InfiniSynapse
What a supplier scorecard is什么是供应商评分卡
A supplier scorecard, also called a vendor scorecard, is a governed record that converts defined supplier performance evidence into comparable KPI results, normalized scores, trends, gates, and review actions for a stated supplier scope and period.
供应商评分卡也称 vendor scorecard，是一项受治理记录：针对明确的供应商范围和期间，把供应商绩效证据转换为可比较的 KPI 结果、标准化得分、趋势、门槛与评审行动。
The scorecard is not just a colored dashboard or spreadsheet total. It must retain the numerator, denominator, event grain, target, scoring rule, sample size, data coverage, evidence source, owner, and version behind every result. A reviewer should be able to move from the total score to the exact order line, receipt, defect, invoice, service case, audit record, or approved judgment that produced it.
评分卡不只是彩色仪表板或表格总分。每项结果都必须保留分子、分母、事件粒度、目标、评分规则、样本量、数据覆盖、证据来源、责任人与版本。评审者应能从总分下钻到产生该结果的订单行、收货、缺陷、发票、服务工单、审计记录或获批判断。
Its purpose is consistent review, not automatic punishment. It supports evidence-based discussion, supplier development, escalation, allocation review, contract conversation, and further investigation under an approved decision policy.
其目的在于一致评审，而非自动惩罚。它在获批决策政策下支持基于证据的讨论、供应商发展、升级处理、份额复核、合同沟通与进一步调查。
Keep scorecard, selection, risk, and management distinct区分评分卡、选型、风险与管理流程
A periodic artifact that calculates and presents governed performance results and actions.
周期性计算并呈现受治理绩效结果与行动的载体。
Compares bids, capabilities, price, risk, and evidence before award or allocation.
在授标或分配前比较投标、能力、价格、风险与证据。
Evaluates potential exposure, likelihood, impact, controls, and treatment.
评估潜在暴露、可能性、影响、控制与处置。
Runs the wider cycle of expectations, review, corrective action, closure, and improvement.
运行期望、评审、整改、关闭与改善的完整循环。
This page focuses on the scorecard asset. The broader supplier performance management process and forward-looking supplier risk assessment require different decisions. Do not silently combine them into one opaque total.
Define the scoring unit before selecting KPIs选择 KPI 前先定义评分单位
A legal supplier may operate multiple sites, supply unrelated categories, serve different plants, and perform under different contracts. Decide whether one row represents supplier, supplier-site, supplier-category, supplier-site-item family, supplier-contract, or supplier-service relationship. Aggregate only when the decision and evidence remain meaningful.
一个供应商法人可能运营多个站点、供应不同品类、服务不同工厂，并在不同合同下履约。必须决定一行代表供应商、供应商—站点、供应商—品类、供应商—站点—物料族、供应商—合同还是供应商—服务关系。只有在决策与证据仍有意义时才聚合。
Publish the time boundary. State transaction dates, posting dates, promise-date hierarchy, timezone, late-arriving records, restatements, and whether rolling trends use the current formula version or the historical version.
公布时间边界。说明交易日期、过账日期、承诺日期层级、时区、迟到记录、重述，以及滚动趋势使用当前公式版本还是历史版本。
Also define eligibility: active versus closed orders, buyer-caused exceptions, samples, emergency purchases, disputed defects, credits, incomplete cases, and entities with too little evidence. Hidden exclusions destroy comparability.
还要定义资格：活动与关闭订单、买方原因例外、样品、紧急采购、有争议缺陷、贷项、未完成案例以及证据过少的实体。隐藏排除项会破坏可比性。
Supplier scorecard template fields供应商评分卡模板字段
| Block模块 | Fields字段 | Why it matters作用 |
|---|---|---|
| Identity身份 | Supplier, site, category, contract, owner, tier, period, currency供应商、站点、品类、合同、责任人、层级、期间、币种 | Prevents unlike relationships from being mixed避免混合不可比关系 |
| Metric contract指标口径 | KPI, purpose, formula, grain, target, direction, scoring curve, versionKPI、目的、公式、粒度、目标、方向、评分曲线、版本 | Makes the result reproducible使结果可复现 |
| Result结果 | Numerator, denominator, actual, score, weight, contribution, trend分子、分母、实际值、得分、权重、贡献、趋势 | Shows how the total was formed说明总分形成过程 |
| Confidence置信 | Sample size, coverage, freshness, missing reason, estimate flag, dispute status样本量、覆盖率、新鲜度、缺失原因、估算标志、争议状态 | Separates score from evidence strength区分分数与证据强度 |
| Decision决策 | Gate, band, reviewer, comment, action, owner, due date, approval, outcome门槛、区间、评审人、评论、行动、责任人、到期日、审批、结果 | Connects measurement to governed follow-up把衡量连接到受治理后续行动 |
A spreadsheet can implement this structure, but keep one metric definition table, one fact table at stable grain, one scorecard output, and one action log rather than copying formulas across many supplier tabs.
电子表格可以实现此结构，但应分别保留指标定义表、稳定粒度事实表、评分卡输出与行动日志，而不是把公式复制到大量供应商工作表。
Choose supplier scorecard KPIs that support decisions选择能够支持决策的供应商评分卡 KPI
| Dimension维度 | Possible KPI可能 KPI | Control question控制问题 |
|---|---|---|
| Quality质量 | Defect PPM, acceptance, return, repeat finding, corrective-action effectiveness缺陷 PPM、验收率、退货率、重复发现、整改有效性 | What is an inspected unit and who owns attribution?什么是受检单位，谁负责归因？ |
| Delivery交付 | On-time delivery, OTIF, lead-time variance, schedule adherence准时交付、OTIF、提前期差异、计划达成率 | Which promise date, grain, tolerance, and exclusion apply?采用哪个承诺日期、粒度、容差与排除项？ |
| Commercial商务 | Price variance, invoice accuracy, contract adherence, validated savings价格差异、发票准确率、合同履约、已验证节省 | Are currency, baseline, volume, and cost scope comparable?币种、基线、业务量与成本范围是否可比？ |
| Service服务 | Acknowledgment, response, resolution, documentation accuracy确认、响应、解决、文件准确率 | Which timestamps are observed versus subjective?哪些时间戳是客观观察，哪些是主观评价？ |
| Risk and compliance风险与合规 | Required document status, verified incidents, control or disclosure completion必要文件状态、已核实事件、控制或披露完成度 | Should this be a score, warning, or non-negotiable gate?它应是得分、警告还是不可妥协门槛？ |
Use a few decision-relevant KPIs. More metrics create administration, double counting, and conflicting incentives. Tie every KPI to an expectation that the supplier can understand and to evidence the buyer can verify.
使用少量与决策相关的 KPI。指标越多，管理负担、重复计算与激励冲突越大。每个 KPI 都应连接到供应商能理解的期望以及采购方能核实的证据。
Write a metric contract before scoring评分前先建立指标口径合同
For every KPI, record purpose, formula, numerator, denominator, source, event grain, time boundary, supplier hierarchy, unit, direction, target, tolerance, scoring curve, cap, exclusions, missing-data policy, minimum sample, weight, gate, owner, refresh, drill path, and effective dates.
为每个 KPI 记录目的、公式、分子、分母、来源、事件粒度、时间边界、供应商层级、单位、方向、目标、容差、评分曲线、上下限、排除项、缺失数据政策、最低样本、权重、门槛、责任人、刷新、下钻路径与生效日期。
Do not average percentages with different denominators. Recompute from comparable underlying events. A 100% result across two receipts does not carry the same evidence as 98% across 20,000 order lines; expose both performance and evidence strength.
不要平均分母不同的百分比。应从可比底层事件重新计算。两次收货的100%与两万条订单行的98%证据强度不同；应同时披露绩效与证据强度。
Normalize unlike KPIs with published scoring curves用公布的评分曲线标准化不同 KPI
Percentages, counts, currency variance, response time, and audit status cannot be weighted directly. Convert each applicable actual result to a governed common scale, often zero to 100. The curve must define higher-is-better or lower-is-better direction, target, stretch, floor, caps, boundary values, and rounding.
百分比、计数、货币差异、响应时长与审计状态不能直接加权。应把每项适用实际结果转换到受治理的共同尺度，通常为0–100。曲线必须定义正向或反向方向、目标、挑战值、底线、上下限、边界值与取整。
Linear curves are not always appropriate. A one-day delay may have little effect inside tolerance but severe effect for a production-critical delivery. Use bands or nonlinear curves when the business impact is nonlinear, and test edge cases before release. Version every change; never rewrite history silently.
线性曲线并非总是合适。一天延迟在容差内可能影响很小，但对生产关键交付可能影响严重。业务影响非线性时应使用区间或非线性曲线，并在上线前测试边界。每次变更都应版本化，不能静默改写历史。
Supplier scorecard weighted formula供应商评分卡加权公式
A weight represents approved decision importance, not the amount of data available or the preference of the loudest reviewer. Avoid double counting: OTIF and on-time delivery may overlap, while defect PPM and return rate may share incidents. Document why each dimension exists and test sensitivity to reasonable weight changes.
权重代表获批的决策重要性，不是可用数据量或声音最大的评审者偏好。避免重复计算：OTIF 与准时交付可能重叠，缺陷 PPM 与退货率可能共享事件。记录每个维度存在的原因，并测试合理权重变化下的敏感性。
Show each weighted contribution beside the total. A score of 88.70 is less useful than knowing delivery created the largest gap. Retain the unrounded calculation for audit and round only the displayed value under a published rule.
在总分旁展示各项加权贡献。仅知道88.70分，不如知道交付造成最大差距有用。审计时保留未取整计算，只按公布规则对显示值取整。
Weighted supplier scorecard calculator供应商加权评分卡计算器
Enter normalized scores and weights. This educational calculator requires weights to total exactly 100% and applies a separate critical-gate result. It does not validate the underlying KPI evidence.
输入标准化得分与权重。本教学计算器要求权重合计恰好为100%，并单独应用关键门槛结果；它不验证底层 KPI 证据。
Worked supplier scorecard example供应商评分卡计算示例
Hypothetical example: normalized quality, delivery, commercial, service, and risk/compliance scores are 94, 86, 80, 88, and 92. Their approved weights are 30%, 25%, 15%, 15%, and 15%. The weighted contributions are 28.20, 21.50, 12.00, 13.20, and 13.80, producing 88.70.
假设示例：标准化后的质量、交付、商务、服务与风险/合规得分分别为94、86、80、88、92；批准权重为30%、25%、15%、15%、15%。加权贡献分别为28.20、21.50、12.00、13.20、13.80，总分为88.70。
Delivery creates the largest weighted gap to a perfect score: (100 − 86) × 25% = 3.50 points. That does not prove root cause or justify reallocating business. Review order-line evidence, tolerances, buyer-caused exceptions, trend, sample size, supplier comment, and operational impact first.
交付产生相对满分的最大加权差距：（100−86）×25%=3.50分。但这并不能证明根因，也不能直接支持调整份额。应先复核订单行证据、容差、买方原因例外、趋势、样本量、供应商评论与运营影响。
Keep non-negotiable requirements outside the average把不可妥协要求置于平均分之外
A weighted average can hide a severe failure. Legal eligibility, product safety, critical quality escape, cybersecurity, sanctions, required certification, business continuity, or unresolved integrity issues may require a hold, escalation, restricted allocation, or specialist review regardless of total points.
加权平均可能掩盖严重失败。法律资格、产品安全、重大质量逃逸、网络安全、制裁、必要认证、业务连续性或未解决诚信问题，可能无论总分多少都需要暂停、升级、限制份额或专业复核。
A failed gate is not automatically proof of supplier fault. Preserve the requirement, evidence, event date, source, reviewer, supplier response, exception authority, temporary control, due date, and resolution. Do not silently convert a failed gate into a small score deduction.
门槛未通过不自动证明供应商有过错。保留要求、证据、事件日期、来源、评审人、供应商回应、例外授权、临时控制、到期日与解决情况。不能把门槛失败无痕转换为小幅扣分。
Handle missing data and small samples explicitly明确处理缺失数据与小样本
Classify missing results as not applicable, not yet due, source unavailable, supplier submission missing, data-quality exception, or unresolved dispute. These states have different meanings. Do not default to zero, carry forward an old score indefinitely, or renormalize remaining weights without a published policy.
把缺失结果分类为不适用、尚未到期、来源不可用、供应商未提交、数据质量例外或争议未解决。这些状态含义不同。不要默认填零、无限期沿用旧分数，也不要在没有公布政策时重标剩余权重。
Show event counts and denominator exposure. Minimum samples may suppress a score, widen a confidence category, or require narrative review. Separate observed performance from evidence confidence so a supplier cannot appear excellent merely because data is sparse.
显示事件数量与分母暴露。最低样本规则可以抑制得分、降低置信类别或要求叙述性复核。把已观察绩效与证据置信度分开，避免供应商仅因数据稀少而显得优秀。
Adapt the template by category and criticality按品类与关键性调整模板
Do not use identical scorecards for direct materials, logistics, contingent labor, facilities, software, and professional services. Their deliverables, risk, evidence, and decision cadence differ. Build approved template families with a common core and category-specific modules.
不要对直接材料、物流、临时用工、设施、软件与专业服务使用相同评分卡。它们的交付物、风险、证据与决策频率不同。应建立带共同核心和品类专用模块的获批模板族。
- Strategic or critical: deeper evidence, frequent review, executive visibility, explicit gates, and action effectiveness.战略或关键：更深证据、更高频评审、高管可见、明确门槛与行动有效性。
- Leverage: consistent commercial, quality, service, and delivery comparison across qualified alternatives.杠杆型：在合格备选之间一致比较商务、质量、服务与交付。
- Routine: compact exception-led scorecard with proportionate cadence and administration.常规型：紧凑的例外驱动评分卡，配合适度频率与管理投入。
Review the evidence with the supplier与供应商共同复核证据
Freeze the period and metric version before review. Distribute results with enough time for verification. In the meeting, separate arithmetic, data quality, attribution, business impact, trend, and action. A supplier disagreement should create a tracked dispute with evidence and resolution authority, not an offline spreadsheet edit.
评审前冻结期间与指标版本，并提前分发结果供核实。会议中区分算术、数据质量、归因、业务影响、趋势与行动。供应商异议应形成带证据和解决权限的可跟踪争议，而不是线下修改表格。
Set cadence by criticality, spend, dependency, volatility, contract, and open actions. Use event-triggered review for severe incidents, gate failures, major deterioration, or material data restatement. Keep approval, comment, override reason, and the supplier's response in the audit trail.
按关键性、支出、依赖、波动、合同与未关闭行动设定频率。严重事件、门槛失败、重大恶化或实质数据重述应触发事件评审。审批、评论、覆盖原因与供应商回应都应进入审计轨迹。
Connect score bands to proportionate action把评分区间连接到适度行动
Bands are organization-specific. An illustrative policy might distinguish maintain, monitor, improvement required, and escalation; it must also consider gates, confidence, trend, criticality, and contract rights. Never treat the numeric boundary as a substitute for judgment or due process.
区间因组织而异。示例政策可以区分维持、监控、需要改善与升级，但还必须考虑门槛、置信度、趋势、关键性与合同权利。不能用数字边界替代判断或正当流程。
No failed gates; evidence is sufficient; continue monitoring and recognize verified strengths.
无门槛失败，证据充分；持续监控并认可已验证优势。
Investigate material decline, weak coverage, or one dimension that conflicts with the total.
调查重大下降、覆盖不足或与总分冲突的单项维度。
Agree cause, owner, action, evidence, milestone, due date, and effectiveness test.
约定原因、责任人、行动、证据、里程碑、到期日与有效性测试。
Apply approved governance for failed gates, severe incidents, repeat failure, or unacceptable exposure.
对门槛失败、严重事件、重复失败或不可接受暴露应用获批治理流程。
Choose spreadsheet or supplier scorecard software deliberately有意识地选择表格或供应商评分卡软件
| Option选项 | Useful when适用情况 | Watch for注意事项 |
|---|---|---|
| Controlled spreadsheet受控表格 | Small scope, stable data, few owners, low workflow complexity范围小、数据稳定、责任人少、工作流复杂度低 | Formula copies, version drift, permissions, evidence links, concurrency, audit公式复制、版本漂移、权限、证据链接、并发与审计 |
| BI or analytics layerBI 或分析层 | Multiple sources, repeatable metrics, portfolio analysis, evidence drill-through多来源、可重复指标、组合分析、证据下钻 | Action workflow, supplier access, definition ownership, writeback boundaries行动工作流、供应商访问、口径责任、回写边界 |
| SPM/SRM suiteSPM/SRM 套件 | Portal, approvals, surveys, reviews, corrective actions, documents, roles门户、审批、调查、评审、整改、文件与角色 | Integration depth, metric transparency, adoption, administration, total cost, exit集成深度、指标透明度、采用、管理、总成本与退出 |
Evaluate connectors, entity hierarchy, metric catalog, scoring curves, gates, missing-data rules, period locking, source drilldown, dispute and approval support, role and row security, audit trail, exports, API, monitoring, recovery, implementation, licensing, and exit cost. Confirm which capabilities are native versus custom.
评估连接器、实体层级、指标目录、评分曲线、门槛、缺失数据规则、期间锁定、源证据下钻、争议与审批支持、角色与行级安全、审计轨迹、导出、API、监控、恢复、实施、许可与退出成本。确认哪些能力原生提供，哪些需要定制。
Validate the scorecard with frozen historical periods用冻结历史期间验证评分卡
Select representative suppliers: strategic, routine, high-volume, low-volume, multi-site, service, direct material, disputed, improving, deteriorating, failed-gate, and incomplete-data cases. Freeze source extracts, metric versions, weights, gates, and expected outcomes before the test.
选择有代表性的供应商：战略、常规、高量、低量、多站点、服务、直接材料、有争议、改善、恶化、门槛失败与数据不完整案例。测试前冻结源提取、指标版本、权重、门槛与预期结果。
- Reconciliation: reproduce counts, denominators, actuals, scores, weights, contributions, gates, and totals.核对：复现数量、分母、实际值、得分、权重、贡献、门槛与总分。
- Edge cases: test zero denominators, missing events, weight errors, exact boundaries, caps, late data, and restatement.边界案例：测试零分母、缺失事件、权重错误、精确边界、上下限、迟到数据与重述。
- Decision replay: confirm whether known historical reviews, disputes, gates, and actions are represented accurately without future information.决策重放：确认已知历史评审、争议、门槛与行动是否在不使用未来信息时得到准确表达。
- Operation: measure analyst time, supplier review effort, permissions, audit, exception precision, administration, and total cost.运营：衡量分析时间、供应商复核工作量、权限、审计、异常精度、管理与总成本。
Common supplier scorecard mistakes供应商评分卡常见错误
- Using one template, target, weight set, and cadence for every supplier category.对所有供应商品类使用同一模板、目标、权重与频率。
- Mixing legal entity, site, category, contract, and item-family evidence in one row.在一行中混合法人、站点、品类、合同与物料族证据。
- Weighting raw KPIs with incompatible scales instead of normalizing them.未标准化不兼容尺度的原始 KPI 就直接加权。
- Allowing a serious compliance, safety, or quality failure to be averaged away.让严重合规、安全或质量失败被平均分掩盖。
- Treating missing or disputed evidence as zero without a published rule.在没有公布规则时把缺失或争议证据视为零。
- Publishing scores without sample size, data coverage, version, or source drilldown.发布得分却不显示样本量、数据覆盖、版本或源证据下钻。
- Changing formulas or weights mid-period and silently rewriting history.期间中途改变公式或权重并静默改写历史。
- Triggering allocation, payment, or termination automatically from the total score.仅按总分自动触发份额、付款或终止决定。
Where InfiniSynapse can support scorecard analysisInfiniSynapse 可以支持哪些评分卡分析
InfiniSynapse can be considered as an analysis layer for exploring authorized supplier, site, category, contract, purchase order, receipt, delivery, quality, defect, return, invoice, service, risk, compliance, scorecard, note, dispute, action, and supporting-document evidence. Analysts can reconcile sources, calculate governed views, compare cohorts and periods, inspect outliers, and prepare traceable review summaries.
InfiniSynapse 可作为分析层，探索获准使用的供应商、站点、品类、合同、采购订单、收货、交付、质量、缺陷、退货、发票、服务、风险、合规、评分卡、备注、争议、行动与支持文件证据。分析人员可核对来源、计算受治理视图、比较分组与期间、检查异常并准备可追溯评审摘要。
Boundary: InfiniSynapse is not presented here as a supplier portal, supplier onboarding or qualification system, procurement execution suite, survey collector, approval workflow, corrective-action management system, audit platform, autonomous allocation or termination engine, operational system of record, or ERP transaction/writeback tool.
边界：本文不把 InfiniSynapse 描述为供应商门户、供应商准入或资质系统、采购执行套件、问卷收集器、审批工作流、整改管理系统、审计平台、自动份额或终止引擎、运营记录系统或 ERP 交易/回写工具。
Prepare a governed supplier scorecard analysis准备一项有治理依据的供应商评分卡分析
Bring supplier hierarchies, category and contract scope, metric definitions, targets, scoring curves, weights, gates, transaction extracts, delivery and quality events, commercial and service records, evidence links, disputes, actions, and historical outcomes. Use InfiniSynapse to explore the connected evidence and determine where analysis can support your review process.
准备供应商层级、品类与合同范围、指标定义、目标、评分曲线、权重、门槛、交易提取、交付与质量事件、商务与服务记录、证据链接、争议、行动与历史结果。使用 InfiniSynapse 探索关联证据，并判断分析可以如何支持评审流程。
Try InfiniSynapse Online在线体验 InfiniSynapseSupplier scorecard FAQ供应商评分卡常见问题
It converts governed supplier performance evidence into comparable KPI results, normalized scores, trends, gates, and review actions for a defined scope and period.
它针对明确范围与期间，把受治理的供应商绩效证据转换为可比较 KPI 结果、标准化得分、趋势、门槛与评审行动。
Common dimensions are quality, delivery, commercial performance, service, and applicable risk or compliance evidence, adapted to category and criticality.
常见维度为质量、交付、商务绩效、服务及适用风险或合规证据，并按品类与关键性调整。
Normalize applicable KPIs, multiply each score by its decimal weight, and sum the products; require weights to total 100% and keep hard gates separate.
标准化适用 KPI，将各得分乘以小数权重后求和；要求权重合计100%，并把硬门槛独立处理。
Include scope, period, formula version, target, actual, numerator, denominator, score, weight, contribution, gate, sample, coverage, source, owner, dispute, action, approval, and outcome.
包括范围、期间、公式版本、目标、实际值、分子、分母、得分、权重、贡献、门槛、样本、覆盖、来源、责任人、争议、行动、审批与结果。
Classify the missing reason, apply a published policy, show coverage, retain evidence, and obtain approval; do not silently enter zero or renormalize weights.
分类缺失原因，应用公布政策，显示覆盖并保留证据及审批；不要静默填零或重标权重。
Set cadence by criticality, spend, dependency, volatility, contract, and action urgency, with event-triggered review for severe changes or failed gates.
按关键性、支出、依赖、波动、合同与行动紧急度设定，并对严重变化或门槛失败触发事件评审。
No. It is positioned here as an analysis layer, not a supplier portal, procurement suite, approval workflow, corrective-action system, audit platform, or ERP writeback tool.
不是。这里将其定位为分析层，而不是供应商门户、采购套件、审批工作流、整改系统、审计平台或 ERP 回写工具。
Sources and limitations来源与局限
Primary references: the ISO 9001 Auditing Practices Group guidance on external providers, which discusses defined criteria and effective provider performance monitoring; U.S. Federal Acquisition Regulation Subpart 42.15, which requires objective performance facts, roles, narratives, and factors such as quality, cost control, schedule, and management; and a NIST Baldrige application summary describing monthly supplier scorecard use for quality, delivery, cost, and service. Sources reviewed August 18, 2026.
主要参考：ISO 9001 审核实践组外部供方指南，其中讨论已定义准则与有效供方绩效监控；美国联邦采购法规第42.15分部，要求客观绩效事实、角色、叙述以及质量、成本控制、进度与管理等因素；以及描述按月使用质量、交付、成本和服务供应商评分卡的 NIST Baldrige 申请摘要。来源核验于2026年8月18日。
The dimensions, formulas, scoring curve, example, bands, and template fields are educational starting points, not universal policy, a supplier qualification, risk acceptance, contract determination, allocation instruction, payment decision, termination recommendation, or legal conclusion. Validate the design with responsible procurement, quality, operations, finance, risk, compliance, legal, security, data, and supplier owners.
本文维度、公式、评分曲线、示例、区间与模板字段是教学起点，不是通用政策、供应商资质结论、风险接受、合同决定、份额指令、付款决定、终止建议或法律结论。应由负责采购、质量、运营、财务、风险、合规、法务、安全、数据与供应商的人员验证设计。