marketing-automation · diff

git:20260412.de03456 to v2.0

178 added, 301 removed. Audit A to A.

---

name: marketing-automation

description: >

- Generate general product and growth marketing deliverables across CRO,

- copywriting, SEO, analytics, lifecycle messaging, pricing, launches, and

- experiment planning using 23 specialized sub-skills. Use when the user needs

- broad marketing help for websites, SaaS, funnels, email, growth, retention,

- or analytics workflows, even if they ask for a launch plan, onboarding copy,

- A/B tests, or SEO assets. Prefer more specific skills for niche domains such

- as Steam game store-page / festival launch ops.

+ Route broad product and growth marketing requests into one clear deliverable lane

+ across CRO, copywriting, SEO, lifecycle messaging, analytics, pricing, launch,

+ retention, and experiment planning. Use when the user needs general marketing

+ help for a website, SaaS product, funnel, launch, onboarding flow, campaign, or

+ reporting question and the first job is to turn a messy ask into a KPI-aware

+ brief, a chosen marketing lane, and an implementation-ready packet. Triggers on:

+ marketing plan, growth help, launch messaging, onboarding emails, landing page

+ conversion, SEO brief, pricing copy, campaign plan, funnel optimization.

+ allowed-tools: Write Read WebSearch WebFetch Task

+ compatibility: >

+ General marketing router for SaaS, product, web, and lifecycle work. Prefer

+ narrower skills when the request is already obviously domain-specific.

metadata:

tags: marketing, cro, copywriting, seo, analytics, growth, automation

platforms: Claude, ChatGPT, Gemini, Codex

+ version: "2.0"

---

- # Marketing Skills Collection

+ # Marketing Automation

- A collection of 23 sub-skills for marketing deliverables. Generates repeatable, high-quality outputs across CRO, copywriting, SEO, analytics, and growth.

+ Use this skill as the **canonical general marketing router** for the repository.

- ## When to use this skill

+ The job is not to dump generic growth advice. The job is to:

+ 1. normalize a messy request into a usable marketing brief,

+ 2. pick the single best marketing lane,

+ 3. produce the next deliverable packet,

+ 4. leave a measurement-aware implementation handoff.

- - **Marketing Deliverables Needed**: CRO, copy, SEO, analytics, growth outputs

- - **Repeatable High-Quality Outputs**: Generate assets tuned to a single KPI

- - **Business Goal → Skill Mapping**: Convert goals to specific marketing sub-skills

+ Read [references/routing-heuristics.md](references/routing-heuristics.md) for lane-selection rules and [references/measurement-handoff.md](references/measurement-handoff.md) before routing broad requests.

- ---

+ ## When to use this skill

+ - The user asks for broad marketing help and the correct lane is not yet obvious

+ - A product, website, funnel, or launch needs one structured marketing brief before detailed execution starts

+ - The request spans multiple possible lanes such as CRO, copy, lifecycle, SEO, analytics, pricing, or launch planning

+ - The user wants a KPI-aware packet that a narrower marketing or execution workflow could pick up next

- ## 23 Sub-Skills Overview

+ ## When not to use this skill

+ - The request is already clearly Steam/store-page/festival launch work for a game → prefer `steam-store-launch-ops`

+ - The request is already a narrow SEO/content optimization workflow with enough detail to stay specialist

+ - The request is really product planning, UX research, or technical implementation rather than marketing routing

+ - The user wants only one atomic copy rewrite and no broader routing or measurement layer

- ### CRO (Conversion Rate Optimization)

- | Sub-Skill | Description |

- |-----------|-------------|

- | `page-cro` | Landing page conversion rate optimization |

- | `signup-flow` | Signup flow optimization |

- | `onboarding` | Onboarding experience improvement |

- | `form-optimization` | Form optimization (fields, UX) |

- | `paywall` | Paywall/pricing page optimization |

+ ## Supported lanes

+ Use these as routing targets inside the skill:

- ### Copywriting

- | Sub-Skill | Description |

- |-----------|-------------|

- | `copywriting` | Ad/marketing copy writing |

- | `copy-editing` | Improve existing copy |

- | `email-sequence` | Email sequence design |

- | `social-content` | Social media content |

+ ### CRO

+ - `page-cro`

+ - `signup-flow`

+ - `onboarding`

+ - `form-optimization`

+ - `paywall`

- ### SEO

- | Sub-Skill | Description |

- |-----------|-------------|

- | `seo-audit` | SEO audit and improvements |

- | `programmatic-seo` | Programmatic SEO pages |

- | `comparison-page` | Comparison page writing |

- | `schema-markup` | Structured data markup |

+ ### Copy & messaging

+ - `copywriting`

+ - `copy-editing`

+ - `email-sequence`

+ - `social-content`

- ### Ads & Analytics

- | Sub-Skill | Description |

- |-----------|-------------|

- | `analytics-tracking` | Analytics tracking setup |

- | `paid-ads` | Paid ads strategy/copy |

- | `ab-test` | A/B test design |

+ ### SEO & content

+ - `seo-audit`

+ - `programmatic-seo`

+ - `comparison-page`

+ - `schema-markup`

- ### Strategy & Growth

- | Sub-Skill | Description |

- |-----------|-------------|

- | `launch-strategy` | Product launch strategy |

- | `pricing-strategy` | Pricing strategy |

- | `retention` | Retention improvement strategy |

- | `churn-analysis` | Churn analysis |

- | `growth-experiments` | Growth experiment design |

- | `referral-program` | Referral program design |

- | `content-strategy` | Content strategy |

+ ### Ads & analytics

+ - `analytics-tracking`

+ - `paid-ads`

+ - `ab-test`

- ---

+ ### Strategy & growth

+ - `launch-strategy`

+ - `pricing-strategy`

+ - `retention`

+ - `churn-analysis`

+ - `growth-experiments`

+ - `referral-program`

+ - `content-strategy`

## Instructions

- ### Step 1: Define Objective and Constraints

+ ### Step 1: Normalize the marketing brief

+ Capture the request in this form before choosing a lane:

```yaml

marketing_brief:

- objective: [single KPI - conversion rate, CTR, activation]

- target_audience:

- segment: [customer segment]

- pain_points: [key pain points]

- terminology: [terms they use]

- channel: [LP, email, social, SEO, ads]

- format: [format]

+ primary_kpi: "the one number that matters most right now"

+ audience:

+ segment: "who is this for"

+ pain_points:

+ - pain 1

+ - pain 2

offer:

- value_prop: [value proposition]

- positioning: [positioning]

- proof_points: [proof points]

- ```

- ### Step 2: Select the Sub-Skill

- Select the appropriate sub-skill for the situation:

- ```bash

- # When CRO is needed

- → page-cro, signup-flow, onboarding, form-optimization, paywall

- # When copy is needed

- → copywriting, copy-editing, email-sequence, social-content

- # When SEO is needed

- → seo-audit, programmatic-seo, comparison-page, schema-markup

- # When ads/analytics are needed

- → analytics-tracking, paid-ads, ab-test

- # When strategy/growth is needed

- → launch-strategy, pricing-strategy, retention, churn-analysis, growth-experiments

+ product: "product / feature / package"

+ value_prop: "main promise"

+ proof_points:

+ - proof 1

+ - proof 2

+ constraints:

+ timeline: immediate | this-week | this-month | longer

+ brand_or_compliance_notes: "voice, legal, or positioning limits"

+ main_question: "what does the user actually need next?"

```

- ### Step 3: Build the Prompt

- Build structured prompt:

- ```markdown

- ## Marketing Asset Request

- ### Product Context

- - **Product**: [product name]

- - **Category**: [category]

- - **Stage**: [stage - early, growth, mature]

- ### Audience

- - **Segment**: [target segment]

- - **Pain Points**: [1-3 pain points]

- - **Current State**: [solution they currently use]

+ If the packet is incomplete, continue with explicit assumptions instead of stalling.

- ### Offer

- - **Value Prop**: [core value proposition]

- - **Differentiator**: [differentiation point]

- - **Proof**: [trust elements - numbers, clients, awards]

+ ### Step 2: Pick one primary lane

+ Choose exactly one lane.

- ### Constraints

- - **Tone**: [tone - professional, casual, bold]

- - **Brand Voice**: [brand voice guide]

- - **Do NOT**: [things to avoid]

+ - **CRO** when the request is about improving a page, flow, form, or paywall

+ - **Copy & messaging** when the main bottleneck is words, positioning, or conversion copy

+ - **SEO & content** when the request centers on search demand, structure, comparison pages, or content strategy

+ - **Ads & analytics** when the request is about tracking, paid acquisition, or test/readout setup

+ - **Strategy & growth** when the user needs launch planning, pricing, retention, churn, referral, or experiment prioritization

- ### Output Format

- - [desired format - table, checklist, bullets]

- ```

+ If more than one lane is relevant, choose the lane that reduces ambiguity fastest and mention the secondary handoff.

- ### Step 4: Generate and Validate

+ ### Step 3: Return one implementation-ready packet

+ Return one of these deliverable types:

+ - `marketing-routing brief`

+ - `channel-ready brief`

+ - `copy/messaging packet`

+ - `measurement + experiment packet`

+ - `launch/growth brief`

- ```bash

- # Generate

- claude task "generate marketing asset with sub-skill name"

+ Do not emit five half-finished plans. Choose the single most useful packet for the current ask.

- # Validation checklist

- - [ ] KPI alignment

- - [ ] Target audience fit

- - [ ] Brand voice consistency

- - [ ] Actionability

- ```

+ ### Step 4: Add explicit handoff and measurement logic

+ Every output must include:

+ - the chosen lane,

+ - the primary KPI,

+ - what assumptions remain,

+ - what execution artifact should be created next,

+ - what to measure after implementation.

- ### Step 5: Handoff + Measurement

+ ### Step 5: Use this output structure

```markdown

- ## Implementation Checklist

- - [ ] Asset publishing

- - [ ] Tracking event setup

- - [ ] Success threshold definition

- ## Tracking Events

- | Event | Description | Success Threshold |

- |-------|-------------|-------------------|

- | page_view | Page view | baseline |

- | cta_click | CTA click | +20% vs control |

- | signup_complete | Signup complete | +15% vs control |

- ## A/B Test Proposals

- 1. [Hypothesis 1]: [variant] vs [control]

- 2. [Hypothesis 2]: [variant] vs [control]

- ```

- ---

- ## Examples

- ### Example 1: Landing Page CRO

- **Prompt**:

- ```

- Optimize the landing page for higher signup conversion.

- Audience: indie founders building side projects.

- Offer: AI co-pilot for product launches.

- Output: prioritized CRO changes + A/B tests.

- ```

- **Expected output**:

- - CRO checklist prioritized by impact/effort

- - 3 A/B test hypotheses with expected lift

- - Hero + CTA copy suggestions (3 variants each)

- ### Example 2: Email Sequence

- **Prompt**:

- ```

- Create a 5-email welcome sequence for a B2B SaaS.

- Audience: ops managers at 50-500 employee companies.

- Goal: drive first workflow setup within 7 days.

- ```

+ # Marketing Routing Brief

- **Expected output**:

- ```markdown

- ## Welcome Sequence

+ ## Scope

+ - Objective: ...

+ - Primary KPI: ...

+ - Audience: ...

+ - Surface / channel: ...

+ - Confidence: high | medium | low

- ### Email 1: Welcome (Day 0)

- - **Subject**: Welcome to [Product] - Let's get started

- - **Goal**: Account confirmation + quick win

- - **CTA**: Complete profile

+ ## Chosen lane

+ - CRO | Copy & messaging | SEO & content | Ads & analytics | Strategy & growth

- ### Email 2: Value Demo (Day 1)

- - **Subject**: See what [Product] can do in 2 minutes

- - **Goal**: Feature awareness

- - **CTA**: Watch demo video

+ ## What matters most now

+ - 2-4 bullets

- ### Email 3: First Workflow (Day 3)

- - **Subject**: Create your first workflow (step-by-step)

- - **Goal**: Activation milestone

- - **CTA**: Create workflow

+ ## Recommended packet

+ - One of: marketing-routing brief | channel-ready brief | copy/messaging packet | measurement + experiment packet | launch/growth brief

- ### Email 4: Use Case (Day 5)

- - **Subject**: How [Customer] saved 10 hours/week

- - **Goal**: Social proof + inspiration

- - **CTA**: Try this template

+ ## Priority decisions

+ |----------|---------|-------|-----------------|

+ | ... | ... | ... | ... |

- ### Email 5: Check-in (Day 7)

- - **Subject**: Need help getting started?

- - **Goal**: Rescue non-activated users

- - **CTA**: Book a call / Reply for help

+ ## Immediate next steps

+ 1. ...

+ 2. ...

+ 3. ...

- ## Metrics

- | Email | Open Rate Target | CTR Target |

- |-------|------------------|------------|

- | Email 1 | 60%+ | 30%+ |

- | Email 2 | 45%+ | 15%+ |

- | Email 3 | 40%+ | 20%+ |

- | Email 4 | 35%+ | 12%+ |

- | Email 5 | 40%+ | 15%+ |

- ```

+ ## Secondary handoffs

+ - Skill / workflow: ...

+ - Why: ...

+ - What packet to pass: ...

- ### Example 3: Programmatic SEO

+ ## Measurement checklist

+ - Event / signal: ...

+ - Baseline or assumption: ...

+ - Success threshold: ...

- **Prompt**:

- ```

- Create a programmatic SEO template for comparison pages.

- Target: "[Tool A] vs [Tool B]" searches.

- Include: H1, meta description, comparison table, CTA.

+ ## What not to do yet

+ - 1-3 bullets that prevent scope drift or premature channel sprawl

```

- **Expected output**:

- - Page template with placeholders

- - Schema markup (JSON-LD)

- - Internal linking strategy

- - Content guidelines per section

- ---

- ## Best practices

- 1. **One KPI per deliverable**: Avoid mixed objectives

- 2. **Audience specificity**: Use segment-specific needs and terminology

- 3. **Instrument measurement**: Set up tracking before launch

- 4. **Iterate with data**: Treat outputs as hypotheses

- ---

- ## Common pitfalls

- - **Mixed Objectives**: Multiple goals in one asset

- - **Missing Audience Context**: Unclear who it's for

- - **No Tracking/Validation Plan**: Unable to measure impact

- ---

- ## Troubleshooting

- ### Issue: Output is generic

- **Cause**: Vague product/audience information

- **Solution**: Provide positioning, competitors, proof points

- ### Issue: Output conflicts with brand voice

- **Cause**: No tone/voice constraints

- **Solution**: Provide brand do/don't list and sample copy

- ### Issue: Can't measure impact

- **Cause**: Tracking events not defined

- **Solution**: Define events and success thresholds in advance

- ---

## Output format

- ```markdown

- ## Marketing Asset Report

- ### Brief Summary

- - **Sub-Skill Used**: [sub-skill]

- - **Objective**: [KPI]

- - **Audience**: [segment]

- ### Deliverable

- [generated asset]

- ### Implementation Checklist

- - [ ] Asset ready

- - [ ] Tracking configured

- - [ ] Success criteria defined

- ### A/B Test Plan

- |------|------------|--------|---------------|

- ```

- ---

- ## Multi-Agent Workflow

- ### Validation & Retrospectives

+ Always return a **short operator-style marketing routing brief**.

- - **Round 1 (Orchestrator)**: Coverage of 23 sub-skills across 5 categories

- - **Round 2 (Analyst)**: KPI alignment, prompt structure review

- - **Round 3 (Executor)**: Output format, actionability check

+ Required qualities:

+ - one KPI, not a vague wish list

+ - one primary lane, not a channel soup

+ - concrete next artifact and handoff

+ - explicit assumptions when context is missing

+ - lightweight measurement plan tied to the chosen lane

- ### Agent Roles

+ ## Examples

- | Agent | Role |

- |-------|------|

- | Claude | Brief composition, asset generation |

- | Gemini | Competitor research, trend analysis |

- | Codex | Tracking code generation, automation |

+ ### Example 1: broad launch ask

+ **Input**

+ > We are launching a new SaaS workflow next month and need help with messaging, landing page copy, onboarding emails, and how to measure whether it worked.

- ---

+ **Output sketch**

+ - Primary lane: `Strategy & growth`

+ - Packet: `launch/growth brief`

+ - Secondary handoff mentions copy/messaging and lifecycle execution

+ - Measurement checklist names launch KPI, activation milestone, and follow-up experiment

- ## Metadata

+ ### Example 2: landing page conversion ask

+ **Input**

+ > Our homepage gets traffic but signups are flat. Help with conversion.

- ### Version

- - **Current Version**: 1.0.0

- - **Last Updated**: 2026-01-21

- - **Compatible Platforms**: Claude, ChatGPT, Gemini, Codex

+ **Output sketch**

+ - Primary lane: `CRO`

+ - Packet: `measurement + experiment packet`

+ - Priority decisions cover page promise, CTA hierarchy, friction points, and test order

- ### Related Skills

- - [presentation-builder](../../documentation/presentation-builder/SKILL.md)

- - [frontend-design](../../frontend/design-system/SKILL.md)

- - [image-generation](../../creative-media/image-generation/SKILL.md)

+ ### Example 3: legacy broad catalog ask

+ **Input**

+ > Use the marketing skills collection to figure out the best lane for this pricing-page refresh.

- ### Tags

- `#marketing` `#cro` `#copywriting` `#seo` `#analytics` `#growth` `#automation`

+ **Output sketch**

+ - Resolve to the canonical `marketing-automation` flow

+ - Choose `CRO` or `Copy & messaging` as the primary lane based on the packet

+ - Leave a short note that the legacy name maps to the same routing logic

- ---

+ ## Best practices

+ 1. **Act like a router, not a slogan generator** — clarify the lane before producing assets.

+ 2. **Choose one KPI and one primary lane** — that constraint makes the output reusable.

+ 3. **Prefer implementation packets over generic advice** — the next team should know what to do.

+ 4. **Keep measurement attached to the packet** — a marketing plan without instrumentation is unfinished.

+ 5. **Defer niche domains to narrower skills** when a dedicated workflow is already stronger.

## References

- - [CXL - Conversion Optimization](https://cxl.com/)

- - [Google Search Central SEO Starter Guide](https://developers.google.com/search/docs/fundamentals/seo-starter-guide)

- - [Google Analytics Documentation](https://support.google.com/analytics)

+ - [references/routing-heuristics.md](references/routing-heuristics.md)

+ - [references/measurement-handoff.md](references/measurement-handoff.md)

+ - `../marketing-skills-collection/SKILL.md`

+ - `../steam-store-launch-ops/SKILL.md`