What this is

The vendor performance scorecard: the dimensions, weighting model, and period-based measurement framework that translates vendor compliance data into a composite score used for purchase allocation, renegotiation, and rationalization decisions.

Purpose

The scorecard converts raw compliance events — a short shipment here, a late ASN there — into a structured performance record that buyers can act on. Without a scorecard, vendor relationship decisions are made on relationships and gut feel. With one, they're made on evidence. The scorecard also defines the chargeback trigger thresholds: poor scorecard performance in a given dimension automatically generates a chargeback at the established administrative fee rate.

Structure

Measurement period — Scorecard runs by month, season, and year. Purchases must be linked to receipts that may cross period boundaries; the scoring methodology must account for this lag.

Core performance dimensions:

Weighting factors — Each dimension is assigned a relative weight that sums to 100. Weights must be established before any evaluation period begins, agreed with vendor leadership. A retailer with a fragile supply chain might weight fill rate and on-time delivery heavily; one with quality issues might weight acceptance criteria.

Composite score — Weighted average of dimension scores produces a single vendor score per period. Composite score drives: (1) purchase allocation — high-scoring vendors receive preferred buying; (2) negotiation posture — score becomes the buyer's opening data point; (3) rationalization candidate list — vendors below a defined threshold for N consecutive periods are flagged for rationalization review.

Chargeback triggers — Each compliance dimension has a defined threshold below which automatic chargebacks are generated. The chargeback matrix specifies the administrative fee rate and occurrence escalation logic. The scorecard is the source data; the chargeback matrix is the enforcement mechanism.

Consumers

Buyers use the scorecard in negotiation sessions and purchase allocation decisions. The Operations Agent monitors scorecard trends and alerts when a vendor approaches a rationalization threshold or a chargeback-trigger threshold. The smart contract layer, for smart-contract-native vendors, encodes scorecard thresholds as on-chain conditions: falling below threshold automatically triggers the contractual consequence without human intervention.

Invariants

• Scorecard dimensions and weights must be agreed with the vendor before the measurement period begins. Retroactive reweighting is not permitted.
• The scorecard must be run on a consistent schedule — monthly at minimum, with seasonal and annual aggregates. Irregular scoring destroys its credibility in negotiation.
• Purchases linked to cross-period receipts must use a defined methodology (receipt date or PO close date) applied consistently. The methodology must be documented.

Platform (2030)

Agent mandate: Operations Agent computes scorecard dimensions continuously — not at period close. Business Agent reads the current scorecard composite in vendor negotiation preparation. Neither agent makes chargeback decisions based on the scorecard; the chargeback matrix and the smart contract engine execute those automatically.

Live scorecard from live contract state. Traditional scorecards are period-close computations: query transaction history at month-end, produce a report. In the Canary Go model, the vendor's smart contract IS the scorecard. Each compliance event submitted to the contract updates the relevant dimension counter in real time. Operations Agent reads contract state continuously — the scorecard composite is always current, not a snapshot. The vendor can query their own contract state at any time to see their current performance, eliminating the "we didn't know we were below threshold" dispute category.

Exception-first alerting. Operations Agent does not generate periodic scorecard reports. It monitors the trend on each dimension and alerts only when: (1) a dimension composite crosses below the chargeback trigger threshold; (2) the rate of decline suggests threshold breach within N periods; (3) the overall composite drops below the rationalization candidate threshold. Alerts surface to the merchant dashboard with one-click context. Normal performance generates no noise.

MCP surface. scorecard(vendor_id) returns current composite score and all dimension scores in a single call. scorecard_trend(vendor_id, n_periods) returns composite and dimension trend data — the input for negotiation preparation. rationalization_candidates() returns vendors below composite threshold with dimension breakdown. Low-bandwidth, agent-optimized.

RaaS. Scorecard inputs — fill rate, on-time delivery, ASN accuracy — are derived from receipt events (receiving confirmations, PO completions, ASN matches). The scorecard is a computed view over an append-only event log, not a mutable record. Sequence matters: a fill rate calculation that uses out-of-order receipt events produces a wrong numerator. scorecard(vendor_id) should be a pre-aggregated view, not a real-time join over raw events — at scale (many vendors × many periods × many line items) real-time join cost is prohibitive. Scorecard history exportable by period for vendor negotiation and rationalization analysis.

Related

• retail-vendor-compliance-standards — the clause matrix that defines what is being measured
• retail-vendor-lifecycle — the lifecycle stage that triggers scorecard evaluation
• retail-chargeback-matrix — translates scorecard dimension failures into financial consequences
• retail-three-way-match — the receipt-side data source for fill rate and ASN accuracy