Skip to content
GREEN FUTURETECH.
Explore

Research libraryCotton to Garment Across Africa

Chapter 12 of 14 · Part III · Cross-Cutting Determinants of Competitiveness Across African Corridors · Draft web edition

European digital product passports and supply-chain due diligence turn farm-to-garment traceability from a marketing story into a market-access gate

European product, labour and due-diligence rules reward suppliers that can prove each material transformation. CmiA offers a useful origin baseline, but an African garment becomes inspectable only when bale, yarn, fabric, production and claims records reconcile.

Sections
7
Exhibits
6
Minutes
27
Sources
12
EX 53 · 2,028: Forced-labour rules. Apply from 14 December 2027

The cloth is woven from this chapter's 6 exhibits: each weft is one exhibit's own data, normalised; each warp group is a section. The probe reads the exhibit's own value where it crosses. Procedural visualisation, not a measurement.

The chapter as a draft: 7 sections, 2 exhibit wefts, one shuttle.

Lead exhibit · 53TimelinesourcedTextile evidence duties arrive through separate legal instruments, not one 2027 deadline
2,025ESPR in force2,025First working plan2,027Unsold-goods ban2,028Textile act planned2,028Forced-labour rules2,028Textile EPR schemes2,030CSDDD application

Read-out2,025: ESPR in force. 18 July 2024

Method and source

The Q4 2027 textile delegated act is planned and indicative, not a fixed garment application date. CSDDD application shown for 26 July 2029 follows the amended timetable. EPR schemes are due by 17 April 2028. Exact textile DPP fields and transition remain subject to delegated and implementing acts.

Unit: calendar year, month fractions used only for spacing. Status: sourced. Sources: S1201, S1202, S1203, S1204, S1205, S1206.

The chapter's warpBar: share of the reading · square: an exhibit

Section 12.0 · Opening · 1 min · no exhibit of its own · one movement Go to section ↓

Overture · 1 min

Farm-to-garment traceability becomes commercially valuable when it can answer a buyer's evidence request without reconstruction. The answer is not a larger sustainability narrative. It is a controlled chain of lot identifiers, physical transformations, source documents and measured outcomes. African cotton can then support product compliance, buyer due diligence and precise origin claims from the same record. A certificate at the farm gate helps, but it cannot carry that burden alone.

¶ 01

The regulatory conclusion needs equal precision. The European Union has established the Digital Product Passport framework and prioritised textiles, but it has not set a blanket requirement for every garment to carry a passport from the start of 2027. The Commission currently plans the textile delegated act for the fourth quarter of 2027, describes that timing as indicative and leaves the exact data requirements and subsequent implementation measures to that act and its supporting specifications.[S1203] GFT therefore recommends building the evidence spine now while treating the final payload, granularity and application date as controlled regulatory variables.

¶ 02 · 1 source

S1203

12.12 min · 7 paragraphs · 1 exhibit

Delegated acts, not a blanket 2027 slogan, define the European gate

Related · chapter 037 paragraphs step it

Related exhibit 11 · chapter 03FlowillustrativeA controlled provenance chain preserves identity from contracted farm to garment record
Contracted farm unitCollection lotGin lot and bale IDSegregated yarn lotFabric roll IDCutting markerGarment bundle IDProduct data record

Read-outContracted farm unit. Starting point. Leads to Collection lot (Register).

Method and source

Illustrative GFT architecture. AbTF supports the HIP chain, now called Segregation, from bale through textile product; the garment data carrier and factory exception controls are proposed extensions. This is not a claim of implementation at a GFT operating factory.

Unit: GFT reference control sequence. Status: illustrative. Sources: S0301, S0305, S10, S34. Illustrative figures are teaching examples, not measured factory results.

Read this exhibit in its chapter ↗
12.1 · Introduction · paragraph 1 of 7

Regulation (EU) 2024/1781 establishes the Ecodesign for Sustainable Products Regulation. It creates product-specific performance and information requirements through delegated acts rather than imposing one finished textile rule immediately. Article 9 makes a passport necessary only in accordance with the applicable delegated acts. It also requires passport data to be accurate, complete and current.[S1201] This distinction matters for an investment committee. The legal direction is clear, but the final textile data dictionary is still being prepared.

The first working plan adopted on 16 April 2025 places textiles and apparel among the priority final products and gives 2027 as the indicative adoption year. It also states that product-specific delegated acts will specify the information collected and made available in the passport.[S1202] The Commission's textile page narrows the current planning point to the fourth quarter of 2027 and repeats that the timeline may evolve.[S1203] Adoption is not the same as application. The framework regulation normally gives economic operators at least eighteen months from a delegated act's entry into force before its requirements apply, except in specified justified cases.[S1201] A prudent programme therefore plans against enacted transition text, not against a trade headline.

The passport is nevertheless a market-access instrument rather than a voluntary product page. Once the relevant acts and systems apply, manufacturers and importers have defined responsibilities, and customs can check the unique registration identifier against registry data before releasing a covered product for free circulation. The regulation also requires a persistent product identifier linked through a data carrier, with open, interoperable and machine-readable data as appropriate.[S1201] A QR code is only the carrier. The legal and operational asset is the maintained record behind it.

Other European duties have separate triggers and should not be collapsed into the passport. Since 19 July 2026, the prohibition on destroying unsold apparel, clothing accessories and footwear has applied to large enterprises, with medium-sized enterprises following in 2030 and micro and small enterprises outside that prohibition.[S1201] The forced-labour regulation applies from 14 December 2027 and covers products made with forced labour at any production, harvest or processing stage, including imported products and online offers targeting EU end users.[S1205] These instruments make different evidence requests. A passport cannot certify away a labour investigation, and a social audit cannot substitute for missing product data.

Corporate due diligence has its own scope. Following the 2026 amendments, the principal thresholds rise to 5,000 employees and EUR 1.5 billion of turnover, rather than placing every small African supplier directly inside the Corporate Sustainability Due Diligence Directive.[S1204] Smaller suppliers can still receive requests through their chains of activities. Their response should be proportionate, contractually clear and capable of showing actual controls. The OECD sector guidance supports that wider risk process, but it is voluntary and does not itself create legal obligations.[S1210]

The chronology in Exhibit 53 separates enacted dates from planned work. This is the compliance calendar GFT would put under change control.

Two implications follow. First, regulatory readiness must be versioned by instrument, product, operator role and date. Second, the underlying evidence should be useful before any passport obligation starts. The same lot record improves receiving inspection, supplier claims, production reconciliation and customer service even if the European timetable moves.

12.22 min · 6 paragraphs · 1 exhibit

CmiA supplies a verified origin baseline, not an automatic compliance certificate

Exhibit 54Line boardsourcedCmiA provides scale, while the physically traceable HIP network has a narrower reported boundary

Read-out · sourcedAfrican output verified: 36.4%. CmiA and CmiA Organic, 2025

Method and source

The 36.4 percent and farmer figures describe CmiA and CmiA Organic in 2025. The 700 partners in 25 countries describe the physically traceable network reported in February 2026. These are different denominators and must not be combined into one coverage ratio.

Unit: reported share, farmers, partners and countries. Status: sourced. Sources: S08, S10.

Cotton made in Africa gives an African supplier a documented starting point. The Aid by Trade Foundation reports that, in 2025, about 715,000 small-scale farmers in ten African countries produced under CmiA or CmiA Organic, representing 36.4 percent of African cotton production. It separately reports more than 2,800 partners processing cotton under its various standards across 50 textile production markets.[S08] These figures describe the standard owner's ecosystem. They do not show that every African bale, every Cameroonian bale or every garment sold with a sustainability reference is verified.

¶ 01 · 1 source

It separately reports more than 2,800 partners processing cotton under its various standards across…S08

Related · Exhibit 25 · chapter 06

S08

The physical traceability network is narrower and should not be confused with the wider processing network. In February 2026, AbTF reported around 700 suppliers and producers in 25 countries able to trace CmiA cotton from product back to origin, more than double the previous year's figure.[S10] The dated figures in Exhibit 54 retain this distinction between total production verification and the physically traceable partner network.

¶ 02 · 1 source

AbTF reported around 700 suppliers and producers in 25 countries able to trace CmiA…S10

Exhibit 54

S10

CmiA's value is specific. Its standard covers cotton cultivation and ginning, while its chain-of-custody models provide different ways to carry a claim through later organisations. The standard owner states that CmiA cotton is rainfed, excludes genetically modified seed and uses independent third parties within its assurance system.[S09] That is useful evidence for origin and farm practice. It is not proof of chemical compliance at a dyehouse, labour conditions on every garment line, or the footprint of a finished shirt.

¶ 03 · 1 source

S09

A buyer must therefore qualify the exact organisation, site, material and claim. AbTF's membership guidelines place a duty on the purchasing entity to check that the selling entity is registered, and they require organisations producing, treating, trading or sourcing CmiA-claimed products to maintain the relevant membership.[S1208] A certificate screenshot without scope, validity and transaction linkage is weak procurement evidence.

¶ 04 · 1 source

S1208

Note

The practical distinction is between an eligible input and an eligible product story. Cameroon can offer the first through its cotton organisation and CmiA participation. The second must be manufactured through controlled procurement and production. GFT's role is to specify that control system, not to claim that a future factory or supplier has already attained it.

¶ 05

12.34 min · 1 sub-steps · 1 exhibit

Physical segregation and volume reconciliation answer different questions

Exhibit 5511 paragraphs step it

Exhibit 55RegistersourcedHIP supports physical-origin claims, while Mass Balance verifies volumes rather than garment content
Control question
Physical content
Hard Identity Preserved
Cotton share must be CmiA from nominated traders
Mass Balance
Verified cotton may be blended or substituted

Read-outControl question: Physical content. Hard Identity Preserved: Cotton share must be CmiA from nominated traders. Mass Balance: Verified cotton may be blended or substituted

Open the full register (5 rows)
Physical contentCotton share must be CmiA from nominated tradersVerified cotton may be blended or substituted
SegregationRequired at every supply-chain stageNot required after registered spinning purchase
Core evidenceLot identity, segregation and transaction recordsPurchase and claimed-volume reconciliation
Assurance emphasisSite controls, documentation and physical traceabilityTracking-system balance and transaction validity
Strong single-origin statementPossible when all lot links and claim rules holdNot supported as a physical content claim
Method and source

Interpretation follows CmiA Membership Guidelines version 1, October 2025, and the AbTF Transparency Standard description reported in February 2026. Buyer and legal teams should confirm current scheme rules before approving a claim.

Unit: custody control comparison. Status: sourced. Sources: S1208, S10.

12.3 · Introduction · paragraph 1 of 11

The two CmiA chain-of-custody models support different statements. Under Hard Identity Preserved, the cotton share of a claimed yarn, fabric or garment must contain only CmiA cotton from nominated traders, and claimed materials must be segregated through the supply chain. Under Mass Balance, verified cotton may be blended or substituted, while the tracking system reconciles purchased and claimed volumes.[S1208] Mass Balance is a volume-control mechanism, not proof that verified fibre is physically present in a particular garment.

This is not a reason to dismiss Mass Balance. It can support demand for verified cotton where physical segregation would be costly or impractical. It is a reason to control language. A Mass Balance claim may support a statement about verified purchasing and volume reconciliation within the scheme's rules. It cannot support the stronger statement that a particular shirt contains identifiable Cameroonian cotton. HIP is the appropriate CmiA route when that stronger physical-origin statement is required.

HIP also demands operating discipline rather than a software label. The guidelines permit physical or temporal segregation, require clear marking and call for bale or lot identifiers on accompanying production, sales and transport records. Site-level assurance reviews documentation, records and physical controls.[S1208] Temporal segregation is especially relevant to a small factory. A dedicated line is not always necessary, but a documented production run, line clearance, status label and reconciliation are necessary.

The comparison in Exhibit 55 converts those requirements into procurement and communication decisions.

A garment factory must preserve identity through many-to-one and one-to-many transformations. Several bales can feed a yarn lot. Several yarn lots can feed a fabric batch. One fabric roll can serve several cutting markers, and one marker can create bundles for several sewing lines. The ledger therefore needs parent and child relationships, not a single copied certificate number. Each event should record input identifiers, output identifiers, site, timestamp, quantity, unit, operator role and source document.

The same structure must handle losses and exceptions. Cutting offcuts, rejected panels, rework, returned garments and subcontracted operations all change either quantity or custody. A manual adjustment should require a reason code and a second role's approval. A missing predecessor should block a physical-origin claim until resolved, even when the commercial invoice is valid. This is where a traceability system becomes an operating control rather than a reporting layer.

Physical markers can strengthen that control, but they do not replace it. A tracer attached to fibre may help test whether material is present, while a transaction ledger explains who transformed it and under which claim. Conversely, a digital token without a controlled material stream can be copied far more easily than the physical lot it purports to describe. GFT's design principle is to make independent evidence types agree: material observation, document, quantity and responsible actor.

12.4 Supply contracts must make the evidence deliverable enforceable

The purchase contract should define the evidence schedule as carefully as the fabric specification. Each order should identify the chain-of-custody model, expected registration scope, required transaction documents, lot-level quantity data and any restrictions on substitution or subcontracting. The supplier should know before production which records must arrive with the material and which discrepancy will place the receipt on hold. AbTF's own guidance makes registration and claim-specific documentation an operating requirement, so these fields should be procurement obligations rather than optional requests after shipment.[S1208]

Change control needs a named trigger. A different dyehouse, a new subcontracted sewing site or a replacement yarn lot can break the approved custody route even when the fabric looks identical. The supplier should notify that change before execution and supply the new site's evidence for review. The receiving team should then compare the actual delivery note, lot labels and transaction record against the approved route. A commercial acceptance of substitute material should not automatically become approval for the same origin claim.

Remedies should distinguish commercial and communications decisions. A buyer may accept a late document as an administrative delay, reject the material, or accept the material while withdrawing the physical-origin statement. Those are different actions with different consequences. The contract should preserve the right to correct the public record if later evidence contradicts an earlier claim, including instructions for labels, product pages and customer notifications.

Finally, retention and access must survive staff changes. The factory should be able to retrieve the complete evidence pack for a historical lot after the merchandiser has left and the supplier has changed account managers. Independent assurance works best when both parties know the sampling method, response window and acceptable source artefacts in advance. This turns due diligence from an annual questionnaire into a repeatable control, consistent with the OECD's emphasis on tracking effectiveness rather than treating assessment as a one-off exercise.[S1210]

Key figures from chapter 03 · Exhibit 10Key figuressourcedCmiA combines substantial farm-level coverage with a smaller physically traceable partner network
  • 36.4%African output verifiedCmiA or CmiA Organic, 2025
  • about 715,000Participating farmersSmall-scale farmers, 2025
  • 10Countries in annual report2025 reporting boundary
  • about 700Traceability partnersAcross 25 countries, February 2026

Read-outAfrican output verified: 36.4%. CmiA or CmiA Organic, 2025

Method and source

AbTF reports 2025 farm and production coverage in its annual report. The partner count is reported as of 19 February 2026 and concerns partners able to trace CmiA cotton, not all CmiA supply-chain participants.

Unit: reported status for 2025 and February 2026. Status: sourced. Sources: S08, S10.

Read this exhibit in its chapter ↗

12.52 min · 6 paragraphs · read with a related exhibit

Rain-fed provenance helps only when the footprint boundary stays visible

CmiA states that its cotton is entirely rainfed with zero groundwater extraction. It reports more than 2,100 litres of water saved per kilogram and up to 13 percent fewer greenhouse-gas emissions than conventional cotton, citing its 2021 lifecycle assessment for both comparisons.[S09] These are relevant fibre-level findings from the standard owner. They are not a finished-garment footprint, a guarantee for every lot or a comparison that a manufacturer can transfer automatically to its own product.

¶ 01 · 1 source

It reports more than 2,100 litres of water saved per kilogram and up to…S09

Related · Exhibit 42 · chapter 09

S09

The boundary explains the limitation. A farm and ginnery assessment does not include every later choice of spinning technology, dye recipe, thermal energy, factory electricity, cutting efficiency, packaging, transport, garment use or end-of-life treatment. A verified agricultural input can therefore sit inside a product with a very different total profile. The proper claim identifies which stages were assessed, in which year, with which method and against which comparator.

¶ 02

GFT recommends three tiers of environmental evidence. Tier A is direct measurement at the processing site, such as a meter reading tied to a production interval. Tier B is supplier-specific evidence whose method and allocation have been reviewed. Tier C is a published secondary factor used only when the better evidence is unavailable. The passport or buyer file should carry the tier, source, period and allocation rule beside each value. A lower tier is not necessarily useless, but presenting it as measured factory performance would be misleading.

¶ 03

Allocation deserves the same scrutiny as measurement. A factory submeter may serve several lines and several customer lots. The energy assigned to a lot can be based on recorded machine time, standard minutes or another stated driver, but the choice can change the result. The calculation should show the meter total, the allocation driver, the lot share and any unallocated remainder. Reconciliation to the meter total prevents a collection of plausible product footprints from exceeding actual consumption.

¶ 04

Wet processing needs particular care because the garment factory may not control it directly. The purchase specification should require the dyehouse or finisher to identify the batch, recipe family, water and energy method, effluent destination and relevant test reports. If those data are unavailable, the record should say so. An estimated dyehouse footprint with a named method is more credible than a false zero inserted to complete a passport field.

¶ 05

The resulting rule is simple. Preserve the agricultural advantage as a sourced farm attribute, then build the manufacturing footprint from measured or explicitly estimated downstream evidence. Never convert rainfed cultivation into an unbounded low-carbon or low-water garment slogan.

¶ 06

12.62 min · 7 paragraphs · read with a related exhibit

Evidence governance prevents provenance from becoming greenwashing

Related · chapter 10 · Exhibit 45Status lampssourcedTraining evidence becomes useful only when outcomes, enrolment and plans stay distinct

Read-out · sourcedProductivity after female supervisor training: +22%. Tufts assessment reported by Better Work; not a guaranteed transfer

Method and source

The 22 percent figure is a Tufts assessment reported by Better Work in September 2016. GDIZ reports sewing machines per operating centre without stating completions or productivity. PIA reports an ongoing recruitment campaign, not graduates. Gaaraadji reports a small retained group, not industrial productivity. These entries are not additive.

Unit: reported programme observations and plans. Status: sourced. Sources: S29, S04, S15, S31.

Read this exhibit in chapter 10 ↗

European consumer-protection rules now make the wording of a claim a controlled output. The relevant amendments have applied since 27 September 2026. The Commission states that environmental claims and sustainability labels in business-to-consumer communications must comply from that date, including for existing products.[S1207] Generic claims, future-performance claims and claims based on offsetting have different tests. The communications team therefore needs legal review and evidence at the point of publication, not a library of interchangeable green phrases.

¶ 01 · 1 source

S1207

A claim register should hold the exact wording, product and lot scope, lifecycle boundary, evidence references, calculation method, approval role, publication channel and review date. It should also hold the prohibited interpretation. For example, a team may be authorised to say that a lot uses physically traceable CmiA cotton, but not to imply that every product in the brand is traceable. A target to reduce future emissions should not be presented as an achieved reduction.

¶ 02

The distinction between CmiA and CmiA Organic is equally important. CmiA Organic requires both the CmiA criteria and certification under an approved organic standard, while the HIP model is mandatory for CmiA Organic.[S1208] A supplier cannot infer organic status from ordinary CmiA verification, rainfed cultivation or the absence of genetically modified seed. Each term has a separate evidence burden.

¶ 03 · 1 source

S1208

Traceability also does not complete due diligence. The OECD garment and footwear guidance describes an ongoing process of embedding responsible conduct, identifying harm, preventing or mitigating it, tracking effectiveness, communicating and supporting remediation. It expressly warns against a tick-the-box approach.[S1210] A lot ledger can show where material went. It cannot, by itself, show whether recruitment fees, excessive hours, harassment or unsafe work occurred at each site.

¶ 04 · 1 source

S1210

The operating response must therefore include worker and supplier channels. Complaints should have an owner, a protection rule, a remediation path and a closure test. Corrective actions should identify the cause and verify effectiveness rather than merely recording training attendance. Where a buyer identifies a severe risk, evidence requests should focus on that risk and the affected sites. Indiscriminate data collection can burden small suppliers while leaving the actual harm untouched.

¶ 05
Note

For GFT, the communications rule follows directly: state what has been measured, what has been verified by another party, what remains an estimate and what is still a target. This is especially important for a programme in formation. No future plant, certification, supplier audit or impact result should be described as already achieved.

¶ 06

12.75 min · 1 sub-steps · 2 exhibits

A lean event ledger can serve the passport without pretending to be an ERP

Step 1 of 5Decision and access

Exhibit 56 · illustrativeA five-layer evidence architecture keeps physical custody separate from publication and access

Decision and accessBuyer evidence roomPublic product viewAuthority and recycler viewsPassport servicePersistent identifierRegistry metadataBackup and role permissionsEvidence repositoryCertificates and scopeTests and declarationsAudits and claim approvalsTransformation ledgerParent and child lotsQuantity reconciliationQuality and utility eventsPhysical operationsBale to yarn and fabricMarker, bundle and garmentOffcut, rework and return
Decision andaccessBuyer evidence roomPublic product viewAuthority and recycler viewsPassportservicePersistent identifierRegistry metadataBackup and role permissionsEvidencerepositoryCertificates and scopeTests and declarationsAudits and claim approvalsTransformationledgerParent and child lotsQuantity reconciliationQuality and utility eventsPhysicaloperationsBale to yarn and fabricMarker, bundle and garmentOffcut, rework and return

Decision and access

Decision and access: Buyer evidence room, Public product view, Authority and recycler views · illustrative

A greenfield factory does not need to buy a monolithic enterprise system before it can preserve evidence.

Passport service

Passport service: Persistent identifier, Registry metadata, Backup and role permissions · illustrative

Evidence repository

Evidence repository: Certificates and scope, Tests and declarations, Audits and claim approvals · illustrative

Transformation ledger

Transformation ledger: Parent and child lots, Quantity reconciliation, Quality and utility events · illustrative

Physical operations

Physical operations: Bale to yarn and fabric, Marker, bundle and garment, Offcut, rework and return · illustrative

The method behind it: Worked example: a lot must reconcile before it can carry a claim.

Method and source · Exhibit 56

Illustrative GFT system design for a future African garment operation. It is not an implemented GFT result or a declaration of legal conformity. Final DPP payload, registry, access and retention rules must be mapped to the applicable delegated acts and technical specifications.

Unit: GFT proposed control layers. Status: illustrative. Sources: S1201, S1203, S1208. Illustrative figures are teaching examples, not measured factory results.

12.7 · the bodyExhibit 56 above; its figures are illustrative where marked.

A greenfield factory does not need to buy a monolithic enterprise system before it can preserve evidence. It needs a disciplined data model, controlled forms and clear ownership. A structured workbook and small relational database can suffice for an initial lot ledger when they enforce unique identifiers, validation, access rights, backups and change logs. The weakness is not the spreadsheet itself. The weakness is an untracked file that permits overwritten lot histories and unlabelled estimates.

GFT proposes the architecture in Exhibit 56. It separates physical custody, transformation events, source evidence, passport publication and user decisions. Each layer has a distinct control owner, so a communications editor cannot silently alter a receiving weight or an energy allocation.

The identifier register comes first. It defines the legal entity, facility, supplier, bale, yarn lot, fabric batch, cutting marker, bundle, production order, carton and product identifiers. The passport's eventual model, batch or item granularity must follow the delegated act, but the internal ledger can already preserve finer relationships.[S1201] Stable identifiers should never be recycled after a year-end reset or a change of commercial customer.

¶ 03 · 1 source

S1201

The event register comes second. Receiving, ginning, spinning, fabric formation, wet processing, cutting, sewing, finishing, packing, shipping and return events use the same minimum fields: event type, site, timestamp, input lot, output lot, quantity, unit, evidence reference and responsible role. Offline capture can be synchronised later, but the original device timestamp and subsequent correction history must remain visible. Connectivity is an implementation constraint, not a reason to invent continuity.

¶ 04

The evidence register comes third. It stores the actual certificate, scope, validity, test report, supplier declaration, transaction document and claim approval. A metadata field saying "certified" is not enough. The reviewer must be able to open the source artefact and see why it applies to the lot. Expired evidence remains in the historical record but cannot support a new claim without review.

¶ 05

The measurement register carries quality and resource data. Fabric inspection, first-pass quality, offcut weight, meter readings and allocation calculations link to the same production event. Where a value is calculated, the record stores its inputs and formula. Where it is estimated, it stores the basis and confidence limitation. This makes the future passport reproducible rather than merely readable.

¶ 06

Access is deliberately uneven. A consumer may need composition, care and origin information. A buyer may also need transaction documents and audit summaries. An authority may need conformity records, while a recycler needs material and disassembly information. The Commission's textile page says access will include online sales and a product data carrier, with exact information and technical specifications still to be defined.[S1203] Role-based views avoid publishing worker identities, sensitive costing or a supplier's confidential process data merely because transparency is required elsewhere.

¶ 07 · 1 source

S1203

Exhibit 57 shows why the arithmetic matters. A batch that cannot explain its quantities should not publish an unqualified physical-origin statement, even if every incoming document is genuine.

The passport record for that illustrative lot can now state the shipped quantity, fabric equivalent, offcut destination and reject disposition, with links to the cutting and finishing events. It still cannot prove farm origin unless the upstream identifiers resolve. It cannot state emissions without metered or estimated energy evidence. And it cannot declare conformity with a future textile delegated act whose final requirements are not yet published. Those limitations are features of an honest system.

¶ 09

Implementation should proceed through a thin pilot. Select one fabric supplier, one product family and one shipment route. Reconcile every lot from receipt to packed garment, test an offline shift, simulate a certificate expiry and ask an independent reviewer to reconstruct the batch. Only after that reconstruction succeeds should the factory scale the template across suppliers and product families.

¶ 10

12.8 Reconciliation tests should challenge the ledger before buyers do

12.8

A useful pilot ends with deliberate failure tests. The reviewer should attempt to register output greater than available input, reuse a child-lot identifier, receive material against an expired scope and mix two lots without recording the new parent. The system should reject or visibly flag each case. A successful test is not the absence of exceptions. It is proof that exceptions cannot pass silently into a published claim.

¶ 11

Quantity controls need both local and cumulative checks. At each event, inputs should reconcile to outputs plus recorded losses within an approved measurement tolerance. Across the order, the sum of child quantities should not exceed the parent quantity without an explained return or adjustment. The tolerance should reflect scale calibration and process knowledge, not a broad allowance chosen to make discrepancies disappear. Any manual balancing entry should display its author, reason and approval.

¶ 12

Temporal segregation requires a different test. The reviewer should inspect the prior and following production runs, clearance record, line status and residual material handling. A timestamp alone cannot prove that the previous fibre was removed. Photographs or checklists may support the clearance, but the responsible supervisor and the quantity remaining still matter. The objective is to show that the claimed production window was physically controlled.

¶ 13

The final test follows evidence into language. The reviewer selects the strongest statement planned for the product page and asks which lot records prove every part of it. If the wording covers a whole collection while the evidence covers one order, the claim is too broad. If the record supports rainfed cultivation but not factory emissions, the environmental statement is too broad. This last gate joins data assurance to editorial control and prevents a correct ledger from producing an incorrect advertisement.

¶ 14

12.92 min · 6 paragraphs · 1 exhibit

Disclosure should follow verified operating gates, not a promotional calendar

A twelve-month programme is useful only if each communication follows evidence. The calendar should never force a claim. GFT therefore recommends an initial year of gated releases: establish governance and the claim register, qualify suppliers, complete a traceability pilot, stabilise production records, commission an independent evidence review, and publish a measured impact report only when the records reconcile. This is an illustrative sequence for GFT's Cameroon textile programme, not a report of completed activity.

Exhibit 58 turns the outline's communications calendar into hold points. Each gate identifies what can be said and what must stop publication.

Exhibit 58RegisterillustrativeExternal disclosure advances only when the underlying evidence gate closes
Planning window
Months 1 to 2
Evidence gate
Owners, methods and claim register approved
Permitted external statement
System design and sourcing policy
Hold condition
Unassigned evidence owner

Read-outPlanning window: Months 1 to 2. Evidence gate: Owners, methods and claim register approved. Permitted external statement: System design and sourcing policy. Hold condition: Unassigned evidence owner

Open the full register (6 rows)
Months 1 to 2Owners, methods and claim register approvedSystem design and sourcing policyUnassigned evidence owner
Months 3 to 4Supplier scopes and transactions verifiedNamed input verification within scopeExpired or mismatched evidence
Months 5 to 6Pilot lot reconstructed independentlyTraceable pilot with stated boundaryBroken parent and child lot links
Months 7 to 9Production and resource records reconcileMeasured operating KPIs with methodsUnexplained quantity or meter variance
Months 10 to 11Evidence review and corrective actions closedVerified claims within reviewer scopeOpen material non-conformity
Month 12Annual ledger, grievances and remediation reconciledMeasured impact report with limitationsAny unresolved material exception
Method and source

Illustrative GFT governance sequence, not completed activity or a commitment to publish on a fixed date. The OECD guidance supplies the wider due-diligence principle. Months are planning windows; a failed gate pauses the associated external claim rather than the underlying corrective work.

Unit: illustrative months and release conditions. Status: illustrative. Sources: S1210. Illustrative figures are teaching examples, not measured factory results.

The measured impact report should be a reconciliation package rather than a glossy assertion. It should show opening and closing lot balances, supplier scopes, production quality, resource methods, waste destinations, workforce measures, complaints and remediation. Targets belong beside actuals with clear labels. Estimates belong beside measured values with their basis. Material exceptions belong in the report, not in an excluded annex.

¶ 01

The buyer evidence room can use the same structure at shorter intervals. A sourcing lead receives supplier scopes and transaction documents. A technical lead receives test reports and process specifications. A sustainability lead receives methods and calculations. This reuse reduces contradictory answers across teams and makes the public claim a controlled extract from the operational record.

¶ 02

The final control is independent challenge. A reviewer who did not build the ledger should select a shipment and trace it backwards to the upstream evidence, then select an input lot and trace it forwards to all affected shipments. The two directions test different failure modes: unsupported product claims and unallocated verified inputs. Findings should enter the same corrective-action process as production defects.

¶ 03

This chapter is industrial research, not legal advice. Product scope, operator responsibility, claim wording and transition dates should be confirmed against the applicable EU acts, Member State measures and buyer contracts before an export decision. The engineering task remains stable even when those rules change: preserve the physical lot, reconcile every transformation and publish only what the evidence can carry.

¶ 04

Check the record

The chapter keeps its sources within reach.

Every knot is a source this chapter cites; select one to read its record, or filter the register by grade. Publication dates and inherited-source limits stay explicit. Read the paper's full evidence summary.

12 sources · 36 citation marks on this pageMost cited: S1201, 6 times

Grade AGrade BGrade CKnot size: citations on this page

All 12 records shown

c12-S08 · Grade B

Aid by Trade Foundation (AbTF) / Cotton made in Africa

Aid by Trade Foundation Presents Annual Report for 2025

In 2025, around 715,000 small-scale farmers in ten African countries cultivated cotton in accordance with Cotton made in Africa or Cotton made in Africa Organic. As a result, 36.4 percent of all cotton produced in Africa is now verified under CmiA or CmiA Organic.

2026-08-20. Register access date: 2026-10-30. Source provenance is supplied by the paper evidence register.

Open the recorded source ↗
Chapter source notes

S1203 European Commission, Directorate-General for Internal Market, Industry, Entrepreneurship and SMEs. 2026-07-20.

S1201 European Union, EUR-Lex. 2024-06-28.

S1202 European Commission. 2025-04-16.

S1205 European Union, EUR-Lex. 2025-10-16.

S1204 European Union, EUR-Lex. 2026-02-26.

S1210 Organisation for Economic Co-operation and Development (OECD). 2018.

S1206 European Union, EUR-Lex. 2025-09-26.

S08 Aid by Trade Foundation (AbTF) / Cotton made in Africa. 2026-08-20.

S10 Cotton made in Africa (CmiA) / Aid by Trade Foundation. 2026-02-19.

S09 Cotton made in Africa (CmiA). 2026-09-28.

S1208 Aid by Trade Foundation (AbTF) / Cotton made in Africa. 2025-10.

S1207 European Commission. 2026-09-27.

Keep the thread · chapter 12 of 14

Chapter 13 picks up the thread.

Move the shuttle along the rack to preview any chapter: its sections, exhibits and reading time. Draft web edition. This research is not investment, legal or tax advice; illustrative models are not measured factory results.

The paper's rack · 14 boltsBolt length: reading time

13Part IV · The GFT Synthesis: Modular Entry, Partnered Midstream and the Measured Factory

A sequenced performance-management and edge-first AI stack lets a greenfield African factory match Asian conversion efficiency without legacy data

A greenfield plant has no operating history, so GFT's proposed stack builds a six-family KPI ledger and validates edge inspection from the first shift, then admits forecasting and generative design only when their data and physical-validation gates are passed.

  • 13.1Artificial intelligence fails when the shop floor has no trustworthy denominators
  • 13.4Six KPI families turn daily production into a governed operating record
  • 13.6A fixed management cadence makes the workbook a system of record
  • 13.8Edge vision can start early, but only local validation earns production use
Reading time
26 min
Exhibits
6
Sections
8
Continue to chapter 13
← Previous: chapter 11, Grid unreliability, captive solar-plus-storage economics and wet-processing water intensity dictate corridor versus cotton-belt factory sitingContinue the threadChapter 13 · A sequenced performance-management and edge-first AI stack lets a greenfield African factory match Asian conversion efficiency without legacy data

Find a thread.

Type to search.