On this page本页目录
What should a return reason report include?退货原因报告应包含什么？
Include report scope and reason-taxonomy version; raw customer response and mapped reason; question wording and channel; missing, other and free-text coverage; physical condition and disposition kept separate; unique return counts; eligible product/variant denominator where rates are shown; refund value and incremental cost under approved definitions; product/channel segments; cumulative contribution; case-review evidence; data limitations; owner and decision. Never present the largest reason category as a verified root cause.
包含报告范围与原因分类版本；原始客户回答与映射原因；问题措辞与渠道；缺失、其他与自由文本覆盖；分开的实体状态与处置；不重复退货数量；显示比例时的合格商品/变体分母；获批定义下的退款金额与增量成本；商品/渠道细分；累计贡献；案例复核证据；数据局限；负责人和决定。绝不能把最大原因类别呈现为已核验根因。
The downloadable XLSX contains synthetic reason rows, formulas and a sorted report. Reason choices and workflow semantics differ by platform and category. Review the current source system and preserve its raw values before applying the template mapping.
可下载 XLSX 包含模拟原因行、公式与排序报告。原因选项与工作流语义因平台和品类而异。应用模板映射前，应复核当前来源系统并保留原始值。
Keep seven layers separate in the reason report workbook在原因报告工作簿中分开七个层
A reusable resource must preserve raw facts, business definitions and calculated outputs as different layers. This keeps an import refresh from rewriting policy or turning a formula into source evidence.
可复用资源必须把原始事实、业务定义与计算输出作为不同层保留，避免导入刷新改写政策或把公式变成来源证据。
Store, channel, market, timezone, currency, period, event and one declared row grain.店铺、渠道、市场、时区、币种、期间、事件及一个声明行粒度。
Order, line, product, variant, return, refund and shipment keys with source system.带来源系统的订单、商品行、商品、变体、退货、退款与发货键。
Sale, shipment, request, receipt, inspection, disposition and refund remain dated facts.销售、发货、申请、收货、质检、处置与退款保持为带日期事实。
Versioned product, channel, reason, status, currency and cost mappings retain originals.版本化商品、渠道、原因、状态、币种与成本映射保留原值。
Numerator, denominator, population, window, cohort, as-of date and exclusions.分子、分母、总体、窗口、群组、截至日期与排除。
Rates, counts, values and costs link to governed inputs and visible formulas.比例、数量、价值与成本关联受治理输入及可见公式。
Coverage, limitations, owner, reviewer, decision, due date and evidence status.覆盖、局限、负责人、审核人、决定、期限与证据状态。
Keep raw imports read-only. Make mappings and assumptions editable in clearly marked cells, and make outputs formula-driven. A free template does not remove the need to validate source semantics.
保持原始导入只读。在清晰标记单元格中编辑映射与假设，让输出由公式驱动。免费模板不能替代来源语义核验。
Define the reason report workbook contract before adding rows添加行前定义原因报告工作簿契约
Freeze each field name, datatype, allowed values, source, business meaning and missing-value rule. Put identifiers and timestamps ahead of labels. Store money as numeric amount plus ISO currency code; conversion policy belongs in a separate assumption.
固定每个字段名、数据类型、允许值、来源、业务含义与缺失规则。标识与时间戳优先于标签。金额使用数值与 ISO 币种代码；换算政策属于单独假设。
| Field group字段组 | Required contract必需契约 | Control控制 |
|---|---|---|
| Raw response原始回答 | Original code, label, free text, question version, channel and timestamp原始代码、标签、自由文本、问题版本、渠道与时间 | Immutable source evidence不可变来源证据 |
| Mapped reason映射原因 | Canonical group, subreason, mapping version, owner and effective date规范组、子原因、映射版本、负责人与生效日期 | Historical remap table历史重映射表 |
| Observed condition观察状态 | Inspection code, evidence, location, inspector and time质检代码、证据、地点、质检人及时间 | Never overwrite customer reason绝不覆盖客户原因 |
| Disposition/outcome处置/结果 | Restock, refurbish, liquidate, recycle, dispose, refund or replacement重新入库、翻新、清算、回收、处置、退款或替换 | Separate operational/transaction fields分开运营/交易字段 |
| Report metric报告指标 | Unique count, eligible rate, value, cost, coverage and cumulative share不重复数量、合格比例、价值、成本、覆盖与累计占比 | Definition beside every output每个输出旁有定义 |
Set up and refresh the reason report workbook设置并刷新原因报告工作簿
- Freeze taxonomy version固定分类版本
Record source question, choices, mapping owner and effective dates.记录来源问题、选项、映射负责人与生效日期。 - Preserve raw responses保留原始回答
Load codes, labels, text and missingness before applying mappings.应用映射前加载代码、标签、文本与缺失。 - Join outcomes carefully谨慎关联结果
Use stable return-line keys to link condition, disposition, refund and product data.使用稳定退货行键关联状态、处置、退款与商品数据。 - Build reason totals建立原因总额
Deduplicate at the declared grain and show counts before percentages.在声明粒度去重，并在百分比前展示数量。 - Add eligible denominators增加合格分母
Use product/variant shipment cohorts only when reason rates require them.仅在原因比例需要时使用商品/变体发货群组。 - Review cases and decide复核案例并决定
Sample raw evidence, record limitations and assign investigation owners.抽样原始证据、记录局限并指定调查负责人。
The free XLSX includes a Reason Report and editable Reason Data with raw responses, mappings, outcomes and synthetic examples.
免费 XLSX 包含原因报告与可编辑原因数据，含原始回答、映射、结果与模拟示例。
Download reason report XLSX下载原因报告 XLSX ↓Rank reasons by count and cumulative contribution按数量与累计贡献排序原因
Show missing and unmapped responses outside and beside the denominator. Sort counts descending and calculate cumulative share for a Pareto view, but use the chart to prioritize review—not to assert an 80/20 law or verified cause.
在分母外及旁边展示缺失与未映射回答。按数量降序并计算累计占比形成帕累托视图，但图表用于确定复核优先级，而不是断言 80/20 规律或已核验原因。
| Output输出 | Required context所需背景 | What it can support可支持内容 |
|---|---|---|
| Unique return count不重复退货数量 | Declared grain and deduplication key声明粒度与去重键 | Reason workload/volume原因工作量/数量 |
| Reason share原因占比 | Mapped eligible count and published missingness已映射合格数量与发布缺失 | Composition among known reasons已知原因中的结构 |
| Return rate by reason按原因退货率 | Matching eligible shipped product/variant denominator匹配合格已发货商品/变体分母 | Cohort risk with limitations带局限的群组风险 |
| Value/cost contribution价值/成本贡献 | Approved transaction/cost boundary and coverage获批交易/成本边界与覆盖 | Materiality for investigation调查重要性 |
Worked example: a Pareto view selects reviews, not causes示例：帕累托视图选择复核而非根因
A synthetic report has 1,000 mapped returns: 360 “size/fit,” 250 “not as expected,” 180 “damaged,” 120 “changed mind” and 90 “other.” Another 140 responses are missing or unmapped and appear in coverage, not silently excluded.
一份模拟报告有 1,000 条已映射退货：360 条“尺码/版型”、250 条“不符合预期”、180 条“破损”、120 条“改变主意”及 90 条“其他”。另有 140 条缺失或未映射回答，展示在覆盖中而非静默排除。
| Mapped reason映射原因 | Count数量 | Share of mapped已映射占比 | Cumulative累计 |
|---|---|---|---|
| Size / fit尺码 / 版型 | 360 | 36% | 36% |
| Not as expected不符合预期 | 250 | 25% | 61% |
| Damaged破损 | 180 | 18% | 79% |
| Changed mind + other改变主意 + 其他 | 210 | 21% | 100% |
The first three groups contribute 79% of mapped responses, so they can receive case-review capacity. The view does not prove size charts, product content or packaging caused the returns; each mechanism needs corroborating evidence.
前三组占已映射回答的 79%，因此可分配案例复核产能。该视图不能证明尺码表、商品内容或包装造成退货；每个机制都需要印证证据。
All store figures and records in this example are synthetic. They illustrate the method and do not represent InfiniSynapse customer results or industry benchmarks.本示例中的商店数字与记录均为模拟，仅用于说明方法，不代表 InfiniSynapse 客户结果或行业基准。
Read reason mix beside missingness and outcomes结合缺失与结果读取原因结构
Reason mix can shift when question wording, menu order, channel, incentives or mapping changes. Publish taxonomy version and missingness for every comparison. Compare customer reason with inspection condition and disposition without forcing agreement. Use free text and support evidence to refine hypotheses.
原因结构可能因问题措辞、菜单顺序、渠道、激励或映射变化。每个比较都发布分类版本与缺失。比较客户原因、质检状态与处置，但不要强迫一致。使用自由文本与客服证据细化假设。
Counts with the eligible denominator and record coverage.带合格分母与记录覆盖的数量。
One event, one cohort window and mature observation.一个事件、一个群组窗口与成熟观察。
Amounts with currency, accounting boundary and double-count checks.带币种、会计边界与重复计算检查的金额。
Missingness, duplicates, joins, mappings, late events and reviewer state.缺失、重复、关联、映射、迟到事件与审核状态。
A Pareto chart shows concentration among recorded categories. It does not establish causal mechanisms, statistical significance or expected savings.
帕累托图展示已记录类别中的集中度，不建立因果机制、统计显著性或预期节省。
Pass twelve controls before sharing the reason report workbook分享原因报告工作簿前通过十二项控制
- One row grain: the input table states exactly what one row represents.一个行粒度：输入表准确说明每行代表什么。
- Stable keys: identifiers are unique at the declared grain and duplicates are visible.稳定键：标识在声明粒度唯一，重复可见。
- Typed dates: timestamps retain offsets and business dates follow a declared timezone.类型化日期：时间戳保留偏移，业务日期遵循声明时区。
- Currency control: amount and currency are separate; exchange rates include source and date.币种控制：金额与币种分开；汇率包含来源与日期。
- Event separation: requests, receipts, refunds, cancellations and exchanges remain distinct.事件分离：申请、收货、退款、取消与换货保持独立。
- Complete denominator: eligibility, exclusions and coverage are published beside every rate.完整分母：每个比例旁发布资格、排除与覆盖。
- Mature cohorts: compared rows have equal opportunity under one as-of date.成熟群组：比较行在统一截至日期下拥有相同机会。
- Versioned mappings: raw product, reason and status values remain recoverable.版本化映射：原始商品、原因与状态值保持可恢复。
- Formula checks: expected blanks stay distinct from zero and unexpected errors remain visible.公式检查：预期空白与零保持不同，意外错误保持可见。
- Reconciliation: input counts and amounts tie to a source control before interpretation.核对：解读前输入数量与金额对上来源控制。
- Privacy: unnecessary direct identifiers and sensitive free text are removed or protected.隐私：不必要直接标识与敏感自由文本被移除或保护。
- Named review: an accountable owner signs definitions, limitations and action status.具名审核：负责所有人确认定义、局限与行动状态。
Avoid nine reason report workbook mistakes避免九个原因报告工作簿错误
- Mixing orders, order items, units, return requests, physical receipts and refunds in one row without a declared grain.把订单、订单商品行、商品件数、退货申请、实体收货与退款混在一行且不声明粒度。
- Using mutable names as join keys instead of stable order, line, product, variant, return and transaction identifiers.使用可变名称作为关联键，而不是稳定订单、商品行、商品、变体、退货与交易标识。
- Replacing missing values with zero and allowing an incomplete denominator to look valid.用零替代缺失值，让不完整分母看起来有效。
- Combining physical returns, cancellations, refunds, exchanges and replacements under one metric.把实体退货、取消、退款、换货与替换合并为一个指标。
- Pasting new exports over formulas, mappings or prior-period evidence without preserving a raw copy.把新导出粘贴覆盖公式、映射或上期证据，且不保留原始副本。
- Showing a percentage without the numerator, denominator, cohort window, as-of date and data coverage.展示百分比却没有分子、分母、群组窗口、截至日期与数据覆盖。
- Treating a template, spreadsheet formula or chart as a verified business conclusion.把模板、表格公式或图表当成已核验业务结论。
- Dropping missing and unmapped reasons from the report without publishing coverage.从报告删除缺失与未映射原因且不发布覆盖。
- Calling the largest Pareto category the root cause.把帕累托最大类别称为根因。
A clean-looking resource can still be wrong. Publish the data contract, coverage and reconciliation beside the output, and keep a reviewer accountable for every interpretation.
外观整洁的资源仍可能错误。应在输出旁发布数据契约、覆盖与核对，并让审核人对每项解读负责。
Convert the top reason groups into evidence reviews把重点原因组转为证据复核
Select by count, value, cost and evidence quality. Review cases before selecting content, fit, packaging, fulfillment or policy changes.
按数量、价值、成本与证据质量选择。选择内容、尺码、包装、履约或政策变化前先复核案例。
| Signal信号 | Evidence to check待检查证据 | Safe next step安全下一步 |
|---|---|---|
| Size/fit group material尺码/版型组重大 | Variant, body/product measurements, chart version, text, condition and exchanges变体、人体/商品测量、尺码表版本、文本、状态与换货 | Run fit mechanism review运行版型机制复核 |
| Damage group material破损组重大 | Condition, photos, packaging version, route, carrier, lot and test evidence状态、图片、包装版本、路线、承运、批次与测试证据 | Separate transit, product and handling damage分开运输、商品与处理破损 |
| Missingness high缺失率高 | Channel question, optionality, event joins and source extraction渠道问题、是否可选、事件关联与来源提取 | Repair collection before comparing mix比较结构前修复采集 |
Prepare governed files for Return Compass为逆向罗盘准备受治理文件
Prepare raw and mapped reason fields, taxonomy version, return-line keys, condition, disposition, product/variant, channel, cohort denominator, refund/cost and quality coverage. Return Compass can analyze governed files; current reason workflow and connector support must be validated.
准备原始与映射原因字段、分类版本、退货行键、状态、处置、商品/变体、渠道、群组分母、退款/成本与质量覆盖。逆向罗盘可分析受治理文件；当前原因工作流与连接器支持必须核验。
Open Return Compass打开逆向罗盘 →return reason report template FAQ退货原因报告模板常见问题
Show raw and mapped reasons, taxonomy version, counts, eligible denominators where used, missingness, value/cost, condition, disposition, segments and review decisions.展示原始与映射原因、分类版本、数量、使用时的合格分母、缺失、价值/成本、状态、处置、细分与复核决定。
No. It is a reported category or text response. Verify a mechanism with case, product, inspection, fulfillment and other relevant evidence.不等。它是报告类别或文本回答。应使用案例、商品、质检、履约及其他相关证据核验机制。
Publish missing and unmapped counts and shares under the same scope. Do not convert them to Other or silently remove them.在相同范围下发布缺失与未映射数量及占比。不要转为“其他”或静默删除。
No. Calculate the actual cumulative distribution. Use concentration to prioritize evidence review, not to impose an 80% threshold or infer causality.不能。应计算实际累计分布。用集中度确定证据复核优先级，不要强加 80% 阈值或推断因果。
Yes. The downloadable XLSX requires no payment or email and contains synthetic examples.是。可下载 XLSX 无需付款或邮件，包含模拟示例。
Sources, evidence labels, and limitations来源、证据标签与限制
- W3C Recommendation: Model for Tabular Data and Metadata on the Web — Primary specification for annotated tables, schemas, datatypes, primary and foreign keys, source provenance and validation of tabular data.W3C 推荐标准：Web 表格数据与元数据模型——关于带注释表格、Schema、数据类型、主键与外键、来源溯源及表格数据验证的一手规范。
- W3C: Data Quality Vocabulary — Primary vocabulary for expressing data-quality measurements, metrics, annotations, policies and provenance without implying one universal definition of quality.W3C：数据质量词汇表——用于表达数据质量测量、指标、注释、政策与溯源的一手词汇表，并不假设存在单一通用质量定义。
- Shopify Help Center: Sales reports — Official definitions distinguishing physical returned quantity and return-line reason from broader sales reversals, cancellations, edits, and refunds.Shopify 帮助中心：销售报告——官方定义区分实体退回数量与退货行原因，以及更广泛的销售冲回、取消、编辑与退款。
- Microsoft Learn: Return reason codes and disposition codes — Official distinction between why a customer requests a return and the condition/action assigned after physical inspection.Microsoft Learn：退货原因码与处置码——官方区分客户申请退货的原因与实体质检后分配的状态/动作。
- Oracle Retail Order Management: Establishing Return Reason Codes — Official reference for reason-code maintenance, storefront use, authorization/receipt/credit stages, history, counts, and returned merchandise value reporting.Oracle 零售订单管理：建立退货原因码——原因码维护、前台使用、授权/收货/贷项阶段、历史、数量与退货金额报告的官方参考。
- Microsoft Support: Create and format tables — Current official guidance for turning a range with headers into an Excel table for grouped, filterable analysis. It does not validate any returns metric or workbook design.Microsoft 支持：创建并格式化表格——关于把带表头区域转为可分组筛选分析的 Excel 表格的当前官方指南；它不验证任何退货指标或工作簿设计。
Evidence statement: W3C tabular-data and data-quality specifications support explicit schemas, keys, provenance and quality measurements; Microsoft sources document the relevant spreadsheet features where cited. The reason workbook and examples are synthetic. Platform sources support field distinctions but not this reason distribution or any root-cause claim. Examples are synthetic. No customer result, universal benchmark, causal claim, or guaranteed improvement is asserted. Sources were reviewed September 15, 2026; named subject-matter review is required before publication.证据声明：W3C 表格数据与数据质量规范支持明确 Schema、键、溯源与质量测量；引用的 Microsoft 来源说明相关表格功能。原因工作簿与示例为模拟。平台来源支持字段区分，但不支持本原因结构或任何根因结论。示例为模拟。本文不声称客户结果、通用基准、因果结论或保证改善。来源核验于 2026 年 9 月 15 日完成；发布前需要具名领域审核。
Use the resource as a governed starting point把资源作为受治理起点
Preserve raw responses, version mappings, separate reason from condition and disposition, publish missingness and denominators, and use ranked counts to allocate case review—not to declare causes.
保留原始回答、版本化映射，分开原因、状态与处置，发布缺失与分母，并使用数量排序分配案例复核，而不是宣布原因。