# Plain-Language Return-Reason FAQ Prompt

> A reusable prompt that turns a supplied return-reason list into a plain-language FAQ using only stated patterns, with no invented policy or frequency.

## Install

Copy the content below into your project:

# Plain-Language Return-Reason FAQ Prompt

A reusable prompt that turns a supplied return-reason list into a plain-language FAQ using only stated patterns, with no invented policy or frequency.

## Start here

Paste your return reasons into an ordinary AI chat that accepts text (for example a standard ChatGPT conversation). No terminal, account setup or integration is needed.

**What to paste**
- The raw return-reason list, one reason per line.
- The product or product category.
- The shop's existing return/refund policy text, if any.
- Who will read the FAQ: customers or frontline staff.

**How to paste it**
1. Copy the full prompt from the original prompt block appended below.
2. Paste it first as your message, or paste it as the first line and add your inputs after it.
3. Send one message. If any input is missing, the prompt tells the tool to label the gap rather than invent details.

**How to check the output**
- The draft should start with a two-sentence intro saying it is based only on your reasons.
- Every FAQ theme should trace back to at least one reason you listed. Read the list again and cross-check.
- No percentages, frequencies or trends should appear.
- Each FAQ should have a "What we need from you" line and a "Not covered here" line.
- The "Questions we cannot yet answer" list should be non-empty when your policy or timing details were missing.
- The draft should end with the staff routing note.

## Introduction

This is TokRepo's original reusable prompt, not a third-party README or an announcement. It is written for general readers in the writing category: small shops and office staff who want an honest, plain-language FAQ about why customers return a product. The prompt only groups the reasons you supply. It never estimates frequency, percentages or trends, and it never invents a cause, policy, remedy or timeline that is not in your input.

## What it does

The prompt reads every supplied reason literally, groups only where the wording clearly supports it, and writes each FAQ in a four-part shape: question, short answer, what we need from you, and not covered here. It then adds a "Questions we cannot yet answer" list and a one-line note for staff on how to handle a return that fits no listed theme: acknowledge it, log it, and route it to the owner rather than guessing.

## Input permissions and limitations

- **Permissions:** none. You paste text and read the result; nothing is edited, sent, posted or scheduled. This produces a draft document only.
- **Privacy:** only paste return reasons and policy text you are allowed to share with the AI tool you use. Redact customer names, order numbers or any personal data before pasting if your policy requires it.
- **Limitations:** the prompt does not give legal, tax or consumer-rights advice; it flags that area as needing the owner or a qualified adviser. If the input is empty or has no usable reasons, it returns only a request for the raw list plus two open questions.
- **Verification note:** source reviewed; runtime not tested.

## Worked example (fictional, for shape only)

These reasons and outputs are fictional and are shown only to illustrate the shape. They are not real data and no results were tested.

| Reason supplied | FAQ theme | What we need from you |
|---|---|---|
| item arrived cracked | What if my item arrives damaged? | Not specified in the supplied policy |
| box was crushed | What if my item arrives damaged? | Not specified in the supplied policy |
| customer ordered wrong size | I ordered the wrong size | Not specified in the supplied policy |

## Manual alternative (no AI feature)

Copy the reason list into a spreadsheet, one reason per row. Give each row a short plain-language label. Sort rows so identical labels sit together. Write one question per label group in the four-part shape above. List every blank you cannot fill from the policy text, then finish with the staff routing note. Apply the same checks line by line.

## FAQ

**What if I only have a few reasons or no policy text?**
The prompt proceeds with what you have and labels the gaps instead of inventing details. With no policy text, the "What we need from you" line reads "Not specified in the supplied policy."

**Can it tell me why returns happen or how often?**
No. It only restates and groups the reasons you paste, and it must not estimate frequency, percentages or trends.

## Source and thanks

Original TokRepo prompt, licensed CC BY 4.0. Reference context: [ChatGPT release notes](<https://help.openai.com/en/articles/6825453-chatgpt-release-notes>). External reference material retains its own rights. The complete prompt and its Chinese translation are appended separately.

## Complete reusable prompt

You are helping a small business write a plain-language FAQ about why customers return a product. Your input is a supplied list of raw return reasons. Use only what is in that list; never estimate frequency, percentages or trends, and never invent a cause, policy, remedy or timeline that is not present in the input.

INPUT TO REQUEST
- The raw return-reason list, pasted one reason per line or as short notes.
- The product or product category it applies to.
- The actual return or refund policy text the shop already uses, if any.
- Who will read the FAQ: customers, or frontline staff answering customers.
If any of these is missing, proceed with what you have and label the gaps instead of inventing details.

YOUR TASK
1. Read every supplied reason literally. Do not merge reasons that differ, and do not add a reason you did not receive.
2. Group the reasons into broad plain-language themes only where the supplied wording clearly supports the grouping. If a reason is ambiguous, keep it ungrouped and say so.
3. For each theme, write one FAQ in this shape:
   - Question: the plain question a customer or staff member would actually ask.
   - Short answer: what the shop can say, drawn only from the supplied reasons and the supplied policy.
   - What we need from you: the specific detail the customer must provide (for example order number, photo, description), only if the policy input supports it. Otherwise write 'Not specified in the supplied policy.'
   - Not covered here: any question the reader will likely ask that the input cannot answer.
4. Add a short 'Questions we cannot yet answer' list naming every gap the input leaves open.
5. Add a one-line note for staff on how to handle a return that fits no listed theme: acknowledge it, log it, and route it to the owner rather than guessing.

OUTPUT FORMAT
Return Markdown with: a two-sentence intro stating the FAQ is based only on the reasons provided; then a numbered FAQ list using the four-part shape above; then 'Questions we cannot yet answer'; then the staff note. Do not add a summary, marketing tone, reassurance or urgency language.

WORKED EXAMPLE (fictional, for shape only — do not treat as real data)
Fictional input reasons: item arrived cracked; box was crushed; customer ordered wrong size; customer changed mind.
Illustrative output shape:
1. Question: What if my item arrives damaged? Short answer: Returns are accepted when a supplied reason is that the item arrived cracked or the box was crushed. What we need from you: Not specified in the supplied policy (no policy text was provided). Not covered here: who pays return shipping on damaged items. Questions we cannot yet answer: cost of return shipping; time limit for reporting damage; whether replacement is offered instead of refund.

CHECKS BEFORE YOU RETURN THE DRAFT (state pass or fail for each, tied to the input)
- Every FAQ theme can be traced to at least one listed reason. Pass or fail.
- No frequency, percentage or trend wording appears anywhere. Pass or fail.
- No policy, remedy, fee or deadline appears that was not supplied. Pass or fail.
- Each FAQ has a 'What we need from you' line, and it either cites supplied policy or says not specified. Pass or fail.
- The 'Questions we cannot yet answer' list is non-empty when the input lacks policy or timing details. Pass or fail.
If any check fails, revise before returning and say what you changed.

BOUNDARIES
- This produces a draft document only. You are not editing any account, website, help centre or live chat, and nothing is sent, posted or scheduled.
- Do not state or imply that return rates, customer satisfaction or complaint volumes are known from the input.
- Do not give legal, tax or consumer-rights advice; flag any question in that area as needing the owner or a qualified adviser.
- If the input is empty or contains no usable reasons, return only a request for the raw list plus the two open questions.

MANUAL ALTERNATIVE WITHOUT AN AI FEATURE
If you have no AI tool available, do this by hand: copy the reason list into a spreadsheet, one reason per row; give each row a short plain-language label; sort rows so identical labels sit together; write one question per label group in the four-part shape above; list every blank you cannot fill from the policy text; finish with the staff routing note. The same checks apply line by line.

## References and reuse

- [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes)

Original TokRepo prompt · [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/). Reference documents retain their own rights.

---

<!-- ZH -->

# 把退货原因整理成通俗退货问答的提示词

一条可复用的提示词，把已提供的退货原因清单整理成通俗易懂的退货问答，只使用清单里写明的模式，不编造政策或出现频率。

## 开始使用

把退货原因粘贴到一个能接收文本的普通 AI 聊天窗口即可（例如标准的 ChatGPT 对话）。不需要终端、账号配置或集成。

**要粘贴什么**
- 原始退货原因清单，每行一条原因。
- 产品名称或产品类别。
- 店铺已经在使用的退货/退款政策文本（如有）。
- 谁会读这份问答：顾客，还是一线接待顾客的员工。

**怎么粘贴**
1. 复制下方单独附上的原始提示词全文。
2. 先把它作为消息粘贴，或把它粘在第一行、再把你的输入接在后面。
3. 发送一条消息。如果有输入缺失，提示词会让工具标注缺口，而不是编造细节。

**怎么检查输出**
- 草稿应以两句话的开头说明：它只基于你提供的原因。
- 每个问答主题都应能追溯到至少一条你列出的原因。把你写的清单再读一遍并对照检查。
- 不应出现任何百分比、频率或趋势的说法。
- 每条问答都应有“你需要提供什么”和“此处未覆盖”两行。
- 当你缺少政策或时限细节时，“暂时无法回答的问题”列表应非空。
- 草稿应以员工处理说明结尾。

## 简介

这是 TokRepo 的原创可复用提示词，不是第三方 README，也不是公告转载。它面向写作类别的普通读者：想要一份诚实、通俗的退货原因问答的小店主和办公室员工。提示词只整理你提供的原因，绝不估算频率、百分比或趋势，也绝不编造输入里没有的原因、政策、补救措施或时限。

## 它能做什么

提示词会逐条按字面读取你提供的原因，只在措辞明确支持时才归组，并把每条问答写成四部分：问题、简短回答、你需要提供什么、此处未覆盖。随后它加上“暂时无法回答的问题”列表，再加一行员工处理说明：对于不属于任何已列主题的退货，先确认、记录，并交给负责人，而不是猜测。

## 输入、权限与限制

- **权限：** 无。你粘贴文本、阅读结果；不会编辑、发送、发布或定时任何内容。它只产出一份草稿文档。
- **隐私：** 只粘贴你被允许与你使用的 AI 工具分享的退货原因和政策文本。如果你的政策要求，粘贴前请先隐去顾客姓名、订单号等个人信息。
- **限制：** 提示词不提供法律、税务或消费者权益建议；这类问题会被标记为需要店主或合资格顾问处理。如果输入为空或没有可用原因，它只返回一句索要原始清单的请求和两个待解问题。
- **验证说明：** 已审阅来源；未做运行测试。

## 示例（虚构，仅示形式）

以下原因和输出均为虚构，仅用于说明形式，不是真实数据，也未做过测试。

| 提供的原因 | 问答主题 | 你需要提供什么 |
|---|---|---|
| item arrived cracked | 商品到货时破损怎么办？ | 提供的政策中未说明 |
| box was crushed | 商品到货时破损怎么办？ | 提供的政策中未说明 |
| customer ordered wrong size | 我买错了尺码 | 提供的政策中未说明 |

## 没有 AI 功能时的手动做法

把原因清单复制到表格里，一行一条原因。给每行写一个简短的通俗标签。按标签排序，让相同标签排在一起。每个标签组按上面的四部分形式写一个问题。把你无法从政策文本中填补的空白全部列出，最后写员工处理说明。逐行套用同样的检查。

## 常见问题

**如果只有几条原因，或者没有政策文本怎么办？**
提示词会用你已有的内容继续，并标注缺口而不是编造细节。没有政策文本时，“你需要提供什么”一行会写“提供的政策中未说明”。

**它能告诉我退货的原因是什么、有多频繁吗？**
不能。它只复述和归组你粘贴的原因，且不得估算频率、百分比或趋势。

## 来源与致谢

TokRepo 原创提示词，采用 CC BY 4.0 许可。参考背景：[ChatGPT 发布说明](<https://help.openai.com/en/articles/6825453-chatgpt-release-notes>)。外部参考材料保留其自身权利。完整提示词及其中文翻译单独附上。

## 完整可复制提示词

你正在帮助一家小企业撰写一份通俗易懂的常见问题解答（FAQ），内容是关于顾客为什么要退货。你的输入是一份提供的原始退货原因清单。只使用清单中的内容；绝不估算频率、百分比或趋势，也绝不编造输入中不存在的原因、政策、补救措施或时限。

需要请求的输入
- 原始退货原因清单，每行粘贴一条原因，或作为简短备注。
- 它适用的产品或产品类别。
- 店铺已经使用的实际退货或退款政策文本（如有）。
- 谁会阅读这份 FAQ：顾客，还是接待顾客的一线员工。
如果其中任何一项缺失，就用你已有的内容继续，并标注缺口，而不是编造细节。

你的任务
1. 逐条按字面读取每个提供的原因。不要合并不同的原因，也不要添加你未收到过的原因。
2. 仅在所提供的措辞明确支持归组时，才把原因归入宽泛的通俗主题。如果某个原因含糊不清，就保持不归组并说明这一点。
3. 对每个主题，按以下形式写一条 FAQ：
   - 问题：顾客或员工实际会问的通俗问题。
   - 简短回答：店铺可以说什么，仅取自所提供的原因和所提供政策。
   - 你需要提供什么：顾客必须提供的具体细节（例如订单号、照片、描述），仅在政策输入支持时才写。否则写“提供的政策中未说明”。
   - 此处未覆盖：读者很可能会问、但输入无法回答的任何问题。
4. 添加一个简短的“暂时无法回答的问题”列表，列出输入留下的每一个缺口。
5. 添加一行员工处理说明：对于不属于任何已列主题的退货该如何处理：先确认、记录，并交给负责人，而不是猜测。

输出格式
返回 Markdown，包含：两句话的开头说明，表明该 FAQ 仅基于所提供的原因；然后是按上述四部分形式编写的编号 FAQ 列表；然后是“暂时无法回答的问题”；然后是员工说明。不要添加摘要、营销语气、安抚或紧迫性语言。

示例（虚构，仅示形式——不要当作真实数据）
虚构的输入原因：item arrived cracked；box was crushed；customer ordered wrong size；customer changed mind。
说明性输出形式：
1. 问题：如果我的商品到货时破损怎么办？简短回答：当提供的原因中含商品到货时开裂或包装箱被压坏时，接受退货。你需要提供什么：提供的政策中未说明（未提供政策文本）。此处未覆盖：破损商品的退货运费由谁承担。暂时无法回答的问题：退货运费；报告破损的时限；是否提供换货而非退款。

返回草稿前的检查（针对每项说明通过或未通过，并联系输入）
- 每个 FAQ 主题都能追溯到至少一条已列原因。通过或未通过。
- 任何地方都没有出现频率、百分比或趋势相关措辞。通过或未通过。
- 没有出现未提供的政策、补救措施、费用或截止日期。通过或未通过。
- 每条 FAQ 都有“你需要提供什么”一行，并且要么引用所提供的政策，要么写明未说明。通过或未通过。
- 当输入缺少政策或时限细节时，“暂时无法回答的问题”列表非空。通过或未通过。
如果任何检查未通过，返回前先修订，并说明你改了什么。

边界
- 这只产出一份草稿文档。你不会编辑任何账户、网站、帮助中心或在线聊天，也不会发送、发布或定时任何内容。
- 不要声称或暗示可从输入中得知退货率、客户满意度或投诉量。
- 不要提供法律、税务或消费者权益建议；将任何属于该领域的问题标记为需要店主或合资格顾问处理。
- 如果输入为空或不含可用原因，只返回一句索要原始清单的请求以及两个待解问题。

## 没有 AI 功能时的手动做法

如果你没有可用的 AI 工具，就手动来做：把原因清单复制到表格里，一行一条原因；给每行写一个简短的通俗标签；按标签排序，让相同标签排在一起；每个标签组按上面的四部分形式写一个问题；把你无法从政策文本中填补的空白全部列出；最后写员工处理说明。逐行套用同样的检查。

## 参考资料与复用

- [ChatGPT release notes](https://help.openai.com/en/articles/6825453-chatgpt-release-notes)

TokRepo 原创提示词 · [CC BY 4.0](https://creativecommons.org/licenses/by/4.0/)。参考资料保留各自原有权利。

---
Source: https://tokrepo.com/en/workflows/plain-language-return-reason-faq-prompt-344dd149
Author: Prompt Lab