands that never bothered, the batch is only as good as its weakest input.
The EU is responding with compliance fines of up to 4% of annual turnover. At the scale of adidas or Nike, that runs into hundreds of millions. But fines alone will not solve this. France puts 891,000 tonnes of textiles on the market each year and collects just 289,000. The remaining 602,000 tonnes are not lost. They are being worn by customers walking into stores to buy more.
The brands investing in recycling plants are often the same brands operating thousands of retail stores. The feedstock is already there. The customer is already there. The question is whether the industry can get out of its own way long enough to connect them.
Paper recycling has operated with contamination tolerances for decades. Under the European standard EN 643, a paper bale is permitted a maximum of 1.5% non-paper components for most grades. Above that threshold, the batch is rejected or downgraded. The economics of the entire sorting and processing chain depend on that tolerance being respected.
Plastics learned the same lesson. Not all plastics are the same. PET, HDPE, and polypropylene each require their own processing stream. A PET recycling line cannot accept polypropylene. A single misidentified bottle in a sorted batch creates a defect in the output. The industry solved this through standardised resin identification codes, automated sorting by material type, and decades of infrastructure investment.
Textiles are now entering the same phase. They are scaling toward industrial recycling capacity, but without the identification infrastructure that packaging built over 30 years. Not all textiles are the same. Polyester is not cotton. Nylon is not wool. A polycotton blend cannot be processed on a line built for pure polyester. Each recycler needs a specific feedstock at a specific purity. And unlike packaging, where the material type is often visible and identifiable, a textile product's composition is hidden inside the garment: the fibres, the coatings, the thread, the elastic, the lining.
But textiles have one advantage that packaging never had. A customer knows what brand their shoes are. Nobody knows what brand their plastic bottle is. That brand identification at the point of return is an asset the textile industry has not yet learned to use.
Paper recycling has operated with contamination tolerances for decades. One contaminant above threshold and the batch is rejected. Textiles are entering the same phase, but without the identification infrastructure that packaging built over 30 years.
The European Commission's 2022 textile strategy identified more than 160,000 companies in the EU's textile and clothing industry , and that figure does not include the non-EU brands selling into European markets. The number with genuine product-level material traceability is vanishingly small. But those that have built it deserve recognition, not because they are more virtuous, but because they found a commercial reason to invest before regulation required it.
H&M Group published its full supplier list in 2013, one of the first major retailers to do so. Since then, it has built a blockchain-based traceability system with TextileGenesis that tracks materials across five to six supply chain tiers, from fibre producer to finished garment. By the end of 2022, the system was tracing over 200 million garments. By 2024, 94% of the polyester in H&M products was from recycled sources.
H&M then went further. In 2024, it co-founded Syre, a textile-to-textile polyester recycling venture, with a $600 million offtake agreement over seven years. Syre's first production plant in North Carolina is expected to be operational in 2026. Construction of a larger facility in Vietnam is planned to begin in 2027. The target: over 3 million tonnes of recycled polyester by 2032, produced from used garments rather than plastic bottles.
This is not philanthropy. H&M has seen first-hand what happens when the recycling supply chain does not hold together. The traceability programme and the Syre investment are, in part, an insurance policy. H&M needed to secure a recycled fibre supply chain it could control. That required knowing exactly what its products are made of, tracking them through the supply chain, and ensuring the material could be verified when it reached the recycling plant. Traceability was not a compliance exercise. It was a prerequisite for the $600 million Syre deal.
adidas built a parallel capability through its partnership with TrusTrace, capturing real-time material data as products move through the supply chain. By 2023, 99% of the polyester in adidas products was recycled. The company now participates in the Closing the Footwear Loop project, a collaborative effort to build post-consumer sorting infrastructure across Europe, and has set a target of sourcing 10% of its polyester from recycled textile waste by 2030.
Patagonia took a different path entirely. It made transparency the product. Every item on its website lists specific material composition, Bluesign-approved status, recycled content, and Fair Trade certification. In 2025, the company published its first comprehensive impact report, and used it to acknowledge that it had not yet met its own emissions reduction targets. In a market where consumers increasingly distrust sustainability claims, being honest about what you have not achieved builds more credibility than overstating what you have. For Patagonia, transparency is not a cost centre. It is the brand.
Each of these companies found a commercial reason to know what their products are made of. H&M needed it for its recycling supply chain. adidas needed it to verify its recycled content commitments. Patagonia needed it because its customers will not accept anything less. They acted before they were required to, and the investment is already paying off in ways that compliance alone cannot explain.
H&M has seen first-hand what happens when the recycling supply chain does not hold together. The traceability programme and the Syre venture are not philanthropy. They are an insurance policy, built to secure a recycled fibre supply chain it can control.
The fashion industry is overwhelmingly contract-manufactured. Brands design products and place orders. Factories produce them. The relationship between a brand and its tier 1 supplier, the factory that cuts and sews the finished garment, is usually well documented. Below that, visibility drops sharply. The mill that wove the fabric is sometimes known. The spinner that produced the yarn rarely is. The farm or petrochemical facility that supplied the raw fibre almost never.
A garment label that reads "65% polyester, 35% cotton" may be broadly correct for the main fabric, but it says nothing about the lining, the thread, the elastic, the buttons, or the coatings. It does not tell a recycler whether the product can be processed. And it cannot be verified against actual production data because, in most cases, that data was never collected.
The tier 2 and tier 3 suppliers who hold the actual material composition data have limited incentive to share it. Opacity in the supply chain protects margins and reduces scrutiny. The system was not built for transparency. It was built for speed and cost.
This is not a criticism. It is a description of how the industry evolved. Material knowledge was not commercially necessary at the brand level. Until now.
Here is where the conversation needs to shift.
The debate is usually framed around individual brand compliance: does this brand know what its products are made of? Can it produce the data a regulation demands? Will it be fined if it cannot? Those are the right questions, but they miss the systemic consequence.
A recycling facility operating at industrial scale needs feedstock at a consistent, verifiable purity. A cotton recycler typically requires a maximum of 5% non-cotton contamination. Syre needs polyester. Infinited Fiber needs a minimum of 88% cotton. These thresholds are not guidelines. They are operating parameters. Fall below them and the batch fails, or worse, the output is defective and the entire production run is compromised.
Now consider what happens in a communal collection stream. An H&M garment, traced from fibre to shelf with composition verified and material identified, enters a collection bin alongside a garment from a brand that has never tracked its material composition. The sorting facility receives both. Without product-level identification, they look the same. The NIR scanner reads fibre content at surface level but cannot verify provenance, cannot confirm what is beneath the face fabric, and struggles with black textiles and multi-layer constructions.
The unverified garment enters the sorted batch. If its actual composition differs from what the label claims, or from what the surface scan suggests, it contaminates the batch. The recycler's feedstock purity drops. The brand that invested in traceability pays the price for other brands investing nothing.
This is the free rider problem. The non-compliant brand does not just accept a risk for itself. It exports that risk to every brand whose material enters the same stream. The compliant subsidise the non-compliant. The system is only as reliable as its weakest input.
All cotton t-shirts are equal. But some are more equal than others. Two identical products, same fibre, same weight, same composition, sit side by side. One has traceability data behind it. One does not. A recycler can trust one. The other is a liability. They may be equal in material. They are not equal in value to the system.
The brand that invested in traceability pays the price for other brands investing nothing. The non-compliant do not just accept a risk for themselves. They export it to everyone whose material enters the same stream.
The industry conversation around textile recycling is dominated by volume. How many tonnes can this facility process? How many collection points does this country have? How much material is placed on the market each year?
France provides the clearest numbers. In 2024, 891,000 tonnes of textiles were placed on the French market: 3.5 billion items, roughly 10 million per day. Of that, 289,000 tonnes were collected. The remaining 602,000 tonnes are sitting in wardrobes, going to general waste, or being exported informally. France is targeting a 60% collection rate by 2028, up from roughly 32% today. At the European level, 132 million tonnes of fibre were produced globally in 2024, with recycled fibre accounting for just 7.6% of the total.
These numbers suggest a volume problem. But the constraint is not volume. The constraint is verified quality.
The textile recycling ventures that have struggled did not fail because they lacked volume. They failed because they could not secure enough of the right volume: material that met purity specifications, at a price that made the economics work, with offtake commitments from brands willing to purchase the output.
The same risk now faces the next generation of recycling investments. Syre is planning for 3 million tonnes of capacity by 2032. Circ has secured €450 million for a plant in France. Infinited Fiber is building a 30,000-tonne facility in Finland. RE&UP is targeting 200,000 tonnes in Turkey. Billions of euros are flowing into downstream capacity.
But the question is not whether these plants can be built. It is whether the feedstock flowing into them will be good enough to operate. And "good enough" now means two things: the right material composition, and verification that it is what it claims to be. 95% cotton purity was always the standard. Now it needs to be 95% verified as well. Both matter.
The industry is building bigger pipes. The question is what is flowing through them.
And here is the part that gets lost in the volume conversation. France's 602,000 tonnes of uncollected textiles, from a single country, could, on volume alone, fill Syre's planned Vietnam facility two to four times over, even before accounting for what share is polyester. Scale that across Germany, the UK, the Netherlands, Italy, and Spain, and the total dwarfs what any recycling plant could process. The volume is not missing. It is in sock drawers and on washing lines and folded in the back of wardrobes. The recycling and brand executives looking for feedstock need to realise that the volume they need is already out there. It is not a supply problem. It is a retrieval problem. And the answer to that is not building more sorting plants. It is getting out of our own way.
From September 2026, the EU's Empowering Consumers for the Green Transition directive begins enforcement. The minimum fine for non-compliance with environmental claims requirements: 4% of annual turnover, or €2 million, whichever is higher. Some member states are already setting the ceiling above this. Additional penalties include confiscation of revenues linked to misleading claims and exclusion from public procurement for up to 12 months.
For context: DWS, Deutsche Bank's asset management arm, was fined €25 million in Germany and $19 million by the US SEC for claiming ESG was "in its DNA" without being able to demonstrate it operationally. The investigation forced a change in CEO, and the fine capped a multi-year overhaul of internal documentation and control processes. In contrast, Shein was fined €41 million across Italy and France, but against estimated annual revenue exceeding $24 billion, that penalty barely registers. The incoming framework is designed to close precisely that gap: to make the fine proportional to the company's size, regardless of how large it is.
To understand the scale of what is at stake: adidas reported €23.7 billion in revenue in 2024. H&M Group reported approximately €21 billion. Nike reported $51.4 billion. Inditex reported €38.6 billion. The fine applies to the "trader" making the claim, which could mean the group or a subsidiary depending on corporate structure, but at the scale of these brands, 4% runs into hundreds of millions or more. These are numbers that change behaviour.
These are not symbolic numbers. And the rationale behind them is worth considering. A fine at this scale is not designed to punish a brand for overstating a marketing claim. It is, in effect, designed to protect the system. If one non-compliant brand's unverified product contaminates feedstock batches across multiple recycling facilities, the cost cascades through the entire value chain, from the sorting plant to the recycler to every brand that relies on that recycler's output. Whether intentional or not, the fine structure makes non-compliance more expensive than the damage it causes downstream. This is infrastructure protection, not brand punishment.
At the scale of adidas, H&M, or Nike, a 4% fine runs into hundreds of millions or more. The fine is not designed to punish a marketing claim. It is designed to protect the infrastructure that the compliant brands are paying to build.
Syre needs polyester feedstock. H&M co-founded Syre and signed a $600 million offtake agreement. H&M operates over 4,000 stores globally and already runs garment collection in 26 countries. The feedstock source and the collection infrastructure are controlled by the same company.
adidas is participating in the Closing the Footwear Loop project, searching for post-consumer sorting solutions, while operating thousands of its own retail stores where customers walk in wearing running shoes whose brand and composition are already known. Patagonia runs Worn Wear as a repair and resale channel, but the same infrastructure could identify and route end-of-life products to matched recyclers.
The recycling facilities being built today are searching outward for feedstock: negotiating with municipalities, partnering with waste processors, building automated sorting capacity. Meanwhile, the most verified, highest-quality collection channel available is the brands' own retail network. The offtake agreement and the take-back programme are two halves of the same loop. They are just not connected yet.
The polyester that Syre's plant needs next year is being worn right now by the customer walking into an H&M store to buy a new pair of jeans. The cotton that the next generation of cellulosic recyclers needs is folded in a drawer two kilometres from an adidas flagship. The feedstock is not in a landfill. It has not been exported. It is living with the customer, and the customer is already walking through the door.
The disconnected loop
Brands have mapped the top half. They have invested in the bottom half. The customer sits in between.
Raw materials Manufacturing Distribution Retail / Sale Customer Wardrobe. Washing line. Sock drawer. Collection point Retail store = verification Sorting Verified = simpler, cheaper Recycling plant Syre, Circ, Infinited Fiber Billions invested MAPPED BY BRANDS INVESTED BY BRANDS THE GAP No control. The top half is mapped. The bottom half is funded. The customer connects them. The offtake agreement and the take-back programme are two halves of the same loop. They are just not connected yet.The distinction the industry needs to make is between trusted and untrusted material loops.
A trusted loop: a product collected at a brand's own retail point, identified at the counter without the need for industrial scanning equipment, composition confirmed by the brand's own supply chain data, and routed to a matched recycler. Every step is known. Every input is verified. The recycler can operate with confidence.
An untrusted loop: an anonymous product, collected in a communal bin, sorted by NIR at surface level, composition estimated rather than confirmed. The product may contain exactly the right material. But it cannot be verified. For a recycler operating at scale, that uncertainty is a risk, not a feedstock.
Both loops are necessary. Communal collection captures volume that retail take-back cannot reach. But communal collection cannot provide the verification that industrial recycling increasingly requires. The two serve different functions. Retail take-back provides the assured base: the stable, long-term input that covers a recycling plant's fixed operating requirements. Communal collection supplements it with additional volume. The mistake is treating them as the same channel when they solve different problems.
Any business building a factory wants to cover its fixed costs with reliable, contracted inputs before going after the remainder through other channels. That is not a circular economy principle. It is a business principle. And it applies to recycling plants the same way it applies to any other manufacturing operation.
Trusted loop vs untrusted loop
The same material can follow two very different paths. Only one gives a recycler confidence.
Untrusted loop Anonymous product Brand unknown. Composition unverified. Communal collection bin Mixed with products from every brand. NIR surface scan Composition estimated, not confirmed. Sorted batch Purity unknown. Contamination risk present. Recycler receives unverified feedstock Batch risk. Output quality uncertain. Uncertainty Trusted loop Identified product Brand known. Product category recognised. Retail collection point Collected at the store. No industrial hardware. Brand data confirms composition Supply chain traceability verifies material. Routed to matched recycler Right material to the right destination. Recycler receives verified feedstock Purity confirmed. Output quality assured. Confidence Both loops may contain exactly the same material. Only one can be verified. The mistake is treating them as the same channel when they solve different problems.Retail take-back provides the assured base: the stable, long-term input that covers a recycling plant's fixed operating requirements. Communal collection supplements it. The mistake is treating them as the same channel when they solve different problems.
This is the question the industry has not yet answered, and this article does not claim to answer it definitively.
The Digital Product Passport is the intended verification mechanism. By mid-2028, every textile product placed on the EU market will be required to carry one. The data will include fibre composition, chemical treatments, recyclability guidance, and environmental footprint. Products without a compliant DPP will be denied CE marking and blocked from entering the EU market.
But the DPP only applies to products manufactured after enforcement. Every product sold before mid-2028, and every product currently sitting in a wardrobe across Europe, has no digital identity. Those products will be reaching end of life for the next five, ten, fifteen years. The gap between the regulation and the physical reality will persist for at least a decade.
Possible mechanisms for managing that gap include market access controls at customs, retailer liability for the products they stock, and EPR fee modulation, where brands that can verify their material composition pay lower levies than brands that cannot. Each of these shifts the cost of non-verification back to the source.
But for products that exist today, the most practical verification checkpoint is the retail store. Not because it requires expensive hardware in every stockroom. It does not. A product can be identified at the counter by brand and category without an NIR scanner or an automated sorting line. The brand is visible. The product type is recognisable. The brand's own data confirms the composition. That is enough to route it to the right destination, and it scales across any retail network without capital equipment per store.
The volumes are out there. They are in wardrobes. France alone has 602,000 tonnes of uncollected textiles every year, and that is one country. It is not a question of whether the feedstock exists. It is a question of where the customer takes it, and whether the system is ready when they arrive.
The packaging industry spent decades learning that contamination kills recycling at scale. Not all plastics are the same. Not all paper grades are the same. The textile industry is learning the same lesson, but faster, under more regulatory pressure, and with billions of euros of recycling infrastructure depending on the answer. The brands that know what their products are made of are not just protecting themselves from fines. They are building the trusted feedstock loops that the next generation of recycling plants needs to survive. The brands that do not know are not just risking non-compliance. They are contaminating the system for everyone else. And the mechanism that connects the data to the product, the brand to the recycler, and the investment to the return? It is the same place the product was first sold. The customer is already there. The feedstock is already there. The question is whether the industry can get out of its own way long enough to connect them.
Sources: Paper recycling contamination tolerances: CEPI EN 643 European List of Standard Grades (max 1.5% non-paper components for most grades). Plastics resin identification and sorting: European Plastics Recyclers (PRE); Plastics Europe. Textile recycler feedstock specifications: cotton recyclers (max 5% non-cotton); Syre (polyester); Infinited Fiber (min 88% cotton, per infinitedfiber.com/faq); Circ (polycotton separation). France collection data: Refashion 2024 (pro.refashion.fr) - 891,309 tonnes placed on market; 289,393 tonnes collected; 4.2 kg per inhabitant; target 60% collection by 2028. Global fibre production: Textile Exchange Materials Market Report 2025 - 132 million tonnes in 2024; recycled fibre 7.6% of total; polyester 59% of global output. DPP timeline and requirements: European Commission ESPR Working Plan 2025-2030; Intertek (May 2025); Carbonfact (Jan 2026); European Parliament study on DPP for textiles (2024); CIRPASS consortium and JRC preparatory studies. DPP phased rollout: Phase 1 (2027), Phase 2 (2030), Phase 3 (2033). ECGT directive and fines: Latham & Watkins (Dec 2025); Lewis Silkin (Feb 2026); K&L Gates (Aug 2024). Minimum fine: 4% annual turnover per EU Modernisation Directive. DWS fine: Bloomberg, ESG Dive, D&O Diary (Apr 2025) - €25M Germany; SEC (Sep 2023) - $19M US. Shein fine: Fashion Dive, Euronews (Aug 2025) - €1M Italy + €40M France; Shein revenue estimated (not officially disclosed). Recycling investment: Syre 3Mt by 2032, $600M H&M offtake, $100M Series A (H&M Group press release Mar 2024; ESG News May 2025); Circ €450M (Textile World, Bloomberg, May 2025); Infinited Fiber 30,000t (infinitedfiber.com); RE&UP 200,000t (H&M Foundation, Apr 2025). H&M traceability: TextileGenesis; H&M Group sustainability pages (hmgroup.com); Sourcing Journal (May 2022); 200M+ garments traced by end 2022; 94% recycled polyester in 2024 (Trellis/GreenBiz, Apr 2025). adidas traceability: TrusTrace (trustrace.com); adidas Annual Report 2024 (report.adidas-group.com); 99% recycled polyester by 2023. Patagonia: patagonia.com/our-footprint; Better Trail (Oct 2025); 92% recycled polyester by weight. Revenue figures: adidas FY2024 €23.7B (adidas-group.com); H&M Group FY2024 SEK 234.5B / ~€21B (hmgroup.com); Nike FY2024 $51.4B (investors.nike.com); Inditex FY2024 €38.6B (inditex.com). EU textile industry: European Commission EU Strategy for Sustainable and Circular Textiles (2022) - more than 160,000 companies.