cro

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

CRO (Conversion Rate Optimization)

CRO（转化率优化）

When to Use This Skill

技能适用场景

Activate this skill when the user's request involves any of the following:

Auditing a landing page for conversion performance
Designing or improving a landing page layout, copy, or user flow
Setting up, analyzing, or interpreting A/B tests or multivariate tests
Optimizing web forms (lead gen, signup, contact, application)
Designing or auditing pricing pages and pricing presentation
Reducing cart abandonment or improving checkout completion rates
Improving any website conversion metric (lead form submissions, signups, purchases, trial starts)
Calculating sample sizes, test duration, or statistical significance for experiments
Prioritizing which conversion improvements to tackle first
Diagnosing why a page or funnel has a low conversion rate
Asking about trust signals, social proof, urgency elements, or CTA optimization

当用户的请求涉及以下任一情况时，激活本技能：

审核着陆页的转化表现
设计或改进着陆页布局、文案或用户流程
设置、分析或解读A/B测试或多变量测试
优化网页表单（潜在客户开发、注册、联系、申请类表单）
设计或审核定价页面及定价展示方式
降低购物车弃购率或提升结账完成率
提升任意网站转化指标（表单提交、注册、购买、试用启动）
计算实验所需的样本量、测试时长或统计显著性
优先确定需处理的转化优化事项
诊断页面或转化漏斗转化率低的原因
询问信任信号、社交证明、紧迫感元素或CTA优化相关问题

Brand Context (Auto-Applied)

品牌上下文（自动应用）

Before producing any marketing output from this module:

Check session context — The active brand summary was output at session start. Use the brand name, industry, voice settings, channels, goals, compliance, and competitors shown there.

If you need the full profile, read:

~/.claude-marketing/brands/{slug}/profile.json

Apply brand voice — Formality, energy, humor, authority levels must shape all content tone and word choices
Check compliance — Auto-apply rules for brand's target_markets and industry using
```
skills/context-engine/compliance-rules.md
```
Reference industry benchmarks — Consult
```
skills/context-engine/industry-profiles.md
```
for the brand's industry
Use platform specs — Reference
```
skills/context-engine/platform-specs.md
```
for character limits and format requirements

Check campaign history — Run

python campaign-tracker.py --brand {slug} --action list-campaigns

before planning new work

If no brand exists, say: "No brand profile found. Use /digital-marketing-pro:brand-setup to create one, or I can proceed with general best practices."
Check brand guidelines — If
```
~/.claude-marketing/brands/{slug}/guidelines/_manifest.json
```
exists, load and enforce:
```
restrictions.md
```
for banned words, restricted claims, and mandatory disclaimers;
```
channel-styles.md
```
for channel-specific tone overrides (may differ from base voice);
```
messaging.md
```
for approved key messages, taglines, and positioning language;
```
voice-and-tone.md
```
for detailed voice rules beyond the 4 numeric scores. If producing content for a specific channel, channel style rules take precedence over base voice settings.

Do not ask the user for information that already exists in their brand profile.

在使用本模块生成任何营销内容之前：

检查会话上下文 —— 会话开始时已输出活跃品牌摘要。使用其中显示的品牌名称、行业、语气设置、渠道、目标、合规要求和竞品信息。

如需完整品牌档案，请读取：

~/.claude-marketing/brands/{slug}/profile.json

应用品牌语气 —— 正式程度、活力、幽默感、权威性必须塑造所有内容的语气和用词选择
检查合规性 —— 使用
```
skills/context-engine/compliance-rules.md
```
自动应用品牌目标市场和行业的规则
参考行业基准 —— 查阅
```
skills/context-engine/industry-profiles.md
```
获取品牌所在行业的相关信息
使用平台规范 —— 参考
```
skills/context-engine/platform-specs.md
```
了解字符限制和格式要求

检查营销活动历史 —— 在规划新工作前运行

python campaign-tracker.py --brand {slug} --action list-campaigns

若无品牌档案，请告知："未找到品牌档案。使用/digital-marketing-pro:brand-setup创建档案，或我可基于通用最佳实践继续操作。"
检查品牌指南 —— 若
```
~/.claude-marketing/brands/{slug}/guidelines/_manifest.json
```
存在，请加载并执行：
```
restrictions.md
```
中的禁用词汇、受限声明和强制免责条款；
```
channel-styles.md
```
中的渠道特定语气覆盖规则（可能与基础语气不同）；
```
messaging.md
```
中的已批准关键信息、标语和定位语言；
```
voice-and-tone.md
```
中超出4项数值评分的详细语气规则。若为特定渠道生成内容，渠道风格规则优先于基础语气设置。

请勿向用户询问其品牌档案中已有的信息。

Required Context

必要上下文

Before executing, gather the following from the user (ask if not provided):

Page URL or description: The specific page or flow being optimized
Current conversion rate: Baseline metric to improve against (if known)
Monthly traffic volume: Needed for test duration and statistical significance calculations
Conversion goal: What counts as a conversion (form submit, purchase, signup, download)
Business model: B2B, B2C, D2C, SaaS, ecommerce, lead gen
Traffic sources: Where visitors come from (paid, organic, email, direct) since source affects intent level
Device split: Percentage of mobile vs desktop traffic
Existing test history: What has been tested before and results
Tech stack: CMS, testing tools (Optimizely, VWO, Google Optimize successor, custom), analytics platform
Constraints: Legal disclaimers required, brand guidelines, compliance restrictions

执行前，请从用户处收集以下信息（若未提供则询问）：

页面URL或描述：需要优化的特定页面或流程
当前转化率：需改进的基准指标（若已知）
月流量：用于计算测试时长和统计显著性
转化目标：什么行为算作转化（表单提交、购买、注册、下载）
商业模式：B2B、B2C、D2C、SaaS、电商、潜在客户开发
流量来源：访客来源（付费、自然搜索、邮件、直接访问），因为来源会影响用户意图
设备分布：移动端与桌面端流量占比
已有测试历史：之前进行过的测试及结果
技术栈：CMS、测试工具（Optimizely、VWO、Google Optimize替代工具、自定义工具）、分析平台
约束条件：所需法律免责声明、品牌指南、合规限制

Capabilities

核心能力

Landing Page Audits

着陆页审核

5-second test: Does the page communicate its value proposition within 5 seconds of loading?
Above-the-fold analysis: Headline clarity, subheadline support, hero image relevance, primary CTA visibility
Trust signal inventory: Logos, testimonials, reviews, certifications, security badges, guarantees
CTA assessment: Clarity, contrast, placement, copy specificity, number of competing CTAs
Form evaluation: Field count, field labels, required vs optional, error handling, multi-step vs single-step
Page speed impact: Load time correlation to bounce rate, Core Web Vitals as conversion factors
Mobile experience: Touch targets, scroll depth, thumb-zone CTA placement, mobile-specific friction points
Content hierarchy: Information architecture, visual hierarchy, F-pattern or Z-pattern scanning support
Objection handling: Whether the page addresses common objections before the conversion point

5-second test：页面加载后能否在5秒内传达其价值主张？
首屏分析：标题清晰度、副标题支撑性、主图相关性、主CTA可见性
信任信号盘点：品牌标志、客户 testimonial、评论、认证、安全徽章、保证条款
CTA评估：清晰度、对比度、位置、文案具体性、竞争性CTA数量
表单评估：字段数量、字段标签、必填/可选设置、错误处理、多步骤/单步骤形式
页面速度影响：加载时间与跳出率的相关性，Core Web Vitals作为转化影响因素
移动端体验：触摸目标大小、滚动深度、拇指区CTA位置、移动端特定摩擦点
内容层级：信息架构、视觉层级、对F型或Z型浏览模式的支持
异议处理：页面是否在转化节点前解决常见用户异议

A/B Testing Framework

A/B测试框架

ICE prioritization: Score potential tests by Impact (1-10), Confidence (1-10), and Ease (1-10) to determine test order
Hypothesis format: Structured as "If we [change], then [metric] will [improve/decrease] because [rationale]"
Sample size calculation: Based on baseline conversion rate, minimum detectable effect, statistical power (80%), and significance level (95%)
Test duration estimation: Accounting for traffic volume, conversion rate, and full business cycles (minimum 1-2 weeks to capture weekly patterns)
Result interpretation: Statistical significance, practical significance, segment analysis, and revenue impact projection
Sequential testing: When to use fixed-horizon vs sequential/Bayesian methods for faster decisions

ICE优先级排序：按Impact（1-10分）、Confidence（1-10分）、Ease（1-10分）对潜在测试打分，确定测试顺序
假设格式：结构化表述为"如果我们[做出某项改变]，那么[某指标]将[提升/下降]，因为[基于证据的理由]"
样本量计算：基于基准转化率、最小可检测效果、统计功效（80%）和显著性水平（95%）
测试时长估算：考虑流量规模、转化率和完整业务周期（至少1-2周以捕捉周度模式）
结果解读：统计显著性、实际显著性、细分分析、收入影响预测
序贯测试：何时使用固定周期 vs 序贯/贝叶斯方法以加快决策

Form Optimization

表单优化

Field reduction: Remove or defer non-essential fields. Each additional field reduces conversion by approximately 4-7%
Progressive profiling: Collect information across multiple interactions rather than all at once
Inline validation: Real-time feedback as users complete fields reduces form abandonment
Smart defaults: Pre-fill known data, use sensible defaults, and provide auto-complete
Multi-step forms: Break long forms into logical steps with progress indicators
Field type optimization: Dropdowns vs radio buttons vs text inputs based on option count and context
Error messaging: Specific, helpful error messages positioned near the relevant field

字段精简：移除或延迟非必要字段。每增加一个字段，转化率约降低4-7%
渐进式信息收集：在多次交互中收集信息，而非一次性收集全部
在线验证：用户填写字段时提供实时反馈，减少表单弃填
智能默认值：预填已知数据、使用合理默认值、提供自动补全
多步骤表单：将长表单拆分为逻辑步骤并添加进度指示器
字段类型优化：根据选项数量和上下文选择下拉框、单选按钮或文本输入框
错误提示：在相关字段附近显示具体、有用的错误信息

Pricing Psychology

定价心理学

Anchoring: Present a higher-priced option first to make target option seem reasonable
Decoy effect: Introduce a strategically inferior option to push users toward the target plan
Charm pricing: $99 vs $100 -- when it works (B2C, impulse purchases) and when it backfires (premium B2B)
Price framing: Annual vs monthly display, per-user vs flat rate, daily equivalency ("less than a cup of coffee")
Plan naming: Naming conventions that guide self-selection (Starter/Growth/Enterprise vs Basic/Pro/Premium)
Feature differentiation: Which features to gate at each tier to create natural upgrade pressure
Social proof on pricing: Showing "Most Popular" badges, customer counts per tier, or logos

锚定效应：先展示高价选项，使目标选项显得更合理
诱饵效应：引入策略性劣势选项，推动用户选择目标方案
魅力定价：$99 vs $100——适用场景（B2C、冲动购买）和反效果场景（高端B2B）
价格呈现：年费vs月费展示、按用户收费vs固定费率、每日等效价（"不到一杯咖啡的价格"）
方案命名：引导自我选择的命名规则（Starter/Growth/Enterprise vs Basic/Pro/Premium）
功能差异化：在各层级设置哪些功能以创造自然升级动力
定价页社交证明：显示"最受欢迎"徽章、各层级客户数量或品牌标志

Checkout Optimization

结账流程优化

Cart abandonment reduction: Exit-intent offers, cart recovery emails, progress indicators, persistent cart
Guest checkout: Always offer guest checkout; forced account creation causes 24% abandonment
Payment method coverage: Credit cards, PayPal, Apple Pay, Google Pay, Buy Now Pay Later, regional methods
Shipping transparency: Show costs early, offer free shipping thresholds, provide delivery estimates
Order summary persistence: Keep order details visible throughout checkout
Security reinforcement: SSL badges, payment logos, money-back guarantees at the point of payment
One-page vs multi-step checkout: Decision framework based on product complexity and information requirements
Post-purchase optimization: Confirmation page upsells, order confirmation email, account creation after purchase

降低购物车弃购率：退出意图优惠、购物车恢复邮件、进度指示器、持久化购物车
访客结账：始终提供访客结账选项；强制注册会导致24%的弃购率
支付方式覆盖：信用卡、PayPal、Apple Pay、Google Pay、先买后付、区域支付方式
运费透明度：提前显示运费、提供免运费门槛、给出配送预估
订单摘要持久化：在整个结账过程中保持订单详情可见
安全性强化：SSL徽章、支付标志、付款节点的退款保证
单页vs多步骤结账：基于产品复杂度和信息需求的决策框架
购后优化：确认页追加销售、订单确认邮件、购后引导注册账户

Process

操作流程

Standard Landing Page Audit (Most Common Use Case)

标准着陆页审核（最常见场景）

5-second scan -- Review the page as a first-time visitor. Can you identify what the company does, who it is for, and what action to take within 5 seconds?
Above-the-fold audit -- Evaluate headline specificity, subheadline support, hero relevance, and CTA prominence. The fold is the most valuable real estate.
Trust and credibility -- Inventory all trust signals (testimonials, logos, reviews, certifications, guarantees). Identify gaps where social proof is missing at critical decision points.
CTA analysis -- Count all CTAs on the page. Check for competing actions, button copy specificity ("Get My Free Trial" beats "Submit"), contrast ratio, and placement frequency.
Content flow -- Walk through the page section by section. Does it follow a logical persuasion sequence? Problem, solution, proof, action?
Form/conversion point -- Evaluate the form or conversion mechanism. Count fields, assess labels, check error handling, and evaluate the micro-copy around the submit button.
Mobile audit -- Review the same page on mobile. Check touch targets (minimum 44x44px), scroll depth to CTA, horizontal scrolling issues, and load time.
Speed check -- Note any visible performance issues. Recommend Core Web Vitals audit if speed appears to be a factor.
Prioritized recommendations -- Deliver findings as a prioritized list using the ICE framework. Quick wins first, structural changes second, redesign-level changes last.

5秒扫描 —— 以首次访客的身份浏览页面。能否在5秒内识别出公司业务、目标受众以及应采取的行动？
首屏审核 —— 评估标题具体性、副标题支撑性、主图相关性和CTA突出性。首屏是最有价值的页面区域。
信任与可信度 —— 盘点所有信任信号（客户评价、品牌标志、评论、认证、保证条款）。找出关键决策点缺失社交证明的缺口。
CTA分析 —— 统计页面上的所有CTA。检查是否存在竞争性操作、按钮文案具体性（"获取免费试用"优于"提交"）、对比度和出现频率。
内容流程 —— 逐段浏览页面。是否遵循逻辑说服序列？问题、解决方案、证据、行动？
表单/转化节点 —— 评估表单或转化机制。统计字段数量、评估标签、检查错误处理、评估提交按钮周围的微文案。
移动端审核 —— 在移动端查看同一页面。检查触摸目标（最小44x44px）、到达CTA的滚动深度、横向滚动问题和加载时间。
速度检查 —— 记录任何可见的性能问题。若速度似乎是影响因素，建议进行Core Web Vitals审核。
优先级推荐 —— 使用ICE框架将发现整理为优先级列表。先快速见效的优化，再结构调整，最后重新设计级别的变更。

A/B Test Design Process

A/B测试设计流程

Identify the problem -- Use data (analytics, heatmaps, session recordings, user feedback) to pinpoint the conversion bottleneck.
Form hypothesis -- Write a structured hypothesis: "If we [change X], then [metric Y] will [increase/decrease] by [estimated amount] because [rationale based on evidence]."
Score with ICE -- Rate Impact, Confidence, and Ease on a 1-10 scale. Prioritize tests with highest composite scores.
Calculate requirements -- Determine sample size needed based on current conversion rate and minimum detectable effect. Estimate test duration based on daily traffic.
Design variation -- Create the test variation. Change only one variable per test (unless running a multivariate test with sufficient traffic).
QA the test -- Verify tracking, check both variations across devices and browsers, confirm that the test does not break downstream flows.
Run and monitor -- Launch the test. Do not peek at results before reaching calculated sample size. Monitor for technical issues only.
Analyze and document -- At the end of the test, evaluate statistical significance, check segment-level results, calculate revenue impact, and document learnings regardless of outcome.

识别问题 —— 使用数据（分析、热图、会话录制、用户反馈）定位转化瓶颈。
形成假设 —— 撰写结构化假设："如果我们[做出X改变]，那么[指标Y]将[增加/减少][预估幅度]，因为[基于证据的理由]。"
ICE评分 —— 对Impact、Confidence、Ease按1-10分评分。优先选择综合得分最高的测试。
计算需求 —— 根据当前转化率和最小可检测效果确定所需样本量。根据日流量估算测试时长。
设计变体 —— 创建测试变体。每次测试仅更改一个变量（除非流量充足可运行多变量测试）。
测试QA —— 验证跟踪设置、在不同设备和浏览器上检查两个变体、确认测试不会破坏后续流程。
运行与监控 —— 启动测试。在达到计算的样本量前不要查看结果。仅监控技术问题。
分析与记录 —— 测试结束后，评估统计显著性、检查细分群体结果、计算收入影响，并无论结果如何都记录经验教训。

Reference Files

参考文件

```
landing-page-audit.md
```
-- Detailed audit checklist, scoring rubric, and benchmark conversion rates by industry
```
ab-testing.md
```
-- Test design templates, sample size calculators, statistical methods, and common testing pitfalls
```
form-optimization.md
```
-- Field-by-field optimization guide, progressive profiling implementation, and form UX patterns
```
pricing-psychology.md
```
-- Pricing page templates, psychological principles with examples, and tier structure frameworks
```
checkout-optimization.md
```
-- Cart abandonment diagnosis, checkout flow templates, and payment optimization strategies

```
landing-page-audit.md
```
—— 详细审核清单、评分标准和各行业基准转化率
```
ab-testing.md
```
—— 测试设计模板、样本量计算器、统计方法和常见测试陷阱
```
form-optimization.md
```
—— 逐字段优化指南、渐进式信息收集实施方法和表单UX模式
```
pricing-psychology.md
```
—— 定价页面模板、带示例的心理学原理和层级结构框架
```
checkout-optimization.md
```
—— 购物车弃购诊断、结账流程模板和支付优化策略

Output Formats

输出格式

Landing page audit report: Section-by-section findings with severity ratings (critical/high/medium/low), screenshots or references to specific elements, and prioritized action items with ICE scores
A/B test plan: Hypothesis, variation description, sample size requirements, estimated duration, success criteria, and segmentation plan
Form optimization spec: Current vs recommended field list, layout wireframe description, validation rules, and error message copy
Pricing page recommendation: Tier structure, pricing presentation, feature matrix, and psychological triggers with rationale
Checkout optimization plan: Funnel stage analysis, drop-off diagnosis, and ordered list of improvements with expected impact

着陆页审核报告：分章节的发现，包含严重程度评级（关键/高/中/低）、特定元素的截图或引用、带ICE评分的优先级行动项
A/B测试计划：假设、变体描述、样本量要求、预估时长、成功标准和细分计划
表单优化规范：当前字段vs推荐字段列表、布局线框描述、验证规则和错误提示文案
定价页面建议：层级结构、定价展示、功能矩阵和带理由的心理学触发因素
结账优化计划：漏斗阶段分析、流失诊断、带预期影响的有序改进列表

Edge Cases

边缘场景

Low Traffic Sites

低流量网站

Sites with fewer than 1,000 monthly conversions often cannot reach statistical significance within a reasonable timeframe. For these sites, skip A/B testing and instead implement best practices directly based on audit findings. Use before/after measurement with awareness of confounding variables. Consider qualitative methods (user testing with 5 users catches 85% of usability issues) over quantitative testing.

月转化量少于1000的网站通常无法在合理时间内达到统计显著性。对于此类网站，跳过A/B测试，直接根据审核结果实施最佳实践。使用前后对比测量，同时注意混淆变量。考虑采用定性方法（5位用户的可用性测试可发现85%的可用性问题）而非定量测试。

B2B Long-Form Pages vs B2C Short-Form

B2B长页面vs B2C短页面

B2B landing pages often need to be longer because purchase decisions involve multiple stakeholders, higher price points, and longer evaluation cycles. Do not default to "shorter is better." Instead, ensure the above-the-fold section qualifies intent quickly, and let the rest of the page handle objections comprehensively. B2C impulse purchases benefit from short, fast, single-CTA pages.

B2B着陆页通常需要更长篇幅，因为购买决策涉及多个利益相关者、更高价格点和更长评估周期。不要默认"越短越好"。相反，确保首屏部分快速筛选用户意图，页面其余部分全面处理异议。B2C冲动购买受益于简短、快速、单CTA的页面。

Mobile-First vs Desktop-First Optimization

移动端优先vs桌面端优先优化

Check the device split before making recommendations. If 70%+ of traffic is mobile, optimize for mobile first and ensure desktop does not break. If traffic is desktop-dominant (common in B2B), optimize for desktop but never neglect mobile. The "responsive" middle ground often serves neither well.

在提出建议前检查设备分布。若70%以上流量来自移动端，优先优化移动端并确保桌面端功能正常。若流量以桌面端为主（B2B常见），优化桌面端但绝不能忽视移动端。"响应式"折中方案往往无法很好地服务任何一端。

Regulated Industries with Required Disclaimers

需强制免责声明的监管行业

Healthcare, finance, legal, and insurance pages often require lengthy disclaimers, disclosures, or compliance text. Do not recommend removing these. Instead, work on formatting (collapsible sections, footnotes, smaller but readable type) and ensure the required content does not visually compete with the primary CTA. Position disclaimers after the conversion point where legally permissible.

医疗、金融、法律和保险页面通常需要冗长的免责声明、披露或合规文本。不要建议移除这些内容。相反，优化格式（可折叠章节、脚注、小但可读的字体）并确保所需内容不会在视觉上与主CTA竞争。在法律允许的情况下，将免责声明放在转化节点之后。

Testing During Seasonal Peaks

季节性高峰期间测试

Avoid launching A/B tests during Black Friday, holiday seasons, or major promotional periods. User behavior during peaks is not representative of normal behavior. Tests run during these periods will produce unreliable results. If a test must run during a peak, note the caveat and plan a validation retest during a normal period.

避免在黑色星期五、节假日或大型促销期间启动A/B测试。高峰期间的用户行为无法代表正常行为。在此期间运行的测试会产生不可靠的结果。若必须在高峰期间测试，需注明局限性并计划在正常时期进行验证性复测。

Related Skills

Agents Used

使用的Agent

cro-specialist — Primary agent for all CRO tasks: landing page audits, A/B test design, form optimization, pricing psychology, checkout optimization, statistical significance calculations, and experiment documentation

cro-specialist —— 所有CRO任务的主要Agent：着陆页审核、A/B测试设计、表单优化、定价心理学、结账流程优化、统计显著性计算和实验文档撰写

cro