全球采购供应商评估模型：从价格、质量到供应链稳定性的完整方法

有效的供应商评估不是寻找“平均分最高”的企业，而是找到最适合特定产品、市场和风险水平的合作伙伴。

一个完整模型至少应覆盖六个维度：

1. 总采购成本；
2. 产品与质量；
3. 交付与产能；
4. 合规能力；
5. 供应链稳定性；
6. 工程与服务。

EY 2025全球CPO调查显示，采购负责人短期仍然重视价值与节省，同时关注供应商韧性和供应商绩效。Deloitte 2025全球CPO调查则显示，采购数字化与风险管理的重要性继续上升。

这意味着价格、绩效与风险应进入同一套决策框架。

一、先按采购场景设定权重

以下是一套适用于一般工业品的起始权重，不是统一行业标准。

权重应随采购场景改变：

• 标准化低风险产品：提高成本权重；
• 季节性产品：提高交付权重；
• 关键工业零部件：提高质量和韧性权重；
• 受监管或安全相关产品：提高合规和变更管理权重；
• 新产品开发：提高工程能力权重。

二、总采购成本：不要只看产品单价

总采购成本通常包括：

• 产品价格；
• 模具、样品与开发费；
• 包装、运输与保险；
• 关税和进口费用；
• 检验与第三方审核；
• 库存资金占用；
• 返工、报废与退货；
• 延期、缺货与停产；
• 售后、备件与召回；
• 更换供应商的重新验证成本。

建议使用：

风险调整后总成本 = 可确定成本 + 各类风险发生概率 × 对应损失

概率难以精确计算时，可设置低、中、高三个情景，比较结论是否会发生变化。

三、产品与质量：验证稳定量产，而不是只验证样品

重点核验：

• 目标材料、工艺和设备是否匹配；
• 样品与量产工艺是否一致；
• 来料、首件、过程和成品检验；
• 测量设备校准；
• 关键参数和批次追溯；
• 不合格品隔离；
• 客户投诉和纠正措施；
• 工程变更审批。

质量评分应优先使用真实订单记录、审核结果和趋势数据，不能只依据体系证书。

四、交付与产能：区分承诺与证据

供应商应提供：

• 标准交期及起算条件；
• 月度标准产能；
• 当前产能利用率；
• 最大可扩展产能；
• 过去12个月准时交付率；
• 旺季排产策略；
• 紧急订单处理方式；
• 关键设备与人员瓶颈。

交付率必须注明统计周期、计算口径和排除条件，否则不同供应商的数据无法比较。

五、合规能力：确认“适用于这个产品”

核验内容包括：

• 目标市场适用法规；
• 认证与测试报告范围；
• 具体型号和生产地址；
• 标签、警示与说明书；
• 材料和化学物质要求；
• 原产地与海关编码；
• 技术文件与版本管理；
• 产品责任、退货与召回机制。

企业体系认证、产品认证与客户采购标准应分别评分，不能用其中一项替代另外两项。

六、供应链稳定性：把抽象风险变成指标

供应链韧性不是“永不发生中断”，而是更早发现、更快响应和更快恢复。

七、工程与服务：判断长期合作价值

根据产品特点评估：

• 可制造性分析；
• 打样与试产；
• 材料和工艺优化；
• 成本改善；
• 项目管理；
• 技术响应；
• 备件与售后；
• 海外仓或本地服务；
• 产品生命周期与停产管理。

标准品采购不必过度提高工程服务权重。高度定制或验证周期长的产品则应重点评估。

八、评分与证据等级

每项按1—5分：

• 1分：明显不符合；
• 2分：存在较大风险；
• 3分：基本符合；
• 4分：表现良好；
• 5分：有完整证据证明表现优秀。

供应商加权总分 = 各项得分 × 对应权重之和

同时给证据分级：

同等得分时，应优先选择证据等级更高的供应商。

九、设置不可加权的淘汰项

以下问题不应通过其他高分抵消：

• 企业身份、生产地点或收款主体虚假；
• 关键认证或报告无效；
• 样品与量产材料、工艺明显不一致；
• 关键产品无法追溯；
• 隐瞒重大质量或交付事件；
• 知识产权、模具或图纸归属不清；
• 拒绝审核关键生产和外包环节；
• 违反商业道德或反贿赂要求。

十、从一次评分升级为动态管理

供应商评分不应在准入后停止。建议：

• 月度跟踪质量、交付与响应；
• 季度更新风险、产能和第二来源；
• 年度复核认证、财务和合规；
• 重大变更后重新评估；
• 对关键供应商制定联合改善计划；
• 对备用供应商保持定期样品或小批量验证。

常见问题

是否应该始终选择总分最高的供应商？

不一定。还要考虑淘汰项、证据等级、供应商组合、切换成本和谈判条件。

两家供应商得分接近时如何选择？

优先比较证据质量、关键风险、改善意愿和产品长期需求。也可以通过小批量试产进一步验证。

评分权重多久调整一次？

产品、市场、法规或供应风险发生变化时应立即调整；稳定品类至少每年复核一次。

可以让供应商自评吗？

可以用于资料收集，但自评分不能作为最终结论。采购、质量、工程和合规团队应分别验证关键证据。

结论

供应商评估的价值不在于得到一个看似精确的分数，而在于把价格、质量、交付、合规和供应风险放到同一框架中，使决策依据透明、可比较、可复核。

参考来源

• EY：Global CPO Survey 2025 Outlook
• Deloitte：2025 Global Chief Procurement Officer Survey

Effective supplier evaluation is not about finding the company with the "highest average score," but about finding the best partner for a specific product, market, and risk level.

A complete model should cover at least six dimensions:

1. Total Procurement Cost;
2. Product and Quality;
3. Delivery and Capacity;
4. Compliance Capabilities;
5. Supply Chain Stability;
6. Engineering and Services.

The EY 2025 Global CPO Survey shows that procurement leaders still prioritize value and savings in the short term, while also focusing on supplier resilience and performance. The Deloitte 2025 Global CPO Survey (https://www.deloitte.com/ch/en/services/consulting/research/procurement-strategy.html) shows that the importance of procurement digitalization and risk management continues to rise.

This means that price, performance, and risk should be integrated into a unified decision-making framework.

I. First, set weights according to the procurement scenario

The following is a set of initial weights applicable to general industrial products, not a uniform industry standard.

Weights should change according to the procurement scenario:

• Standardized, low-risk products: Increase cost weight;
• Seasonal products: Increase delivery weight;
• Critical industrial components: Increase quality and resilience weight;
• Regulated or safety-related products: Increase compliance and change management weight;
• New product development: Increase engineering capability weight.

II. Total Procurement Cost: Don't Just Look at the Unit Price

Total procurement cost typically includes:

• Product price;
• Mold, sample, and development costs;
• Packaging, transportation, and insurance;
• Customs duties and import fees;
• Inspection and third-party audits;
• Inventory capital tied up;
• Rework, scrap, and returns;
• Delays, stockouts, and production stoppages;
• After-sales service, spare parts, and recalls;
• Re-verification costs for changing suppliers.

Recommended method:

Risk-adjusted total cost = Determinable costs + Probability of occurrence of each risk × Corresponding loss

When probabilities are difficult to calculate precisely, set up low, medium, and high scenarios and compare whether the conclusion will change.

III. Product and Quality: Verify Stable Mass Production, Not Just Samples

Key Verification Areas:

• Matching of target materials, processes, and equipment;
• Consistency between sample and mass production processes;
• Incoming material, first article, in-process, and finished product inspections;
• Calibration of measuring equipment;
• Traceability of key parameters and batches;
• Isolation of nonconforming products;
• Customer complaints and corrective actions;
• Engineering change approval.

Quality scoring should prioritize the use of actual order records, audit results, and trend data, and should not rely solely on system certificates.

IV. Delivery and Capacity: Distinguishing Between Commitments and Evidence

Suppliers should provide:

• Standard delivery dates and starting conditions;
• Monthly standard capacity;
• Current capacity utilization rate;
• Maximum scalable capacity;
• On-time delivery rate over the past 12 months;
• Peak season production scheduling strategy;
• Emergency order handling methods;
• Key equipment and personnel bottlenecks.

Delivery rates must specify the statistical period, calculation method, and exclusion conditions; otherwise, data from different suppliers cannot be compared. ## V. Compliance Capability: Confirming "Applicability to this Product"

Verification includes:

• Applicable regulations in the target market;
• Scope of certification and testing reports;
• Specific model and manufacturing address;
• Labels, warnings, and instructions;
• Material and chemical substance requirements;
• Country of origin and customs code;
• Technical documentation and version management;
• Product liability, return, and recall mechanisms.

Enterprise system certification, product certification, and customer procurement standards should be scored separately; one cannot substitute for the other two.

VI. Supply Chain Stability: Transforming Abstract Risks into Metrics

Supply chain resilience is not about "never experiencing interruptions," but about earlier detection, faster response, and faster recovery.

VII. Engineering and Services: Assessing Long-Term Cooperation Value

Evaluate based on product characteristics:

• Manufacturability analysis;
• Prototyping and pilot production;
• Material and process optimization;
• Cost improvement;
• Project management;
• Technical response;
• Spare parts and after-sales service;
• Overseas warehousing or local service;
• Product lifecycle and discontinuation management.

For standard product procurement, engineering services should not be overemphasized. Highly customized products or those with long validation cycles should be given priority in evaluation.

VIII. Scoring and Level of Evidence

Each item is scored from 1 to 5:

• 1 point: Clearly non-compliant;
• 2 points: Significant risk exists;
• 3 points: Basically compliant;
• 4 points: Good performance;
• 5 points: Excellent performance with complete evidence.

Supplier Weighted Total Score = Score of Each Item × Sum of Corresponding Weights

Evidence is also graded as follows:

In case of equal scores, suppliers with higher evidence levels should be prioritized.

IX. Setting Non-Weighted Elimination Items

The following issues should not be offset by other high scores:

• False company identity, production location, or payee;
• Invalid key certifications or reports;
• Significant inconsistencies between sample and mass-production materials and processes;
• Untraceable key products;
• Concealment of major quality or delivery incidents;
• Unclear ownership of intellectual property, molds, or drawings;
• Refusal to audit key production and outsourcing processes;
• Violation of business ethics or anti-bribery requirements.

X. Upgrading from One-Time Scoring to Dynamic Management

Supplier scoring should not stop after approval. Recommendations:

• Monthly tracking of quality, delivery, and responsiveness;
• Quarterly updates on risk, capacity, and secondary sources;
• Annual review of certification, finance, and compliance;
• Reassessment after significant changes;
• Develop joint improvement plans for key suppliers;
• Regular sample or small-batch verification with backup suppliers.

Frequently Asked Questions

Should the supplier with the highest total score always be selected?

Not necessarily. Other factors to consider include elimination criteria, evidence level, supplier mix, switching costs, and negotiation terms.

How to choose when two suppliers have similar scores?

Prioritize comparing evidence quality, key risks, willingness to improve, and long-term product demand. Further verification can be done through small-batch trial production.

How often should scoring weights be adjusted?

Adjustments should be made immediately when product, market, regulatory, or supply risks change; stable product categories should be reviewed at least annually.

Can suppliers self-assess?

It can be used for data collection, but self-assessments should not be considered final conclusions. The purchasing, quality, engineering, and compliance teams should each verify key evidence separately.

Conclusion

The value of supplier evaluation lies not in obtaining a seemingly precise score, but in placing price, quality, delivery, compliance, and supply risks within the same framework, making the decision-making process transparent, comparable, and verifiable.

References

• EY: Global CPO Survey 2025 Outlook
• Deloitte: 2025 Global Chief Procurement Officer Survey