PURCHASING & SUPPLY CHAIN CHEAT-SHEET - GUIDINEE.COM
Reference points to keep at hand during the capstone project.

=== FUNDAMENTALS ===
- PURCHASING (strategic: choosing/negotiating suppliers) vs SUPPLY/PROCUREMENT (operational: ordering, tracking) vs SUPPLY CHAIN (end-to-end flows).
- Three flows to synchronize: PHYSICAL (goods) | INFORMATION (orders, forecasts) | FINANCIAL (payments).
- BULLWHIP effect: demand variations amplify up the chain -> sharing information reduces the swings.
- CQD triangle (cost/quality/lead time): you ARBITRATE, you do not maximize all three. + a 4th dimension: RISK.
- TCO (total cost of ownership) = price + transport + customs + stock + non-quality + admin + RISK. Compare on TCO, not on the sticker price.
- Segmentation (Kraljic): SIMPLE (efficiency) | LEVERAGE (cost, competition) | BOTTLENECK (security) | STRATEGIC (partnership).
- Vocabulary: RFQ (price request) | PO (order) | DN (delivery) | LEAD TIME (total delay) | MOQ (minimum order) | SKU (item) | OTIF (on time AND in full).

=== PURCHASING PROCESS ===
- ~80% of cost is locked at the SPECIFICATION of the need -> CHALLENGE the need (overspecification, "just in case" quantities, false urgencies, imposed brands).
- Express the need as an expected RESULT, not a brand. FUNCTIONAL spec (results) > TECHNICAL spec (imposed solution) as far as possible.
- SOURCING: aim for several CREDIBLE candidates; QUALIFY (legal, finances, references CONTACTED, capacity, sample + test order). Promises != proof.
- Mono-sourcing (simplicity + dependence) vs multi-sourcing (security + management): arbitrate by CRITICALITY. Far-away cheaper per unit -> judge on TCO (MOQ, lead times, customs, risk).
- CONSULTATION: 3 to 5 suppliers, the SAME information for all, a weighted MULTI-CRITERIA grid defined BEFORE opening. Choose best value, not lowest bid. Compare on identical scope.
- P2P: PR -> PO -> confirmation -> CONTROLLED RECEIPT -> invoice -> 3-WAY MATCH -> payment.
  Golden rules: no purchase without a PO | no receipt without a check | no payment without matching. Delivery reserves: IMMEDIATE, precise, WRITTEN.

=== SUPPLIERS & NEGOTIATION ===
- SELECTION on explicit weighted criteria (CQD + finances + capacity + certifications + CSR). On-site AUDIT for high stakes (the visit proves, the file declares).
- CONTINUOUS EVALUATION: OTIF, non-conformities, disputes, price -> retention / progress plan / exit.
- NEGOTIATION: won BEFORE the table (info, objective, BREAK POINT, screened BATNA). Without an alternative, you beg.
  Negotiate the TCO (payment, MOQ, free freight, lead times, warranty, price revision), not price alone. EXCHANGE each concession for a counterpart.
  Lasting relationship -> INTEGRATIVE deal (widen the scope) > brutal power play (a squeezed supplier makes it up elsewhere).
- CONTRACT: price + STABILITY (revision clause), lead times + penalties APPLIED, quality tied to the spec, warranty, liability, EXIT/reversibility.
  GTS (seller) vs GTP (buyer): settle BEFORE signing. Read the clauses that only apply when there is a problem. Law VARIES by country.
- INCOTERMS 2020 (always rule + PLACE + version): EXW (all buyer) | FOB (switch at loading) | CIF (seller pays transport+insurance but RISK already transferred!) | DAP | DDP (all seller).
- SRM: segment (deep relationship for the STRATEGIC), performance reviews, RECIPROCAL progress plans. Disputes: factual, fast, graduated, traced.

=== STOCK ===
- Stock is a BUFFER with functions (service, safety, cycle, anticipation). No function -> no reason to exist. Holding costs 15-30% / year.
- Three costs in tension: HOLDING (too much) | ORDERING (too-frequent orders) | STOCKOUT (not enough, the most expensive).
- ACCOUNTING valuation (FIFO / WAC; LIFO often banned) != PHYSICAL rotation (FIFO always; FEFO for dated products). Two independent topics.
- REORDER POINT = (average demand x average lead time) + SAFETY STOCK. Safety stock pays for variability (demand, lead time) per the target service level (95% << 99% in cost).
- Order quantity: ordering/holding trade-off (Wilson, square-root shaped), aligned with MOQ, packaging, free freight, discounts.
- ABC (Pareto): ~20% of items = ~80% of value -> fine management of A, simple for C. Cross with CRITICALITY (a vital C deserves an A's safety).
- ROTATION (consumption / average stock) and COVERAGE (days/weeks): in trend and comparison, ALWAYS paired with the SERVICE LEVEL. Hunt down DORMANT items.
- RELIABILITY (IRA >= 95%) before any optimization. CYCLE COUNTING (errors near the cause) > annual only. Discrepancy: fix the figure AND the process cause.

=== LOGISTICS & TRANSPORT ===
- WAREHOUSE: ZONING (flow without backtracking, I- or U-shaped) | ADDRESSING (prerequisite for everything) | SLOTTING by rotation (fast movers near shipping, golden zone) | SAFETY (rack loads!).
- PICKING: >50% of the time in TRAVEL. Picking path, grouping (by batches for many small orders). Organization first, technology next.
  Errors: labeling, scan at the item, proportionate control. Packing = the customer's last image.
- TRANSPORT: road (flexibility) | sea (cost, weeks) | air (days, expensive + carbon) | rail/river | multimodal.
  CONSOLIDATE (group, fill, right segment) | ANTICIPATE (urgency = the budget's enemy) | DOCUMENT/INSURE (carrier liability is CAPPED -> ad valorem insurance). Freight forwarder for international.
- TRACEABILITY: where it is / where it comes from (by BATCH). Standards: EAN-13/GS1, SSCC, 2D codes, RFID. A product recall REQUIRES batch traceability (often regulatory).
- LAST MILE: the most expensive (reversed consolidation). Modes: home / pickup point-locker / click&collect. RETURNS = a process to design and to reduce upstream (product sheet).

=== STEERING & SPECIALIZATIONS ===
- KPI: FEW but KEPT, in ANTAGONISTIC PAIRS (service+stock, savings+OTIF, transport cost+lead time). Each: owner + decisions. Define precisely once and for all.
  OTIF BOTH ways (suppliers to you / you to your customers). SAVINGS: honest baseline, real volumes, TCO scope, validated by finance.
- TOOLS: ERP (SINGLE base, end of re-keying) + WMS (warehouse) + TMS (transport) + e-procurement. Sequence: PROCESS -> DATA -> TOOL. Clean master data.
- FORECAST: always WRONG -> measure the error (it sizes safety stock), build it with others (S&OP), gain in RESPONSIVENESS (short lead times = short horizon).
- LEAN: 7 wastes (overordering, stock, transport, motion, waiting, over-quality, defects). Tools: 5S, kanban, VSM, KAIZEN (small steps). JIT = a reliability machine.
  Doctrine: balance LEAN / RESILIENCE. Stock that protects against a real risk is NOT waste.
- CSR: most of the footprint is in the CHAIN. Embed it into the processes (scoring, supplier charter, lifecycle/TCO). Require VERIFIABLE evidence (anti-greenwashing).
  Cascading duty of vigilance (varies by country). Circular economy: refurbished, rental/usage, repairability. Local sourcing reassessed.
- RISKS: map (probability x impact), hunt SINGLE POINTS OF FAILURE (sole supplier/route/site/person).
  Treat: DIVERSIFY (double sourcing multi-region, nearshoring) | BUFFER (costed strategic stock) | VISIBILITY (tiers 2-3, monitoring, early warning) | TRANSFER (insurance) | ACCEPT consciously.
  CONTINUITY PLAN per major risk (signals, first actions, owner, PRE-QUALIFIED alternative). Resilience costs little and saves everything on the day.

REMINDERS
- Reason in TCO and RISK, never price alone.
- Challenge the need BEFORE buying; consult and compare on a grid; lock the P2P.
- Measure to steer (paired KPIs); build reliability of stock and suppliers (measurement changes everything).
- Be lean on costs WITHOUT A FUNCTION, resilient on PROVEN RISKS.
- Precise rules (law, customs, CSR, accounting) VARY by country: check the real stakes locally.