InfiniSynapse
What is spend analysis?什么是支出分析？
Spend analysis is the structured process of collecting, cleansing, normalizing, classifying, and analyzing expenditure data so an organization can understand what it buys, from whom, through which business unit, at what price, and under which commercial conditions.
支出分析是系统收集、清洗、归一、分类并分析支出数据的过程，帮助组织理解购买了什么、向谁购买、由哪个业务单元发起、支付什么价格以及适用哪些商业条件。
The output is not merely a chart of total spend. A useful analysis creates a traceable decision base for category strategy, supplier consolidation, negotiation, compliance review, budgeting, payment terms, risk investigation, and demand management. It should reveal the underlying transactions and data-quality limitations behind every summary.
结果并非只有总支出图表。有效分析应为品类战略、供应商整合、谈判、合规审查、预算、付款条款、风险调查与需求管理提供可追溯的决策基础，并能够从汇总结果回溯到底层交易与数据质量限制。
Spend analysis is narrower than procurement analytics, which may also examine sourcing cycles, supplier performance, contract milestones, risk, requisition-to-order time, and operational outcomes. It is also different from spend management: analysis informs decisions; management includes policies, workflows, approvals, and execution.
支出分析比采购分析更聚焦，后者还可能研究寻源周期、供应商绩效、合同节点、风险、请购到下单时间及运营结果。它也不同于支出管理：分析为决策提供依据，而管理还包含政策、流程、审批与执行。
The business questions a spend analysis should answer支出分析应该回答哪些业务问题？
Spend by category, subcategory, item or service, period, quantity, and price.
按品类、子品类、商品或服务、期间、数量和价格分析支出。
Supplier families, duplicates, concentration, fragmentation, terms, and cross-category relationships.
供应商集团、重复主体、集中度、碎片化、条款及跨品类关系。
Business unit, cost center, buyer, channel, PO coverage, contract usage, and approval path.
业务单元、成本中心、采购员、渠道、PO 覆盖、合同使用与审批路径。
Consolidation, sourcing, price harmonization, compliance, term improvement, demand control, and risk review.
整合、寻源、价格统一、合规、条款改善、需求控制与风险复核。
A chart may suggest an opportunity but cannot prove savings by itself. For example, fragmented spend across suppliers may reflect unnecessary duplication—or necessary geographic coverage, technical specialization, resilience, or minority-supplier objectives. Analysis must preserve enough context for category owners and business stakeholders to validate the interpretation.
图表可以提示机会，却不能单独证明节省。例如，供应商分散可能意味着不必要的重复，也可能源于必要的地域覆盖、技术专长、韧性或多元供应商目标。分析必须保留足够背景，让品类负责人和业务利益相关者验证解释。
Build the data foundation and spend cube建立数据基础与支出立方体
Start with the decision scope: legal entities, regions, business units, source systems, spend types, currencies, and reporting period. Purchase orders provide requisition, buyer, item, quantity, and negotiated-price detail; invoices and accounts payable capture actual payments, including non-PO spend; purchasing-card and expense data expose channels often missed by ERP procurement tables. Contracts, supplier masters, commodity codes, and organizational hierarchies enrich the transactions.
先定义决策范围：法人、地区、业务单元、源系统、支出类型、币种与报告期间。采购订单提供请购、采购员、物料、数量及议价信息；发票和应付账款反映实际付款，包括无 PO 支出；采购卡与费用数据揭示 ERP 采购表经常遗漏的渠道。合同、供应商主数据、商品编码和组织层级用于丰富交易。
| Dimension维度 | Example fields示例字段 | Control question控制问题 |
|---|---|---|
| Transaction交易 | Document ID, line ID, date, quantity, amount, currency, tax, credit flag单据 ID、行 ID、日期、数量、金额、币种、税额、贷项标记 | Can duplicates and reversals be detected?能否识别重复与冲销？ |
| Supplier供应商 | Raw name, supplier ID, normalized parent, location, status原始名称、供应商 ID、归一集团、地点、状态 | Are aliases and subsidiaries consolidated appropriately?别名和子公司是否被正确合并？ |
| Category品类 | Item, description, internal code, taxonomy level, confidence物料、描述、内部编码、分类层级、置信度 | Can a reviewer explain the assigned category?审查者能否解释分类依据？ |
| Organization组织 | Entity, business unit, site, cost center, requester, buyer法人、业务单元、地点、成本中心、申请人、采购员 | Does ownership follow the actual buying decision?责任归属是否对应实际采购决策？ |
| Commercial商业 | PO, contract ID, price, payment term, incoterm, sourcing eventPO、合同 ID、价格、付款条款、贸易术语、寻源事件 | Can compliance be tested at line level?能否在行级检验合规？ |
The classic spend cube connects supplier, category, and organizational unit, with amount or quantity as the measure and time as an additional analytical axis. Keep the detailed transaction fact table beneath the cube; prematurely aggregating invoices removes the evidence needed to correct classification, validate contract use, or explain anomalies.
经典支出立方体连接供应商、品类与组织单元，以金额或数量为度量，并把时间作为额外分析轴。立方体下方必须保留交易事实明细；过早汇总发票会丢失修正分类、验证合同使用或解释异常所需的证据。
A repeatable spend analysis process可重复执行的支出分析流程
- Scope and inventory sources.定义范围并盘点数据源。 Agree entities, periods, spend types, exclusions, currency rules, owners, and decisions the analysis must support.确定法人、期间、支出类型、排除项、汇率规则、负责人及分析需支持的决策。
- Extract and reconcile.提取并核对。 Load PO, invoice, AP, card, expense, contract, and master data; reconcile totals to finance control accounts before analysis.加载 PO、发票、AP、采购卡、费用、合同和主数据；分析前与财务控制账户核对总额。
- Clean and normalize.清洗并归一。 Standardize dates, currencies, signs, units, tax treatment, supplier identities, and organizational mappings while preserving raw values.标准化日期、币种、正负号、单位、税务口径、供应商身份与组织映射，同时保留原始值。
- Classify with review controls.分类并设置复核控制。 Apply an agreed taxonomy, rule precedence, confidence levels, and an exception queue for ambiguous descriptions.应用统一分类体系、规则优先级与置信度，并为模糊描述建立例外队列。
- Analyze patterns and hypotheses.分析模式与假设。 Examine supplier, category, unit, channel, price, terms, concentration, compliance, time, and transaction-level outliers.研究供应商、品类、单位、渠道、价格、条款、集中度、合规、时间及交易异常。
- Validate, prioritize, and track action.验证、排序并跟踪行动。 Confirm context with owners, distinguish estimates from validated opportunities, assign actions, and refresh the baseline.与负责人确认背景，区分估算与已验证机会，分配行动并刷新基线。
Clean suppliers, currencies, and transactions before measuring先清洗供应商、币种与交易，再计算指标
Supplier normalization is more than uppercasing names. One supplier may appear under abbreviations, local subsidiaries, remittance addresses, legacy IDs, spelling errors, or factoring entities. Create a crosswalk from raw supplier record to normalized legal entity and, where appropriate, ultimate parent. Keep both levels because consolidation for negotiation may be useful while legal, risk, tax, and performance analysis still require the actual contracting entity.
供应商归一不只是把名称转为大写。同一供应商可能以缩写、当地子公司、汇款地址、旧 ID、拼写错误或保理主体出现。应建立从原始供应商记录到归一法律实体及适用时最终母公司的映射，并保留两个层级，因为谈判整合可能使用集团视图，而法律、风险、税务和绩效分析仍需实际签约主体。
Define a currency policy: transaction currency for price analysis, ledger currency for reconciliation, and a common reporting currency for portfolio totals. State exchange-rate source and date. Decide whether amounts include tax, freight, rebates, credits, and intercompany transfers. Detect duplicate invoice lines using supplier, invoice, date, amount, and reference combinations, but send candidates to review rather than deleting them automatically.
应定义币种政策：价格分析使用交易币种，对账使用账簿币种，组合汇总使用统一报告币种，并说明汇率来源与日期。还需决定金额是否包含税费、运费、返利、贷项和公司间交易。可用供应商、发票号、日期、金额与参考号组合识别疑似重复行，但应送人工复核，而非自动删除。
Reconciliation gate: document the difference between the analytical population and the finance baseline. Unexplained gaps undermine every percentage that follows.
对账门槛：记录分析数据总体与财务基线之间的差异。无法解释的缺口会削弱后续所有百分比的可信度。
Classify spend with a governed taxonomy使用受治理的分类体系划分支出
Choose a category hierarchy that matches procurement decisions. A broad taxonomy can support enterprise reporting, while category teams may need deeper internal levels. Classification can use item codes, GL accounts, supplier-category mappings, contract references, description rules, or reviewed model outputs. Record rule precedence and the evidence behind each result.
选择与采购决策匹配的品类层级。宽泛分类体系可支持企业报告，品类团队则可能需要更深的内部层级。分类可以使用物料编码、总账科目、供应商—品类映射、合同参考、描述规则或经过复核的模型输出，并应记录规则优先级与结果依据。
A supplier-only rule is risky when one vendor sells products across categories. A GL-only rule may be too coarse or reflect accounting rather than sourcing logic. Prefer line-level classification where descriptions and item data are adequate. Report classified spend coverage, unresolved amount, low-confidence amount, and manual overrides separately; “100% classified” is not meaningful if uncertain records were forced into convenient buckets.
当一个供应商跨多个品类销售时，仅按供应商分类风险较高；仅按总账科目可能过粗，且反映的是会计逻辑而非寻源逻辑。若描述和物料数据充分，应优先使用行级分类。分别报告已分类覆盖、未解决金额、低置信度金额和人工覆盖；如果把不确定记录强行塞入方便的类别，“100% 已分类”并无意义。
Spend analysis metrics and formulas支出分析核心指标与公式
| Metric指标 | Formula公式 | Interpretation解读 |
|---|---|---|
| Data coverage数据覆盖率 | Analyzable spend ÷ in-scope spend | How much of the agreed baseline is usable约定基线中有多少可用于分析 |
| Classification coverage分类覆盖率 | Classified spend ÷ analyzable spend | Report confidence and unresolved value with it应同时报告置信度和未解决金额 |
| Spend under management受管理支出 | Managed spend ÷ in-scope spend | Requires a documented definition of “managed”需书面定义“受管理” |
| Contract compliance合同合规率 | Compliant spend ÷ eligible managed spend | Eligibility and contract dates matter适用范围与合同日期很重要 |
| Supplier concentration供应商集中度 | Top-N supplier spend ÷ category or total spend | Can indicate leverage, dependency, or both可能表示议价能力、依赖风险或二者兼有 |
| Validated savings rate已验证节省率 | Finance-approved realized savings ÷ baseline spend | Keep pipeline, negotiated, and realized stages separate应区分机会、谈判与已实现阶段 |
Other useful measures include PO coverage, invoice-without-PO rate, tail-spend share, supplier count per category, single-source exposure, price variance for comparable specifications, payment-term distribution, transaction count, low-value order frequency, and spend by risk or sustainability attribute. Every metric needs a numerator, denominator, population, time period, currency, owner, and refresh timestamp.
其他有用指标包括 PO 覆盖率、无 PO 发票率、尾部支出占比、每品类供应商数、单一来源暴露、可比规格价格差异、付款条款分布、交易数量、低价值订单频率，以及按风险或可持续属性划分的支出。每项指标都需明确分子、分母、总体、期间、币种、负责人和刷新时间。
Spend analysis KPI calculator支出分析指标计算器
Use one currency and one reporting period. Values below are a hypothetical example, not a benchmark. “Compliant spend” must be part of managed spend, and each other component must not exceed total in-scope spend.
请使用统一币种与报告期间。以下数值是假设示例，不是行业基准。“合规支出”必须属于受管理支出，其他组成金额不得超过范围内总支出。
The calculator treats unclassified spend as total minus classified spend and off-contract managed spend as managed minus compliant spend. These are arithmetic views, not automatic diagnoses. Check population eligibility, contract dates, approved exceptions, and classification confidence before labeling a result as non-compliance or an opportunity.
计算器把“总支出减已分类支出”视为未分类金额，把“受管理支出减合规支出”视为受管理范围内的非合同金额。这些只是算术视图，不是自动诊断。将结果标记为不合规或机会前，应检查总体适用性、合同日期、批准例外及分类置信度。
Hypothetical spend analysis example假设支出分析算例
Consider a hypothetical organization with $12 million of in-scope annual spend. It classifies $10.8 million, so classification coverage is 90% and $1.2 million remains unresolved. Procurement defines $8.4 million as under management, equal to 70% of total spend. Of that managed population, $7.2 million is confirmed compliant with an eligible contract: compliance is 85.7%, leaving $1.2 million for exception review.
假设某组织年度范围内支出为 1,200 万美元，其中 1,080 万美元已分类，因此分类覆盖率为 90%，仍有 120 万美元未解决。采购部门把 840 万美元定义为受管理支出，占总支出的 70%。在该受管理总体中，720 万美元确认符合适用合同，合规率为 85.7%，另有 120 万美元需要例外复核。
The top-N suppliers account for $6 million, or 50% of total spend. That concentration invites category-level review, not an automatic conclusion: it could strengthen leverage or create dependency. Validated realized savings of $360,000 equal 3% of baseline spend. Pipeline ideas, quoted price reductions, avoided cost, and realized P&L savings should remain separate stages with finance-approved definitions.
头部 N 家供应商支出为 600 万美元，占总支出 50%。该集中度值得在品类层面复核，但不能自动得出结论：它可能增强议价能力，也可能形成依赖。已验证的实现节省为 36 万美元，相当于基线支出的 3%。机会线索、报价降幅、成本规避和已实现损益节省应作为不同阶段，并使用财务批准的定义。
Turn spend patterns into testable opportunity hypotheses把支出模式转化为可验证的机会假设
Compare fragmented demand, specifications, locations, capacity, risk, and switching cost before aggregating.
整合前比较分散需求、规格、地点、产能、风险与切换成本。
Normalize units, specifications, freight, currency, volume, terms, taxes, and timing before comparing prices.
比较价格前统一单位、规格、运费、币种、数量、条款、税费与时间。
Test valid contracts, eligible users, approved exceptions, catalog availability, and emergency purchases.
检验有效合同、适用用户、批准例外、目录可用性与紧急采购。
Examine specifications, consumption, low-value orders, duplicate activity, payment terms, and channel design.
分析规格、消耗、低价值订单、重复活动、付款条款与渠道设计。
Prioritize opportunities with value, feasibility, timing, service impact, risk, stakeholder readiness, and data confidence. A high theoretical value with weak specifications or no category owner may rank below a smaller, evidence-rich action. Keep an opportunity register with baseline, hypothesis, owner, next validation step, status, value type, approval, and realized outcome.
应按价值、可行性、时间、服务影响、风险、利益相关者准备度和数据置信度排序。理论价值很高但规格薄弱或没有品类负责人的机会，可能应排在较小但证据充分的行动之后。机会台账应记录基线、假设、负责人、下一验证步骤、状态、价值类型、审批和实际结果。
Common spend analysis errors and limitations常见支出分析错误与限制
- Starting with dashboards before reconciliation: attractive totals can still be incomplete or duplicated.先做仪表板、后做对账：漂亮的总额仍可能不完整或重复。
- Forced classification: hiding low-confidence and unresolved spend creates false certainty.强制分类：隐藏低置信度和未解决支出会制造虚假确定性。
- Comparing unlike prices: unit, specification, freight, tax, location, quantity, currency, and period differences must be controlled.比较不可比价格：必须控制单位、规格、运费、税费、地点、数量、币种和期间差异。
- Calling every gap “savings”: opportunity estimates are not negotiated, implemented, or realized value.把所有差异都称为节省：机会估算不等于已谈判、已实施或已实现价值。
- Ignoring negative and non-PO records: credits, reversals, cards, expenses, and services can materially change the picture.忽略负值与无 PO 记录：贷项、冲销、采购卡、费用和服务可能显著改变结果。
- Using one annual snapshot forever: supplier identities, contracts, prices, categories, and demand change.长期使用单次年度快照：供应商身份、合同、价格、品类和需求都会变化。
Spend analysis explains expenditure patterns but does not by itself diagnose supplier performance, causality, market competitiveness, specification suitability, or operational risk. Combine it with category knowledge, contract review, market intelligence, supplier data, and stakeholder validation before changing sourcing or controls.
支出分析解释支出模式，但无法单独诊断供应商绩效、因果关系、市场竞争力、规格适配性或运营风险。在改变寻源或控制措施前，应结合品类知识、合同审查、市场情报、供应商数据与利益相关者验证。
Govern, validate, and refresh the analysis治理、验证并持续刷新分析
Assign owners for source extraction, finance reconciliation, supplier normalization, taxonomy, contract logic, category interpretation, savings validation, and dashboard publication. Version the supplier crosswalk and category rules. Record source periods, load timestamps, row counts, exclusions, exchange rates, classification coverage, unresolved value, and sign-off status in every release.
应明确源数据提取、财务对账、供应商归一、分类体系、合同逻辑、品类解读、节省验证与仪表板发布的负责人。对供应商映射和分类规则进行版本管理，并在每次发布中记录数据期间、加载时间、行数、排除项、汇率、分类覆盖、未解决金额及批准状态。
Refresh frequency should match decisions. Monthly updates support operational monitoring; quarterly category reviews allow deeper validation; volatile commodities or urgent risk may require faster cadence. Track changes caused by new activity separately from changes caused by revised mappings. A focused data dashboard should expose definitions and drill-through rather than presenting isolated percentages.
刷新频率应与决策匹配。月度更新支持运营监控，季度品类审查允许更深验证，波动品类或紧急风险可能需要更快频率。应区分新业务活动引起的变化与映射修订引起的变化。聚焦型数据仪表板应展示定义并支持下钻，而不是只呈现孤立百分比。
Explore spend patterns across connected procurement and finance data跨关联采购与财务数据探索支出模式
Prepare purchase orders, invoices, accounts-payable extracts, card and expense records, supplier masters, contracts, category mappings, organizational hierarchies, currency rules, and documented exclusions. InfiniSynapse can be used as an analysis layer to explore connected databases and files, compare suppliers and categories, investigate transaction-level exceptions, and support evidence-based review.
请准备采购订单、发票、应付账款导出、采购卡与费用记录、供应商主数据、合同、品类映射、组织层级、币种规则及书面排除项。InfiniSynapse 可作为分析层，探索关联数据库与文件、比较供应商和品类、调查交易级例外，并支持基于证据的复核。
Capability boundary: this page does not present InfiniSynapse as a source-to-pay or procurement execution system. It does not automatically classify or approve every transaction, create requisitions or purchase orders, run sourcing events, negotiate with suppliers, manage contracts, approve savings, release payments, or write transactions back to ERP.
能力边界：本页不把 InfiniSynapse 描述为从寻源到付款或采购执行系统。它不会自动分类或批准每笔交易、创建请购或采购订单、运行寻源活动、与供应商谈判、管理合同、批准节省、释放付款或向 ERP 回写交易。
Open the InfiniSynapse analytics workspace打开 InfiniSynapse 在线分析工作区Spend analysis FAQ支出分析常见问题
What is spend analysis?什么是支出分析？
It is the structured collection, cleansing, supplier normalization, classification, and analysis of expenditure data to understand what was bought, from whom, by whom, for how much, and where action may be warranted.
它是系统收集、清洗、供应商归一、分类和分析支出数据的过程，用于了解购买了什么、向谁购买、由谁购买、花费多少以及哪里值得采取行动。
What data is needed for spend analysis?支出分析需要哪些数据？
Typical sources include POs, invoices, AP, cards, expenses, contracts, supplier masters, taxonomies, business units, cost centers, currencies, and payment terms.
常见来源包括 PO、发票、应付账款、采购卡、费用、合同、供应商主数据、分类体系、业务单元、成本中心、币种和付款条款。
What are the main spend analysis metrics?主要支出分析指标有哪些？
Useful metrics include data coverage, classified spend, spend under management, contract compliance, off-contract spend, concentration, tail spend, price variance, payment terms, and validated savings.
有用指标包括数据覆盖、已分类支出、受管理支出、合同合规、非合同支出、集中度、尾部支出、价格差异、付款条款和已验证节省。
How often should spend analysis be updated?支出分析应多久更新一次？
Match the refresh to decision cadence and data availability. A practical pattern is monthly operational refreshes and quarterly category reviews, with faster updates for volatile or high-risk categories.
刷新应匹配决策频率与数据可用性。实用做法是月度运营更新加季度品类审查，对波动或高风险品类采用更快更新。
Is spend analysis the same as procurement analytics?支出分析等同于采购分析吗？
No. Spend analysis focuses on expenditure visibility and opportunities; procurement analytics is broader and may include sourcing, supplier performance, cycle time, risk, contracts, and operational outcomes.
不等同。支出分析聚焦支出可视性和机会；采购分析范围更广，还可能包括寻源、供应商绩效、周期时间、风险、合同和运营结果。
Does InfiniSynapse automatically execute procurement actions?InfiniSynapse 会自动执行采购行动吗？
It is positioned here as an analysis layer—not as a source-to-pay system that creates POs, runs sourcing events, approves savings, manages contracts, or writes ERP transactions.
本页将其定位为分析层，而不是创建 PO、运行寻源活动、批准节省、管理合同或回写 ERP 交易的从寻源到付款系统。
Official spend analysis sources and editorial limitations支出分析官方来源与编辑限制
The United States Postal Service describes spend analysis as collecting, refining, and analyzing spend data to understand amount, category, and potential savings in its Conduct Spend Analysis guidance, and provides a structured spend analysis process. The National Institute of Governmental Purchasing publishes a Spend Analysis best-practice standard covering data collection, cleansing, classification, and analysis.
美国邮政署在其支出分析指南中将其描述为收集、整理和分析支出数据，以理解金额、品类及潜在节省，并提供了结构化的支出分析流程。美国政府采购协会发布的支出分析最佳实践标准涵盖数据收集、清洗、分类与分析。
SAP's official Ariba Spend Analysis Reporting Basics illustrates analysis by supplier, commodity, organization, compliance, diversity, and risk. Product documentation describes SAP capabilities, not InfiniSynapse capabilities. This page is educational and does not replace procurement, accounting, finance, tax, legal, audit, data-governance, or supplier-risk advice.
SAP 官方Ariba 支出分析报告基础展示了按供应商、商品、组织、合规、多元化与风险进行分析。该产品文档描述的是 SAP 能力，而非 InfiniSynapse 能力。本页仅供教育参考，不替代采购、会计、财务、税务、法律、审计、数据治理或供应商风险建议。