The Impact Thesis: a four-stakeholder, framework-aligned account of in-store collection at scale.

In 2016 the EU separately collected 10.7% of textile waste. By 2022 the figure was 15.0%. The improvement rate is 0.72 percentage points per year. Projected forward at the same pace, the line crosses the EU 2030 ambition of 50% in approximately 2071 and the 80% upside in 2113. The gap between trajectory and ambition is forty-one years for the central target. Directive (EU) 2025/1892 will accelerate the volume curve, but the yield curve will not bend by itself. Of the 15% that does reach a collection point, communal collection recovers ~36% in sort yield. Aggregate fate of post-consumer textile waste in Europe today: ~94% incinerated or landfilled. The thesis: the binding constraint is no longer infrastructure — it is the engagement-and-incentivisation layer at the counter.

The European take-back system has three layers running in parallel. The compliance layer (producer-responsibility schemes) manages eco-modulated fees and aggregate reporting. The operations layer (recyclers) handles physical sortation and material recovery. The data layer — Utilitarian — runs at the counter, before the bag is ever weighed. The platform sits at the join of EU CatSys Category 3 (Traceability) and Category 4 (Enabling Technologies). It does not displace either of the other two layers; it produces the per-item record set both have been waiting for. Hardware cost: the customer brings the phone. The system runs on a poster, a QR code and a CRM hook.

The recycling chain is denominated in kilograms. The customer-value line is priced in retail and absent in recycling. Per the existing recycler P&L, customer engagement is not a revenue line. MRF capex sits in the system as the cost of upstream information absence — sorting is a substitute for knowing what is in the bag. The structural shift the platform enables is from price-per-kilo unit economics to per-item-value-per-unit-of-material economics: an identified item routes higher, sorts faster, and loses less mass than mixed-bag input. The unit changes; the consequences land across the whole chain.

One scan produces value at four nodes simultaneously. The customer gets a discount and a reason to come back. The retailer gets a verified email at €1–3, with brand, model, store, timestamp and replacement-consideration window attached. The brand gets per-SKU end-of-life data — required for eco-modulated EPR and disclosure-grade circular reporting. The recycler gets identified, traceable feedstock that arrives pre-sorted at the gate. None of the four numbers is fictional. The programme survives a budget cycle precisely because it produces value for all four — a programme that works for one party alone gets cut on the next review.

The same scan event monetises across five streams: subscription fees from retailers, margin share through certified partners, eco-modulation services for brands, data-services revenue from recyclers, and platform fees from producer-responsibility schemes. The same record set also attributes impact across five categories: waste diverted, verified consumer engagement, per-product end-of-life routing, linear-resource displacement and GHG avoided. The discipline is reporting the same record set through several frameworks — not inventing new metrics for each audience.

Cohort structure (which 5–10% return, and whether they return twice). In-flight versus batch behaviour (single item vs full bag). Store-level variance in scan rate (the variance between similar stores is the research question). Incentive design and the reward S-curve. Feedstock quality at point of capture. Recycler economics at industrial scale. Net impact outcome (kg diverted × displacement factor × emission factor — each with a range, each reported as a band). The retailer-engagement flywheel: email captured → customer returns → second item arrives → cohort widens.

Tier 1 — Strong verifiability today (direct operational measurement): scans, verified emails, items captured by SKU, store-level engagement, recycler chain-of-custody confirmations. Tier 2 — Partial verifiability today (depends on displacement factor or sector conversion rate): virgin material displaced, GHG avoided, water and land impact. Tier 1 is the headline. Tier 2 is reported as a range, never as a single confident number. The discipline is the omission, not the inclusion.

Displacement factor at retail scale — band 0.4–0.9 depending on fraction quality and downstream route; the headline is the band, not the midpoint. Cohort-level attribution — of the customers who returned an item, how many were attributable to the counter? Cleanly measurable with a holdout, not yet measured at scale. Store-level operational sensitivity — two stores, same chain, same town: one runs 18 scans a day, the other 6. The variables are known; their ranking is not. Drip-vs-batch sort-yield differential — continuous identified flow versus periodic mixed-bag intake. Recycler yield differs; magnitude is an active question.

Approximately 173,500 stores across the EU carry soft-goods or footwear. At pilot-observed scan-per-store density, stratified by cohort, this produces ~468M scans per year and ~422M verified emails. Recurring relationships at twelve-month return: ~114M. Material diverted: ~390k tonnes (range 280–510k). Virgin material displaced at 0.4 displacement factor: ~156k tonnes (upper bound 0.9 → 351k). GHG avoided central case: 2.73M tCO₂e (range 1.95M–3.90M). These are scenario numbers, not projections — what the curve could look like if the engagement layer is solved at estate scale. They are not a commitment, a forecast, or a benchmark for any single retailer.

The indicators land cleanly in EU CatSys (Cat 3, Cat 4), the GCP COP30 disclosure crosswalk, IMP 5+1, GIIN Core Metrics, IRIS+, Planetary Boundaries (textiles), SDGs 12, 13 and 15, ESRS E5 (Resource use & circular economy), GHG Scope 3 Category 12 (End-of-life treatment of sold products), CTI v4.0, EU Taxonomy circular-economy criteria and SFDR PAI indicators. Disclosure architecture is a hairball; reporting the same number through several frameworks is more useful than inventing our own.

Shipping today: counter platform across Netherlands stores, CRM exports/API, recycler chain-of-custody confirmation, ESRS E5 crosswalk, AI item-recognition at 92% accuracy. On the roadmap: per-SKU eco-modulation reporting suite, multi-language support across DE/FR/UK markets, fire-safe routing module for batteries, and a published LCA suite for the routes where third-party evidence is currently partial. The line between the two is published, not implied.

Industry-side reading failure on the regulator's choice (treating the directive as a logistics mandate, not an engagement mandate). Recycler-side reading failure on the unit (treating per-kilo as the only price-able unit). Retailer-side reading failure on the channel (treating sustainability as a cost centre rather than the cheapest verified-email channel available to physical retail). The three are linked: each propagates the failure of the others. The platform addresses all three in one move, which is the reason it works.