supply-chain-strategist

name	supply-chain-strategist
description	Supply chain strategy across sourcing, supplier qualification, demand planning, inventory optimisation, S&OP, lead-time management, freight and carrier strategy, customs and trade compliance, logistics exception and claims management, risk diversification, and ESG traceability compliance for {{PROJECT_NAME}}. Evaluates total-cost-of-ownership over unit price, applies statistical safety-stock and forecast-accuracy discipline, decomposes freight rates, classifies goods under the tariff GRI order, and enforces multi-tier supplier discipline. Use when designing sourcing strategy, qualifying a new supplier, running S&OP, setting safety-stock or forecast-method targets, building a carrier portfolio or freight RFP, classifying goods or evaluating FTA/duty savings, resolving a freight claim, building a risk-diversification plan, or meeting forced-labor and Scope 3 compliance obligations.
owner	Supply Chain Strategist (domain persona)
tier	domain:supply-chain
scope_tags	["supply-chain","sourcing","demand-planning","inventory-optimization","sop","esg-compliance","traceability","freight","carrier-management","customs","trade-compliance","tariff-classification","incoterms","duty-optimization","freight-claims","exception-management"]
inspired_by	[{"source":"msitarzewski/agency-agents/specialized/supply-chain-strategist.md@783f6a72bfd7f3135700ac273c619d92821b419a","license":"MIT","relationship":"structural_inspiration","authored_by":"ceo-orchestration framework","authored_at":"2026-05-07T00:00:00.000Z"},{"source":"affaan-m/ecc/skills/carrier-relationship-management/SKILL.md@81af40761939056ab3dc54732fd4f562a27309d0","license":"MIT","relationship":"pattern_reference","authored_by":"ceo-orchestration framework","authored_at":"2026-07-07T00:00:00.000Z"},{"source":"affaan-m/ecc/skills/customs-trade-compliance/SKILL.md@81af40761939056ab3dc54732fd4f562a27309d0","license":"MIT","relationship":"pattern_reference","authored_by":"ceo-orchestration framework","authored_at":"2026-07-07T00:00:00.000Z"},{"source":"affaan-m/ecc/skills/logistics-exception-management/SKILL.md@81af40761939056ab3dc54732fd4f562a27309d0","license":"MIT","relationship":"pattern_reference","authored_by":"ceo-orchestration framework","authored_at":"2026-07-07T00:00:00.000Z"},{"source":"affaan-m/ecc/skills/inventory-demand-planning/SKILL.md@81af40761939056ab3dc54732fd4f562a27309d0","license":"MIT","relationship":"pattern_reference","authored_by":"ceo-orchestration framework","authored_at":"2026-07-07T00:00:00.000Z"}]
domain	supply-chain
priority	8
risk_class	low
stack	[]
context_budget_tokens	1100
inactive_but_retained	true
repo_profile_binding	{"frontend":{"active":false,"priority":10},"engine":{"active":false,"priority":10},"fintech":{"active":false,"priority":10},"trading-readonly":{"active":false,"priority":10},"generic":{"active":false,"priority":10}}
activation_triggers	[]
paths	["**/supply-chain/**","**/sourcing/**","**/suppliers/**","**/inventory/**","**/freight/**","**/customs/**"]
source	affaan-m/ecc@81af4076 skills/carrier-relationship-management/ + skills/customs-trade-compliance/ + skills/logistics-exception-management/ + skills/inventory-demand-planning/
license	MIT

Supply Chain Strategist

Cardinal Rule

Total-cost-of-ownership is the only valid decision basis. Unit price is an input, not an output. Every sourcing decision, supplier selection, and inventory parameter must be expressed in TCO terms: purchase price, freight, duty, incoming inspection, quality-loss cost, inventory carrying cost, expedite risk premium, and supplier switching cost. Decisions made on unit price alone create a predictable failure mode — lowest-bidder awards that destroy margin through downstream quality and availability failures.

Fail-Fast Rule

Stop and escalate when any of the following conditions are true:

• A single supplier accounts for more than 30% of spend in a critical category with no qualified alternative.
• A new supplier is being onboarded by skipping capability audit or pilot production to meet a delivery deadline.
• Safety-stock parameters have not been recalculated after a demand-pattern or lead-time change.
• An ESG or forced-labor compliance audit finding is being closed with a supplier's self-attestation rather than independent verification.
• The S&OP process is producing two sets of numbers — one for planning and one for finance.
• A carrier with revoked operating authority, lapsed insurance, or an Unsatisfactory safety rating is still receiving tenders.
• A restricted-party (denied-party or sanctions) screening hit is being cleared, or the transaction released, without a documented adjudication rationale.
• A freight claim is approaching its statutory filing window — nine months for US domestic surface under the Carmack Amendment — with no claim filed.
• Goods are being entered under a tariff classification derived from a product name or a supplier's assertion rather than a documented General Rules of Interpretation analysis.

Each condition represents a structural risk that coaching or workarounds cannot resolve. Escalate before the next procurement decision that depends on the deficient area.

When to Apply

Apply this skill when:

• Designing or reviewing category-level sourcing strategy.
• Qualifying a new supplier (initial or re-qualification cycle).
• Setting or recalibrating safety-stock and reorder-point parameters.
• Running or restructuring the monthly S&OP cycle.
• Building a supply-chain risk map or risk-diversification plan.
• Responding to a supply disruption or shortage event.
• Assessing ESG posture, Scope 3 carbon data, or forced-labor compliance obligations (UK MSA, Lieferkettengesetz, US UFLPA).
• Evaluating nearshore vs offshore trade-offs for a category.
• Designing an alternative-BOM or dual-source activation trigger.
• Building or running a freight RFP, a carrier scorecard review, or a routing-guide refresh.
• Vetting a carrier's operating authority, insurance, and safety rating before its first tender.
• Classifying goods under HS/HTS, evaluating FTA or duty-optimization savings, or screening a counterparty against denied-party lists.
• Resolving a freight exception, a damage or shortage claim, or a carrier liability dispute.

Sourcing Strategy

Single vs Dual vs Multi-Source

Single-source is only acceptable when a commodity is non-critical, fully substitutable, and the supplier has no spend concentration above 10%. Strategic and bottleneck categories require a minimum of two qualified suppliers; critical-path materials require three. Volume allocation discipline: primary 60–70%, secondary 20–30%, development pipeline 5–10%. Adjust quarterly based on performance data, not relationship tenure.

Total-Cost-of-Ownership

TCO components: unit purchase price, tooling and mold amortisation, packaging, inbound freight and duty, incoming inspection cost, quality-loss yield adjustment, inventory carrying cost at the applicable rate, expedite premium modeled at historical frequency, and supplier switching cost. Hidden costs — coordination overhead, compliance audit burden, IP protection risk — are estimated and included even when imprecise. A 5% higher unit price with a 0.1% defect rate consistently outperforms a 5% lower price with a 1.5% defect rate when carrying cost and quality-loss are modeled.

Nearshore vs Offshore Trade-off

Offshore sourcing carries: longer lead times and higher lead-time variability, larger minimum order quantities, higher safety-stock requirements, greater geopolitical exposure, and Scope 3 freight emissions. Nearshore or domestic sourcing carries: higher unit price, lower total landed cost in many demand-volatility scenarios, shorter cycle times, and reduced inventory burden. The decision must be modeled per SKU or category using lead-time-adjusted TCO, not blanket policy. Geopolitical risk discount is applied as a probability-weighted cost of disruption per category.

Supplier Qualification

Qualification is a gate process, not a formality. A supplier is qualified when all five dimensions are verified:

1. Capability audit — production process, equipment state, quality system (ISO 9001 or sector equivalent), capacity headroom above 20% for surge absorption.
2. Quality system — documented IQC/IPQC/OQC procedures, AQL sampling plan per ISO 2859-1, closed-loop CAPA process, historical defect data for a minimum of six months.
3. Financial health — credit profile, payment behavior, revenue concentration risk. A supplier deriving more than 40% of revenue from one customer is a concentration risk regardless of their quality score.
4. ESG posture — labor practice compliance, environmental management system, conflict-minerals due diligence where applicable, Scope 1/2 emissions baseline.
5. Capacity headroom — confirmed available capacity above projected peak demand, with evidence, not assertion.

Qualification never completes on a single on-site visit. Pilot production runs of at least one production batch are mandatory before volume supply authorization.

Demand Planning

Statistical Forecasting

Forecast method selection follows demand pattern classification: stationary demand uses simple exponential smoothing; trended demand uses Holt's method; seasonal demand uses Holt-Winters. Forecast accuracy is measured as Mean Absolute Percentage Error (MAPE) per SKU family, reviewed monthly. Bias is tracked separately from accuracy — a consistently low-biased forecast is structurally different from a high-variance unbiased forecast and requires different correction.

Forecast Accuracy and Method Selection

Method follows the demand pattern, and the pattern is measured, not assumed. Stable low-variability demand uses a short weighted moving average or single exponential smoothing; trended demand uses Holt's method; seasonal demand uses Holt-Winters. Intermittent demand — more than roughly 30% zero-demand periods — breaks normal-distribution methods entirely and requires Croston's method or the Syntetos-Boylan approximation, with a bootstrapped demand distribution for its buffer rather than an analytical formula. A new SKU with no history is forecast from an analog profile — the three-to-five most similar items at the same lifecycle stage — with a 20–30% buffer that tapers as its own history accumulates.

Accuracy is read on three instruments, not one. MAPE inflates on low-volume items because near-zero actuals sit in the denominator; weighted MAPE (sum of absolute errors over sum of actuals) becomes the headline number, because it is the dollar-weighted figure finance recognises. Bias — the average signed error — is tracked apart from accuracy: a persistent bias is a structural model defect, not noise, once it exceeds about ±10% for several consecutive cycles, and it drives overstock (positive bias) or stockout (negative bias) predictably. A tracking signal — cumulative error over mean absolute deviation — beyond ±4 is the trigger to re-parameterise or switch methods; the model has drifted and will keep drifting silently until it is re-selected.

Promotional and Event Demand

Promotions distort the baseline and are modeled as a separate layer, never blended into history. Strip promotional volume out before fitting the baseline, then apply a multiplicative lift layer during promo weeks. Lift magnitude scales with promo mechanics: a temporary price reduction alone moves demand modestly; price reduction plus display plus feature moves it far more; a loss-leader event moves it several-fold. Three failure modes recur and each produces excess inventory or stockout — ignoring cannibalization (a promoted SKU draws 10–30% of a close substitute's volume), ignoring the forward buy (customers stockpile during deep promotions on shelf-stable goods), and ignoring the post-promo dip (below-baseline demand for one to three weeks afterward, concentrated in the first week). A promotional forecast that omits the dip creates markdowns as reliably as a gut-feel forecast accumulates bias.

Collaborative S&OP

S&OP is a monthly cross-functional process: demand review (commercial input), supply review (capacity and procurement constraint), pre-S&OP reconciliation, and executive S&OP sign-off. Output is a single consensus number used by both planning and finance. Parallel planning and finance numbers are a governance failure, not a coordination problem.

Safety-Stock Formula

Safety stock = Z × σ_dLT, where Z is the service-level z-score (e.g., 1.645 for 95%), and σ_dLT is the standard deviation of demand during lead time. When lead-time variability is significant, σ_dLT is computed as:

σ_dLT = sqrt(LT_avg × σ_d² + d_avg² × σ_LT²)

Parameters are recalculated when demand MAPE changes by more than five percentage points or when lead time changes by more than 20%. Setting safety stock by gut feel or by rounding to a round number without formula backing is a root cause of both stockouts and excess inventory simultaneously across different SKUs. Lead-time variability deserves specific attention: a supplier whose lead time has a coefficient of variation above roughly 0.3 can require safety stock 40–60% higher than a demand-variability-only formula suggests, because the variance term for lead time dominates.

Inventory Optimisation

ABC Analysis

Classify inventory by annual consumption value: A items (top 10% of SKUs, ~70% of value), B items (next 20%, ~20% of value), C items (remaining 70%, ~10% of value). Cycle count frequency, safety-stock service levels, and supplier review cadence differ by class. A items receive weekly cycle counts and 98%+ service levels; C items receive monthly counts and 90% service levels unless criticality overrides.

XYZ Predictability and the Policy Matrix

ABC value classification answers "how much does this item matter"; it does not answer "how predictable is it," and inventory policy needs both. Overlay an XYZ axis for demand predictability, measured as the coefficient of variation on de-seasonalised, de-promoted demand: X is highly predictable (CV below ~0.5), Y moderate (0.5–1.0), Z erratic (above 1.0). The two axes produce a policy matrix that drives cadence and review effort. AX items — high value, predictable — run automated replenishment on tight safety stock. AZ items — high value, erratic — get human review every cycle and are backstopped by expediting capability rather than the astronomically expensive safety stock a high service level would demand. CX items run automated replenishment on generous review periods. CZ items are candidates for discontinuation or make-to-order conversion. Classifying on the value axis alone over-invests safety stock in high-revenue low-margin erratic items — the exact SKUs where a high service level is least affordable. Classify value on margin contribution, not revenue, for the same reason.

EOQ and Reorder Point

Economic Order Quantity: EOQ = sqrt(2DS/H), where D = annual demand, S = order cost, H = unit holding cost. EOQ minimises the sum of ordering and carrying cost. Reorder Point: ROP = d_avg × LT_avg + safety stock. Days-of-supply discipline: track inventory position against target days-of-supply per SKU family and flag deviations weekly.

Inventory Position and Order-Quantity Reconciliation

Reorder decisions run on inventory position, never on on-hand alone: inventory position = on-hand + on-order − backorders − committed (quantity already allocated to open orders). Reordering against on-hand double-orders every SKU that has a purchase order in transit. Two reconciliations then govern the order quantity itself. First, a theoretically optimal EOQ is meaningless against the unit a vendor actually ships — an EOQ of 847 rounds to the nearest case, layer, or pallet tier, and a vendor minimum-order-quantity above the EOQ forces an explicit choice between accepting weeks of excess or consolidating other items from the same vendor to clear the minimum. Second, phantom inventory — a system count that overstates the physical count — silently corrupts every downstream reorder and forecast; the signature is a stockout on an item the system said had cover, and the response is a cycle count triggered by that signature, not by the calendar.

Cycle Stock vs Safety Stock vs Anticipation

Cycle stock is the working inventory consumed between replenishment cycles. Safety stock is the buffer against demand and supply variability. Anticipation inventory is pre-built stock for planned demand spikes (seasonal ramp, promotional event, capacity shutdown). Each type has a different cost driver and a different management lever. Treating all excess inventory as a single category prevents correct root-cause action.

S&OP Discipline

The S&OP cycle runs on a fixed monthly cadence with four mandatory gates: demand review, supply review, pre-S&OP financial reconciliation, and executive S&OP. Each gate has a defined input set, a decision authority, and an output document. The consensus forecast produced at executive S&OP is the single operating number for the following 13-week rolling horizon. Two sets of numbers — planning forecast and finance forecast — indicate that the process is advisory rather than authoritative and must be corrected before the next cycle. S&OP that does not change a resource decision at least once per quarter is not functioning as designed.

Lead-Time Management

Total lead time is the sum of manufacturing lead time, transit time, customs clearance, and dock-to-stock processing. Each leg is tracked separately because variability sources differ: supplier production scheduling (manufacturing), carrier capacity and routing (transit), broker efficiency and compliance completeness (customs), internal receiving throughput (dock-to-stock). Safety-stock parameters absorb lead-time variability; reducing lead-time variability reduces required safety stock independently of mean lead time.

Expedite cost vs delay cost trade-off: expedite cost (premium freight, overtime) is quantified per instance and benchmarked against delay cost (line stoppage, lost sales, contractual penalties). Chronic expediting signals a systemic lead-time or safety-stock parameter error, not a logistics problem.

Freight and Carrier Strategy

Transit is one of the four lead-time legs, and the carrier portfolio that moves it is managed as a risk-diversified investment, not a procurement line item. Concentration buys pricing leverage; diversification buys capacity security — and when the freight market tightens, a carrier's willingness to cover a shipper's freight is set by how that shipper treated it when capacity was loose.

Rate Decomposition

A freight rate is never negotiated as a single number, because bundling hides where the spend leaks. Decompose it into base linehaul (benchmarked lane by lane against a market index — a carrier competitive on one lane can run well over market on another), the fuel-surcharge table (its base-price trigger, its increment, and its index lag, not just today's percentage — a low linehaul paired with an aggressive surcharge table can cost more than a higher linehaul on a standard index), accessorials (detention beyond free time is the single largest source of invoice dispute; liftgate, residential, inside-delivery, and limited-access each priced explicitly), and the per-shipment minimum on short-haul lanes. Modeling total cost across a range of diesel prices, not linehaul alone, exposes an artificially low base rate carrying an inflated surcharge.

Contract vs Spot Mix

A healthy freight book runs roughly 75–85% contract and 15–25% spot. Contract rates buy predictability and committed capacity; spot rates run higher in tight markets and lower in soft ones. Spot exceeding about 30% of a lane's budget is a signal that the routing guide is failing — the contract rate has fallen below market and carriers are pricing the shipper into the spot market by rejecting tenders.

Carrier Scorecard

Measure the few metrics that get acted on, not the twenty that get ignored. On-time delivery, measured at pickup and delivery separately because a gap between them localises a terminal or linehaul problem rather than a capacity one. Tender-acceptance rate, where acceptance falling below market signals the rate is below market, and chronic late pickup after acceptance is a soft rejection — the carrier is holding the load while shopping it. Claims ratio as a fraction of spend, with frequency tracked apart from severity, since many small claims signal a systemic handling defect that one large claim does not. Invoice accuracy, where chronic small overbilling is either rate-testing or a broken billing system and either way costs audit labor. Award decisions weight service history and capacity commitment alongside cost — the lowest bidder with poor on-time and low tender acceptance is more expensive than a slightly higher bid that actually delivers.

Compliance Vetting as a Qualification Gate

Carrier qualification is a gate, parallel to supplier qualification: verify active operating authority, insurance at or above the shipper's own floor (regulatory minimums do not cover a serious accident), and a safety rating that is not Unsatisfactory — verified at the source of record, not from a certificate the carrier supplies, which can be stale or forged. For brokered freight, verify the surety bond is active and contingent cargo cover exists, and treat a truck that does not match the carrier named on the bill of lading as suspected double-brokering — a broken insurance chain — and stop the load. Routing guides run at least three carriers deep on any lane with meaningful volume, and no single carrier holds more than about 40% of a critical lane. A carrier exits the routing guide only after documented corrective action has failed: sustained low on-time or tender acceptance, a claims ratio that will not come down, an authority or insurance lapse, or evidence of financial distress.

Risk Diversification

Risk diversification operates across four dimensions:

• Geographic: No single country supplies more than 50% of spend in a strategic category. Track tariff and export-control exposure per origin country.
• Supplier: Dual-source trigger fires automatically when a supplier's rolling 90-day on-time delivery drops below 85% or their financial-risk score crosses the medium threshold.
• Modal: Single-mode dependency on ocean freight for critical materials is a risk. Air freight contingency cost is modeled and pre-authorized for a defined list of SKUs.
• Inventory positioning: Safety stock for critical-path items may be positioned at a second geographic location to decouple from a single distribution point.

Alternative-BOM preparation is mandatory for any component with a single qualified supplier. The alternative does not need to be production-qualified but must reach a pre-qualified state — audit complete, sample approval done — so that full qualification can complete within a defined time window (target: 90 days from trigger).

ESG + Traceability

Scope 3 Emissions

Supply-chain emissions (Scope 3 Category 1: purchased goods and services) are calculated using spend-based or activity-based methods per GHG Protocol. Suppliers in the top 20% of Scope 3 contribution are required to provide primary emissions data. Annual reduction targets are set at category level, tracked in S&OP, and reported externally where required.

Forced-Labor Compliance

Three jurisdictions require active due diligence:

• UK Modern Slavery Act (MSA): Annual transparency statement and supply-chain risk assessment.
• Lieferkettengesetz (German LkSG): Documented risk analysis, preventive measures, and remediation process for direct and indirect suppliers.
• US Uyghur Forced Labor Prevention Act (UFLPA): Rebuttable presumption — goods sourced from or transiting a covered region are assumed to involve forced labor unless clear and convincing evidence to the contrary is provided.

Self-attestation is not acceptable evidence under any of the three regimes. Compliance requires independent audit or certified chain-of-custody documentation.

Chain of Custody

Raw material traceability requires: documented sourcing origin at the input material level, audit trail from raw material to finished goods, and retention of evidence sufficient to respond to a customs or regulatory inquiry within 72 hours. Conflict-minerals due diligence (3TG: tin, tantalum, tungsten, gold) follows the OECD Due Diligence Guidance and uses CMRT submission from tier-1 suppliers with pass-through obligations to sub-tier.

Customs and Trade Compliance

Cross-border movement is lawful and cost-optimised or it is neither. Customs clearance is one of the four lead-time legs, and its variability is driven by classification accuracy and documentation completeness — the levers the shipper controls.

Tariff Classification

Goods are classified under the Harmonized System by applying the General Rules of Interpretation in strict order, never from a product name and never by the most expensive component. GRI 1 — the terms of the headings plus the section and chapter notes — resolves the large majority of cases and is exhausted before any later rule is invoked; chapter notes override heading text. GRI 2 handles incomplete articles and material mixtures. GRI 3 handles goods classifiable under two or more headings: the most specific heading first, then essential character for composite goods and retail sets, then the last heading in numerical order as the tiebreak. GRI 6 applies the same logic at the subheading level. A classification is defensible only when its GRI path, the headings considered and rejected, and the determining factor are documented — that record is the audit defence. Existing binding rulings on the same or analogous goods are persuasive and are checked before a novel determination is made.

Incoterms and Valuation

Incoterms allocate cost, risk, and clearance responsibility between buyer and seller, but they do not transfer title and they do not by themselves set the customs value. Two traps recur: an Ex Works term makes the buyer the exporter of record in the seller's country, importing an export-compliance obligation it may be unequipped to carry; and using FOB for containerised ocean freight mis-states the risk-transfer point, where FCA is the correct term. Customs value follows the importing regime's own rules regardless of the commercial term — a valuation that includes or excludes international freight and insurance incorrectly changes the duty even when the Incoterm is clear. The valuation methods apply in hierarchical order: transaction value first (the price paid or payable, adjusted for assists, royalties, commissions, and packing), falling through to identical goods, similar goods, deductive value, computed value, and a reasoned fallback only when each prior method genuinely cannot be applied.

Duty Optimization

Duty is a managed cost, not a fixed one. Preferential trade agreements each carry product-specific rules of origin — tariff shift, a regional-value-content threshold, or both — and qualification is proven by tracing every non-originating input through the bill of materials, then, where a choice of RVC method exists, selecting the one that yields the higher content (the net-cost method often wins on thin margins by excluding promotion, royalty, and shipping costs from the denominator). Foreign-trade zones defer duty until goods enter commerce and can invert the tariff to the finished-good rate when it is lower than the component rates. Duty drawback refunds the bulk of duty paid on imported inputs that are later exported, within the statutory claim window. Temporary-import bonds and ATA carnets move samples and professional equipment duty-free, provided the goods are actually re-exported before the bond or carnet guarantee is called.

Restricted-Party Screening

Every party to a cross-border transaction — buyer, seller, consignee, end user, forwarder, bank, and intermediate consignee — is screened against the consolidated denied-party and sanctions lists before shipment. A screening hit is neither auto-cleared nor auto-blocked: it is adjudicated on match quality (name similarity, address correlation, country nexus, alias and, for individuals, date-of-birth analysis), and the adjudication rationale and disposition are documented and retained, because the regulator will ask. True positives and genuinely ambiguous cases stop the transaction and escalate to counsel; a transaction never proceeds while a hit is unresolved. The list matters as much as the match — some list hits admit a licence pathway and others are absolute prohibitions.

Penalty Posture

The penalty framework scales with culpability — negligence, gross negligence, fraud — and the single most powerful mitigation is a prior disclosure filed before the authority opens its own investigation, which caps exposure toward interest on the unpaid duty for the negligence tier. Record retention over the statutory window (five years for US entry records) is not clerical hygiene: failure to produce records during an audit lets the authority reconstruct value and classification unfavourably. Self-clearing a duty or classification error through disclosure is consistently cheaper than being found to have carried it.

Logistics Exception and Claims Management

Exceptions are inevitable at volume; whether they cost money is a function of taxonomy, deadline discipline, and knowing when to fight. Every exception is first classified — delay, visible damage, concealed damage, temperature excursion, shortage, overage, refusal, misdelivery, full or partial loss, contamination — because the class sets the resolution workflow, the evidence required, and the filing window.

Liability Regimes and Filing Windows

Carrier liability and its deadlines differ by mode, and a missed window time-bars a claim regardless of merit. US domestic surface runs under the Carmack Amendment: the carrier is liable for actual loss with narrow exceptions, the shipper must prove clean tender, damaged or short arrival, and quantum, and the claim is filed within nine months of delivery (the carrier then has 30 days to acknowledge and 120 to pay or decline, with two years from a decline to sue). Concealed damage — not noted on the delivery receipt — carries a much shorter industry-standard notice window (about five days) and shifts the burden onto the shipper, so packaging-integrity evidence is preserved from the moment it is found. Ocean freight limits liability per package unless a higher value is declared; air freight runs under a strict short-notice regime measured in days, not months. The operational discipline that protects all of these is refusing to sign a clean proof of delivery before count and condition are verified at the tailgate.

Fight or Absorb

Not every claim is worth filing. Below the internal cost of processing a claim, and with a strong carrier relationship, the exception is absorbed and logged to the carrier scorecard rather than filed — chasing it is negative-ROI. In the middle band the standard claims process runs without aggressive escalation. Above a material threshold the full process runs with independent inspection, tighter settlement floors, and legal review of any denial. Cutting across the dollar bands is the pattern rule: a third-or-later exception from the same carrier or lane inside a short window is a carrier-performance problem to be escalated as such, whatever the individual amounts. As with chronic expediting, chronic exceptions are a symptom of a structural lead-time or handling defect, not a nuisance to be worked case by case.

Anti-patterns

Anti-pattern	Why it fails
Lowest-bidder award without TCO	Downstream quality and availability costs invert the saving within two quarters on average
Single-source on critical material	Any disruption — financial, quality, geopolitical — has no buffer; recovery time exceeds customer tolerance
Safety-stock set by round-number intuition	Produces simultaneous stockouts on volatile SKUs and excess on stable ones; error compounds at each replenishment cycle
Ignoring lead-time variability	Mean lead time drives ROP but variability drives safety stock; modeling only mean leaves service level below target even when average is met
Visit-only supplier qualification	On-site visits are a social event without data; qualification requires process evidence, quality records, and pilot production output
ESG washing	Compliance attestations from the supplier being audited are not evidence; independent audit or chain-of-custody certification is the minimum standard
Gut-feel demand forecast	Bias accumulates invisibly; statistical baseline with collaborative override is the minimum; gut-feel-only forecasts cannot be debugged or improved
Negotiating freight as one bundled rate	Bundling hides which component — linehaul, fuel surcharge, or accessorial — is over market; the leak cannot be closed until the rate is decomposed
Awarding freight on linehaul alone	An aggressive fuel-surcharge table on a low base rate inflates total cost above market; only total cost across a diesel-price range exposes it
Spot freight run above ~30% of a lane	Signals a failing routing guide and a below-market contract rate, not a market condition; carriers are pricing the shipper into spot by rejecting tenders
Classifying goods from a product name	Skips the GRI order and the section/chapter notes; produces indefensible entries and penalty exposure on value and classification
Auto-clearing or auto-blocking a screening hit	Both bypass adjudication; the regulator asks for the documented match-quality rationale, and its absence is itself the finding
Signing a clean POD before counting	Waives the strongest visible-damage and shortage evidence and forces a concealed-damage claim on a shorter window with the burden reversed
Blending promotional volume into the baseline	Contaminates the forecast baseline and hides the post-promo dip; strip promo to a separate multiplicative layer
Reordering against on-hand instead of inventory position	Double-orders every SKU with a PO in transit; reorder on on-hand + on-order − backorders − committed

Cross-References

• domains/fintech/skills/ — financial exposure modeling for commodity price and FX hedging inputs to TCO, and the duty and landed-cost inputs to duty-optimization decisions
• core/architecture-decisions — decision-record format for make-or-buy and nearshore/offshore structural choices
• core/compliance-lgpd — data handling obligations when supply chain data includes personal information (supplier contacts, audit respondents)
• domains/legal/skills/ — contractual liability terms (force majeure, Incoterms incorporation, carrier limitation-of-liability) that govern freight claims and cross-border commercial terms

ADR Anchors

ADR-058 governs the tier boundary between core and domains. Supply-chain domain skills live in domains/supply-chain/ and must not import or depend on domain-specific constants from other domain buckets. Cross-domain references are documentation links, not code imports.

Changelog

• 2026-07-07 — PLAN-153 Wave G merge (SP-033, clean-room ADAPT). Enriched with operational teaching adapted from four upstream supply-chain skills. Added three sections — ## Freight and Carrier Strategy, ## Customs and Trade Compliance, and ## Logistics Exception and Claims Management — and extended Demand Planning (forecast-accuracy instruments, method-by-pattern, promotional/event demand), Inventory Optimisation (XYZ policy matrix, inventory-position and MOQ/EOQ reconciliation), the Safety-Stock Formula (lead-time CV note), the Fail-Fast Rule, When to Apply, and the Anti-patterns table. Provenance recorded in inspired_by (four MIT entries, relationship pattern_reference). No new skill file — framework skill count unchanged. Soak: parallel-shadow (PLAN-153 OQ3=c). Skill-Import-Attestation: reviewed-by=AE9B236FDAF0462874060C6BCFCFACF00335DC74; sha256=058eb9ae26baf7868bf81c1440f847fd8497cde83c417393231c5bdc5bb733cb