copywriting

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

When this skill is activated, always start your first response with the 🧢 emoji.

激活此技能后，你的第一条回复必须以🧢表情符号开头。

Copywriting

文案写作

Copywriting is the craft of writing words that persuade people to take action - buy, sign up, click, subscribe, or believe. Unlike content writing (which informs), copywriting converts. Every word earns its place. The best copy feels effortless to read because enormous effort went into writing it. This skill gives an agent the frameworks, formulas, and judgment to produce headlines, landing pages, CTAs, value propositions, email subject lines, and product descriptions that actually work.

文案写作是撰写能够说服人们采取行动的文字艺术——比如购买、注册、点击、订阅或相信。与内容写作（以传递信息为目的）不同，文案写作的核心是转化。每一个字都有其存在的价值。优秀的文案读起来轻松自然，背后却倾注了大量的心血。此技能为Agent提供框架、公式和判断能力，使其能够创作出真正有效的标题、着陆页文案、CTA、价值主张、邮件主题行和产品描述。

When to use this skill

何时使用此技能

Trigger this skill when the user:

Asks to write or improve a headline, title, or hook
Needs landing page copy, above-the-fold text, or hero sections
Wants to write or optimize a call-to-action (CTA)
Asks for email subject lines or preview text
Needs a value proposition or positioning statement
Wants to apply AIDA, PAS, BAB, or other persuasion frameworks
Asks to write product descriptions that sell
Wants to A/B test copy and needs variant ideas
Asks for "compelling," "persuasive," or "converting" copy

Do NOT trigger this skill for:

Long-form editorial or journalistic content (informing, not converting)
Technical documentation or developer-facing copy where clarity trumps persuasion

当用户有以下需求时，触发此技能：

请求撰写或优化标题、名称或钩子内容
需要着陆页文案、首屏文本或核心展示区内容
希望撰写或优化行动号召（CTA）
请求撰写邮件主题行或预览文本
需要价值主张或定位声明
希望应用AIDA、PAS、BAB或其他说服框架
请求撰写具有销售力的产品描述
想要进行文案A/B测试并需要不同变体的创意
请求创作“有吸引力的”“有说服力的”或“高转化的”文案

以下情况请勿触发此技能：

长篇社论或新闻类内容（以传递信息为目的，而非转化）
技术文档或面向开发者的文案（此类内容优先考虑清晰性而非说服力）

Key principles

核心原则

Clarity over cleverness - If a reader has to think about what you mean, you've lost them. Plain language outperforms clever wordplay in almost every A/B test. Say the obvious thing, clearly.
Benefits over features - Features describe the product. Benefits describe what the customer gains. Never lead with a feature. Translate every feature into the outcome it produces for the buyer.
One CTA per piece - Multiple calls to action dilute attention and reduce conversions. Every piece of copy has one job. One desired action. Everything else is noise.
Write for scanners, then readers - Most visitors scan before they read. Structure copy so the headline + subheads + CTAs tell the complete story. Readers who want detail will find it in the body.
Test everything - Copy intuition improves with data. Treat every headline as a hypothesis. A/B test subject lines, CTAs, and hero copy. Small wording changes routinely shift conversion rates by 20-40%.

清晰优先于巧妙 - 如果读者需要思考才能理解你的意思，你已经失去了他们。在几乎所有A/B测试中，直白的语言都比巧妙的文字游戏表现更好。清晰地说出显而易见的事情。
利益优先于功能 - 功能描述产品，利益描述客户能获得什么。永远不要以功能开头。将每一项功能转化为能为买家带来的结果。
单篇文案仅一个CTA - 多个行动号召会分散注意力，降低转化率。每篇文案只有一个目标，一个期望的行动。其他所有内容都是干扰项。
先为浏览者写作，再为阅读者写作 - 大多数访客会先浏览再阅读。结构化文案，让标题+副标题+CTA就能完整传达核心信息。想要了解细节的读者会在正文中找到内容。
测试所有内容 - 文案直觉会随着数据积累而提升。将每个标题视为一个假设。对主题行、CTA和核心展示区文案进行A/B测试。微小的措辞变化通常能将转化率提升20-40%。

Core concepts

核心概念

Persuasion frameworks

说服框架

Three battle-tested structures cover 90% of copywriting situations:

AIDA (Attention - Interest - Desire - Action) The classic funnel. Hook with attention, build interest with relevance, create desire by amplifying the benefit, then drive action with a clear CTA. Best for long-form sales pages and email sequences.

PAS (Problem - Agitate - Solve) Lead with the reader's pain, twist the knife by amplifying consequences, then position your product as the solution. Especially effective for cold audiences who don't yet know they have a problem.

BAB (Before - After - Bridge) Paint the reader's current frustrating situation (before), describe the ideal transformed state (after), then present your product as the bridge. Works well for testimonials, case studies, and social ads.

经过实战检验的三种结构可覆盖90%的文案写作场景：

AIDA（注意力-兴趣-欲望-行动） 经典的漏斗模型。以注意力钩子吸引读者，通过相关性建立兴趣，放大利益点激发欲望，然后用清晰的CTA推动行动。最适合长篇销售页和邮件序列。

PAS（问题-激化-解决） 从读者的痛点切入，通过放大后果来加剧痛苦，然后将你的产品定位为解决方案。对尚未意识到自己存在问题的陌生受众尤其有效。

BAB（之前-之后-桥梁） 描绘读者当前令人沮丧的处境（之前），描述理想的转变状态（之后），然后将你的产品呈现为实现转变的桥梁。在推荐信、案例研究和社交广告中效果良好。

Voice of customer (VoC) research

客户声音（VoC）研究

The most powerful copy uses the customer's own words. Before writing, collect:

Exact phrases from support tickets and sales call transcripts
Amazon/G2/Trustpilot reviews for the product category
Reddit/forum threads where the target audience describes their problem
Survey responses, especially to "What almost stopped you from buying?"

Mirror this language verbatim in headlines and body copy. It creates instant recognition: "they're talking about me."

最有说服力的文案会使用客户自己的语言。写作前，收集以下内容：

支持工单和销售通话记录中的精确措辞
产品类别的亚马逊/G2/Trustpilot评论
目标受众讨论问题的Reddit/论坛帖子
调查回复，尤其是“什么差点阻止你购买？”的答案

在标题和正文中直接使用这些语言。这会立刻让读者产生共鸣：“他们在说我。”

Power words

强力词汇

Words that reliably increase emotional engagement and clicks:

Category	Examples
Urgency	Now, Today, Instantly, Before it's gone
Exclusivity	Only, Private, Members-only, Invitation
Curiosity	Secret, Hidden, Surprising, What most people miss
Credibility	Proven, Backed by, Trusted by, Certified
Ease	Simple, Effortless, One-click, Without the hassle
Benefit	Free, Save, Boost, Double, Eliminate

Use sparingly. Overuse makes copy feel like spam.

能可靠提升情感参与度和点击率的词汇：

类别	示例
紧迫感	Now, Today, Instantly, Before it's gone
排他性	Only, Private, Members-only, Invitation
好奇心	Secret, Hidden, Surprising, What most people miss
可信度	Proven, Backed by, Trusted by, Certified
便捷性	Simple, Effortless, One-click, Without the hassle
利益点	Free, Save, Boost, Double, Eliminate

谨慎使用。过度使用会让文案看起来像垃圾邮件。

Social proof

社交证明

Proof reduces skepticism. Layer it throughout copy:

Numbers - "14,000 teams use this" beats "used by teams worldwide"
Named testimonials - Full name + company + photo outperforms anonymous quotes
Results-based - "Reduced churn by 31%" beats "great product"
Logos - Well-known customer logos do the work of a thousand words
Press mentions - "As seen in..." for trust with new audiences

证明能减少怀疑。在文案中层层植入：

数字 - “14000个团队正在使用”比“全球团队都在使用”效果更好
署名推荐信 - 全名+公司+照片比匿名引语效果更好
结果导向 - “将客户流失率降低31%”比“很棒的产品”效果更好
品牌标志 - 知名客户的标志抵得上千言万语
媒体报道 - “被...报道”能为新受众建立信任

Common tasks

常见任务

Write headlines

撰写标题

A headline's only job is to earn the next line. Use these 10 formula templates:

How-to - "How to [achieve desired outcome] without [common obstacle]"
- "How to Write Emails People Actually Open Without Sounding Desperate"
Number list - "[Number] [things] that [outcome]"
- "7 Landing Page Mistakes Killing Your Conversions"
The secret - "The [unusual/counterintuitive] way to [desired outcome]"
- "The Counterintuitive Reason Your Freemium Users Never Convert"
Question - "[Question that implies reader has the problem]"
- "Still Paying Full Price for Software Nobody Uses?"
Direct benefit - "[Do X] and [get Y]"
- "Schedule Once and Let the System Handle All Follow-ups"
Before/After - "From [undesirable state] to [desirable state] in [timeframe]"
- "From Spreadsheet Chaos to Automated Reports in 15 Minutes"
Social proof hook - "How [customer type] [achieved result] with [product]"
- "How a 3-Person Agency Closed $2M Using Cold Email Templates"
Warning - "Don't [do X] until you [read/know/try Y]"
- "Don't Launch Your Landing Page Until You Fix These 5 Copy Errors"
Fascination - "[Number] [surprising things] about [familiar topic]"
- "9 Things Your Bounce Rate Is Actually Telling You"
Challenge - "What if you could [outcome] without [sacrifice]?"
- "What if You Could Double Response Rates Without Writing More Emails?"

Headline checklist: Is the benefit specific? Is the audience implied? Does it create curiosity or urgency? Is it under 12 words? Could it stand alone?

标题的唯一作用是让读者愿意继续阅读下一行。使用以下10种公式模板：

方法型 - "如何[实现期望结果]，无需[常见障碍]"
- "如何撰写让人们愿意打开且不显得急切的邮件"
数字列表型 - "[数字]个[事物]，帮你[达成结果]"
- "7个正在毁掉你转化率的着陆页错误"
秘密型 - "[不寻常/反直觉的]方法，帮你[实现期望结果]"
- "导致你的免费版用户从未转化的反直觉原因"
问题型 - "[暗示读者存在该问题的问句]"
- "还在为没人用的软件全额付费？"
直接利益型 - "[做X]，就能[获得Y]"
- "只需安排一次，系统将处理所有跟进工作"
前后对比型 - "从[不理想状态]到[理想状态]，只需[时间周期]"
- "从电子表格混乱到自动化报表，仅需15分钟"
社交证明钩子型 - "[客户类型]如何通过[产品]实现[成果]"
- "3人小团队如何使用冷邮件模板完成200万美元营收"
警告型 - "在你[阅读/了解/尝试Y]之前，不要[做X]"
- "在修复这5个文案错误之前，不要发布你的着陆页"
趣味吸引型 - "关于[熟悉话题]的[数字]个[令人惊讶的事实]"
- "你的跳出率真正在告诉你的9件事"
挑战型 - "如果无需[牺牲]就能[实现结果]，会怎样？"
- "如果无需撰写更多邮件就能将回复率翻倍，会怎样？"

标题检查清单： 利益点是否具体？是否隐含目标受众？是否能引发好奇心或紧迫感？是否在12字以内？是否能独立表意？

Write landing page copy

撰写着陆页文案

Structure every landing page using this section-by-section framework:

1. Hero section (above the fold)

Headline: primary benefit or transformation (use a formula above)
Subheadline: one sentence clarifying who it's for and how
CTA: verb + specific outcome ("Start Free Trial" not "Submit")
Optional: social proof hook ("Join 14,000 teams...")

2. Problem section

Name the pain clearly, using VoC language
Agitate: what's the cost of not solving it?
Transition: "There's a better way."

3. Solution section

Introduce the product as the logical answer to the pain
Three core benefits (not features) with one supporting sentence each
Show, don't tell: screenshots, demo GIFs, or short video

4. Features section (benefits-led)

Lead with the outcome ("Never lose a lead again"), then the feature beneath it
Limit to 3-6 features - more creates fatigue

5. Social proof section

Two to three testimonials with specific, result-oriented quotes
Customer logos if available
Case study summary with before/after numbers

6. Objection handling (FAQ or dedicated section)

Address the top 3-5 objections from sales calls
Write each answer as a mini persuasion piece, not just a fact

7. Final CTA section

Restate the primary benefit
Repeat the CTA with slightly different framing from the hero
Add urgency or de-risk with a guarantee if available

使用以下分节框架构建每个着陆页：

1. 核心展示区（首屏）

标题：核心利益或转变（使用上述公式）
副标题：一句话明确受众和实现方式
CTA：动词+具体结果（用“开始免费试用”而非“提交”）
可选：社交证明钩子（“加入14000个团队...”）

2. 问题区

清晰点明痛点，使用VoC语言
激化：不解决问题会付出什么代价？
过渡：“有一种更好的方法。”

3. 解决方案区

将产品作为痛点的合理解决方案引入
三个核心利益（而非功能），每个配一句支持性说明
展示而非讲述：截图、演示GIF或短视频

4. 功能区（以利益为导向）

以结果开头（“再也不会丢失潜在客户”），下方再列出功能
限制在3-6个功能——过多会让读者疲劳

5. 社交证明区

2-3个带有具体、结果导向引语的推荐信
如有可用的客户标志
包含前后对比数据的案例研究摘要

6. 异议处理区（FAQ或专门板块）

解决销售通话中最常见的3-5个异议
将每个答案写成小型说服文案，而非仅仅陈述事实

7. 最终CTA区

重申核心利益
使用与核心展示区略有不同的表述重复CTA
如有可能，添加紧迫感或无风险承诺（如“随时可取消”）

Craft CTAs that convert

打造高转化CTA

The CTA is the moment of conversion. Apply these rules:

Use first person - "Start My Free Trial" outperforms "Start Your Free Trial" by 90% in most A/B tests (ownership effect)
Be specific - "Download the 2024 Report" beats "Download Now"
Reduce friction language - "See How It Works" is lower commitment than "Buy"
Add micro-copy beneath - "No credit card required. Cancel anytime." eliminates the risk objection right at the decision point
Match CTA to page temperature - Cold traffic: "See a Demo." Warm: "Start Free Trial." Hot: "Get Instant Access."

Funnel stage	Low-friction CTA examples
Awareness	"See How It Works", "Watch the Demo"
Consideration	"Start Free Trial", "Get the Guide"
Decision	"Get Instant Access", "Claim My Spot"

CTA是转化的关键时刻。遵循以下规则：

使用第一人称 - 在大多数A/B测试中，“开始我的免费试用”比“开始你的免费试用”的表现高出90%（所有权效应）
具体明确 - “下载2024年报告”比“立即下载”效果更好
减少摩擦性语言 - “查看工作原理”比“购买”的承诺感更低
在下方添加微型文案 - “无需信用卡。随时可取消。” 在决策点直接消除风险异议
匹配页面热度 - 陌生流量：“查看演示”；意向流量：“开始免费试用”；高意向流量：“立即获取权限”

漏斗阶段	低摩擦CTA示例
认知阶段	“查看工作原理”、“观看演示”
考虑阶段	“开始免费试用”、“获取指南”
决策阶段	“立即获取权限”、“抢占我的席位”

Write value propositions

撰写价值主张

A value proposition answers: "Why should I buy from you, not your competitor?"

Template: [Product] helps [target customer] [achieve outcome] by [differentiating mechanism] so they can [deeper benefit].

Examples:

"Loom helps remote teams communicate faster by replacing long email threads with instant screen recordings so they can move projects forward without another meeting."
"Superhuman helps founders reach inbox zero by combining keyboard shortcuts and AI triage so they spend under 30 minutes on email each day."

Test your value prop against this checklist:

Is the customer and their job-to-be-done clear?
Is the outcome specific and measurable?
Is there a mechanism that distinguishes you from alternatives?
Would this resonate with VoC language from real customers?

价值主张要回答：“为什么我要从你这里购买，而不是你的竞争对手？”

模板：[产品] 帮助 [目标客户] [实现结果]，通过 [差异化机制]，让他们能够 [更深层次的利益]。

示例：

“Loom帮助远程团队更快地沟通，通过即时屏幕录制替代冗长的邮件线程，让他们无需再开会就能推进项目。”
“Superhuman帮助创始人实现收件箱清零，通过结合键盘快捷键和AI分类，让他们每天花在邮件上的时间不到30分钟。”

使用以下清单测试你的价值主张：

客户及其核心需求是否清晰？
结果是否具体且可衡量？
是否有与竞品区分开的机制？
是否与真实客户的VoC语言产生共鸣？

Write email subject lines

撰写邮件主题行

Subject lines determine 47% of whether an email gets opened. Apply these formulas:

Formula	Template	Example
Curiosity gap	"[Intriguing partial statement]..."	"I almost didn't send this..."
Number	"[N] [things] for [person]"	"3 headlines worth stealing"
Personal	"Quick question, [name]"	"Quick question, Sarah"
Benefit	"[Specific outcome] in [timeframe]"	"Double open rates in 7 days"
Re-engage	"Did you see this?"	"Did you see this case study?"
Fear of missing out	"Last chance for [thing]"	"Last chance: free audit closes Friday"
Controversy	"Unpopular opinion: [contrarian take]"	"Unpopular opinion: shorter emails convert better"

Subject line rules:

Keep under 50 characters (mobile preview cutoff)
Avoid ALL CAPS and excessive punctuation (spam signals)
Match the tone of what's inside or people feel tricked
Test preview text - it's a second headline

主题行决定了47%的邮件打开率。使用以下公式：

公式	模板	示例
好奇心缺口型	“[引人好奇的部分陈述]...”	“我差点没发这封邮件...”
数字型	“[数字]个[事物]，送给[人群]”	“3个值得借鉴的标题”
个性化型	“快速问一句，[姓名]”	“快速问一句，莎拉”
利益型	“[具体结果]，只需[时间周期]”	“7天内将打开率翻倍”
重激活型	“你看到这个了吗？”	“你看到这个案例研究了吗？”
错失恐惧型	“[事物]的最后机会”	“最后机会：免费审计将于周五截止”
争议型	“Unpopular opinion: [contrarian take]”	“Unpopular opinion: 更短的邮件转化率更高”

主题行规则：

控制在50字符以内（移动端预览截断点）
避免全大写和过多标点（垃圾邮件信号）
与邮件内部内容语气一致，否则读者会觉得被欺骗
测试预览文本——它是第二个标题

Write product descriptions

撰写产品描述

Product descriptions must do two things: help the customer picture ownership, and pre-empt the objection "but does it work for someone like me?"

Structure:

Lead line - The single most compelling benefit, in one sentence
Sensory or outcome detail - Help the reader feel or visualize the result
Feature list - Bulleted, brief, benefits-first
Social proof - One inline stat or quote
CTA - Specific to the product

Bad: "Premium noise-cancelling headphones with 30-hour battery life." Good: "Silence your open office and stay in flow for 30 hours straight. Adaptive ANC adjusts to your environment in real time - no more manually tweaking settings when you move from desk to commute. Rated #1 by Wirecutter for three consecutive years."

产品描述必须完成两件事：帮助客户想象拥有产品的场景，以及预先打消“但这对我这样的人有用吗？”的异议。

结构：

导语 - 最具吸引力的核心利益，一句话概括
感官或结果细节 - 帮助读者感受或想象使用后的结果
功能列表 - 项目符号形式，简洁，以利益为导向
社交证明 - 一个嵌入式数据或引语
CTA - 针对产品的具体行动

反面例子：“高级降噪耳机，续航30小时。” 正面例子：“让你在开放办公室里保持专注，连续30小时不受干扰。自适应ANC会实时调整以适应你的环境——从办公桌到通勤途中，无需手动调整设置。连续三年被Wirecutter评为最佳。”

A/B test copy effectively

有效进行文案A/B测试

A/B testing copy without structure produces noise. Follow this process:

One variable at a time - Headline OR CTA OR hero image, never all three
Set a hypothesis - "Changing the CTA from 'Get Started' to 'Start My Free Trial' will increase clicks because it implies ownership and specificity"
Define the metric first - Click-through rate, conversion rate, or revenue
Run to statistical significance - Minimum 100 conversions per variant; use a significance calculator before calling a winner
Document learnings - Winning insights compound. Keep a swipe file of what works for your audience
Iterate, don't pivot - If variant B wins, make it the control and test variant C. Small improvements compound over time.

For subject line testing: 20% to variant A, 20% to variant B, 60% held for the winner (send within 4 hours).

无结构的文案A/B测试只会产生噪音。遵循以下流程：

一次只测试一个变量 - 仅测试标题、CTA或核心展示区图片中的一个，不要同时测试三个
设定假设 - “将CTA从‘开始使用’改为‘开始我的免费试用’会提升点击率，因为它暗示了所有权和明确性”
先定义指标 - 点击率、转化率或营收
运行至统计显著性 - 每个变体至少完成100次转化；在宣布获胜者前使用显著性计算器
记录经验 - 成功的见解会不断积累。为你的受众建立一个有效的文案素材库
迭代而非转型 - 如果变体B获胜，将其设为对照组，然后测试变体C。微小的改进会随着时间推移不断积累

主题行测试：20%流量给变体A，20%给变体B，60%留给获胜者（4小时内发送）。

Anti-patterns

反模式

Anti-pattern	Why it fails	Fix
Feature-first copy	Readers don't care about features, they care about their life after buying	Lead with the outcome; features are proof the outcome is possible
Jargon overload	Industry terms create distance and confusion outside expert audiences	Use the simplest word that carries the meaning; test with someone outside the industry
Weak CTAs ("Submit", "Click here")	Generic CTAs provide no motivation to act	Use first-person, benefit-specific verbs ("Get My Free Report")
Multiple CTAs per page	Splits attention and decision energy; often reduces total conversions	Pick the single most valuable action and optimize the entire page for it
Burying the benefit	Long intros before the value prop cause readers to bounce before they understand the offer	State the primary benefit in the headline or first sentence
Superlative abuse ("best", "world-class", "revolutionary")	Unsupported superlatives signal low credibility; every competitor says the same	Replace with specific, verifiable claims and social proof

反模式	失败原因	修复方法
以功能为核心的文案	读者不关心功能，他们关心购买后的生活变化	以结果开头；功能只是实现结果的证明
术语过载	行业术语会与非专业受众产生距离并造成困惑	使用能表达意思的最简单词汇；让行业外的人测试文案
弱CTA（如“提交”、“点击这里”）	通用CTA无法提供行动动机	使用第一人称、明确利益的动词（如“获取我的免费报告”）
单页多CTA	分散注意力和决策精力；通常会降低总转化率	选择最有价值的单一行动，并优化整个页面以实现该行动
隐藏利益点	在价值主张前添加冗长的引言，会导致读者在理解优惠前就离开	在标题或第一句话中陈述核心利益
滥用最高级（如“最佳”、“世界级”、“革命性”）	无依据的最高级会降低可信度；每个竞争对手都这么说	替换为具体、可验证的声明和社交证明

References

参考资料

For ready-to-use templates, formulas, and structures, load the reference file:

```
references/swipe-file.md
```
- Headline formulas, CTA templates, and landing page structures ready to adapt and deploy

Only load the reference file when the task requires specific templates or examples rather than framework guidance.

如需现成可用的模板、公式和结构，请加载参考文件：

```
references/swipe-file.md
```
- 可直接改编使用的标题公式、CTA模板和着陆页结构

仅当任务需要特定模板或示例而非框架指导时，才加载参考文件。