本文へ移動
cccskills
無料GitHub で公開

campaign-brief

Plan a marketing campaign end-to-end using Allan Dib's Before/During/After framework (The 1-Page Marketing Plan). Produces a phase-specific campaign brief with goal, audience, message, channels, assets, timeline, budget, and success metrics. Use when the user asks to plan a campaign, brief a campaign, "we need a campaign for X", launch plan, GTM plan, marketing plan for <product>, demand gen plan, or any multi-channel marketing initiative. Reads ICP, services, and KPIs from knowledge/. For a full quarterly strategy rather than one campaign, see marketing-plan scope in positioning-doc and kpi-review. For the content schedule inside the campaign, see social-calendar. For the post-campaign review, see retro.

インストール方法を見る

含まれるファイル(1)

  • SKILL.md12.7 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

campaign-brief

Produces a campaign brief that an empowered team can execute without needing a follow-up meeting. Every brief is anchored to one phase of the customer journey (Before/During/After) using Allan Dib's 1-Page Marketing Plan. Phase determines channels, assets, and success metrics - not the other way around.

When to use

  • "Plan a campaign for our Q3 launch"
  • "Build a GTM plan for <product>"
  • "Brief a demand gen campaign targeting CMOs"
  • "Write the launch plan for the integration with X"
  • "We need a marketing plan for <event/feature/audience>"

Framework: Before/During/After (Allan Dib, The 1-Page Marketing Plan)

Every campaign primarily serves one phase. Force the choice before building anything else.

BEFORE - Prospects who don't know you exist.

  • Goal: get strangers to raise their hand
  • Focus: reach, awareness, USP clarity
  • Channels: paid social, SEO, PR, cold outreach, events, partner content
  • Success metrics: reach, impressions, hand-raisers (leads generated), CPL

DURING - Leads who know you but haven't bought.

  • Goal: convert interest into purchase
  • Focus: lead capture, nurture, trust-building, sales conversion
  • Channels: email sequences, retargeting, demos, webinars, case studies
  • Success metrics: MQLs, SQLs, demo bookings, trial starts, conversion rate

AFTER - Customers who have already bought.

  • Goal: create raving fans who expand and refer
  • Focus: world-class experience, upsell, cross-sell, referrals
  • Channels: customer comms, CSM touchpoints, loyalty programs, referral incentives
  • Success metrics: NPS, expansion revenue, referrals generated, churn rate, LTV

A campaign cannot serve all 3 phases at once. If the user wants all 3, split it into 3 campaigns.

Inputs needed

Walk the user through these one at a time. Do not skip.

  1. Campaign phase: Before, During, or After? If the user isn't sure, ask: "Are you trying to reach new people, convert known leads, or grow existing customers?" Force a single answer.
  2. Campaign goal: one sentence, tied to the phase.
  3. Primary KPI: the one number that defines success. With a target.
  4. Audience: which persona from knowledge/icp/personas.md. If multiple, force-rank.
  5. Window: start date, end date (DD-MM-YYYY).
  6. Budget: total, or "no paid budget".
  7. Constraints: legal, brand, channel restrictions, dependencies on other teams.
  8. What's already decided: any creative, channel, or partner choices the user has locked in.

Process

  1. Load context. Read all reads: files. If knowledge/icp/personas.md is missing, stop and run /brand-context.

  2. Identify the phase. Before generating any brief content, confirm the Before/During/After phase with the user. State it explicitly: "This is a BEFORE campaign. The goal is to generate hand-raisers from people who don't yet know us."

  3. Check knowledge/learnings.md for any relevant past campaign retros. Reference them in the brief.

  4. Generate the brief in this format:

    # <Campaign name>
    
    **Phase**: BEFORE | DURING | AFTER (Allan Dib framework)
    **Phase goal**: <get hand-raisers | convert leads | grow customers>
    **Campaign goal**: <one sentence, single-objective>
    **Primary KPI**: <metric> -> target <number> by <DD-MM-YYYY>
    **Window**: DD-MM-YYYY to DD-MM-YYYY
    **Budget**: $<amount> or <none>
    **Owner**: <person, ask the user>
    **Status**: draft
    
    ## Audience
    - **Primary persona**: <from knowledge/icp/>
    - **Top 3 pains we are speaking to**:
      1. ...
      2. ...
      3. ...
    - **Where they are**: channels, communities, publications
    
    ## Message
    - **Core message** (one sentence the audience should remember):
    - **Proof points** (3 max). Each must be lifted from `knowledge/company.md` verified proof points
    

with the source named inline. If fewer than three verified proof points exist, ship fewer: write [NO VERIFIED PROOF POINT] and move it to Open questions. Never generate a proof point - these flow straight into /linkedin-post and /landing-page-writer and become public claims: 1. ... 2. ... 3. ...

  • What we are NOT saying: explicit anti-messages to avoid scope creep

Strategy

  • Why now (timing rationale):
  • Differentiation vs <competitor> (how this plays where we win):
  • What we learned from <past campaign> (from knowledge/learnings.md):

Channel mix

Channels must match the campaign phase. BEFORE = reach channels. DURING = conversion channels. AFTER = retention channels.

ChannelRoleAsset countSpendOwner
LinkedIn organicAwareness8 posts$0<name>
LinkedIn adsDemand3 ad sets$X<name>
Email nurtureConversion5-email sequence$0<name>
...

Assets needed

Group by team. Be specific.

  • Content: 1 pillar article (1800w), 8 LinkedIn posts, 3 quote graphics
  • Design: 1 landing page hero image, 3 ad creatives, 1 email header
  • Video: 1 explainer (60s)
  • Sales enablement: 1 one-pager, 1 deck update
  • Web: 1 landing page

Timeline

Working backward from launch date.

WeekMilestoneOwner
-4Brief approved, assets briefed
-3First drafts in review
-2Final assets, landing page live
-1Soft launch, internal alignment
0Launch
+1First metrics review
+4Campaign retro

Success metrics

Phase-specific. Do not mix metrics from other phases.

BEFORE campaigns:

  • Primary: <reach | CPL | hand-raisers generated> = <target>
  • Secondary: impressions, CTR, email sign-ups
  • Leading indicators (visible by week 1): ad CTR, organic post reach
  • Lagging indicators (visible by week 4): total leads captured, CPL

DURING campaigns:

  • Primary: <MQLs | SQLs | demo bookings | trials> = <target>
  • Secondary: email open rate, demo-to-close rate
  • Leading indicators (visible by week 1): email opens, landing page CVR
  • Lagging indicators (visible by week 4): total SQLs, pipeline generated

AFTER campaigns:

  • Primary: <NPS | referrals | expansion revenue | churn rate> = <target>
  • Secondary: product usage, upsell conversion, CSAT
  • Leading indicators (visible by week 1): feature adoption, support tickets
  • Lagging indicators (visible by week 4): NPS score, referrals, expansion ARR

Risks and mitigations

  • Risk: <what could go wrong> Mitigation: <what we do about it>

Open questions

  • List anything you do not have an answer to. Do not invent.

Approvals

  • Marketing lead
  • Product
  • Sales
  • Legal (if applicable)

5. **Self-check**:
- Phase is declared and all elements serve that phase
- Channels match the phase (no conversion channels in a BEFORE campaign, no awareness channels in an AFTER campaign)
- Single primary KPI - not three
- Specific persona, not "marketers"
- Working-backward timeline, not aspirational
- Concrete assets list with counts
- References at least one prior learning if `knowledge/learnings.md` exists
- Phase-specific success metrics only

6. **Save** to `output/campaign-brief/<DD-MM-YYYY>-<slug>.md` with frontmatter:
```yaml
---
format: campaign-brief
campaign: <name>
phase: <before|during|after>
goal: <awareness|demand|expansion|retention|recruitment>
start: DD-MM-YYYY
end: DD-MM-YYYY
primary-kpi: <metric>
target: <number>
budget: <amount or none>
created: DD-MM-YYYY
---
  1. Offer next actions:
    • Generate the asset list as individual skill calls (/linkedin-post x 8, /landing-page-writer x 1, etc.)
    • Build the launch deck (/ppt-maker)
    • Remind the user: "Set a calendar reminder to run /retro on [launch date + 28 days] to capture learnings."

Rules

  • Declare the phase before writing a single line of the brief. Phase determines everything else.
  • One primary KPI. If the user names three, force them to rank and pick one.
  • Channel mix must match the phase. A BEFORE campaign with a demo CTA is a contradiction. Call it out.
  • Never copy a previous campaign brief without reading knowledge/learnings.md first. Past failures should change current plans.
  • If budget is zero, do not propose paid channels. Reallocate to organic.
  • Timeline must be working backward from launch. If the user gives a launch date, derive milestones from it. If not, ask.
  • If the user wants to run a BEFORE and DURING campaign simultaneously, produce two separate briefs. Mixed-phase campaigns are unfocused campaigns.

Handling [UNVERIFIED] and [NEEDS INPUT] tags

brand-context tags any number the user was unsure of as [UNVERIFIED], and several skills tag gaps as [NEEDS INPUT]. Those tags are a contract, and it only works if this skill honours it.

  1. Carry the tag forward. If a baseline, benchmark or proof point arrives tagged, every figure derived from it is tagged too. A target built on an unverified baseline is an unverified target.
  2. Never silently promote. Do not drop the tag because the number looked confident in the source file.
  3. Say it in the output. List every tagged input in its own line near the top, so a reader knows which numbers are measured and which are estimates before they act on them.
  4. A decision that would change if the tagged number were wrong must say so explicitly.

The Before / During / After model (Allan Dib)

Allan Dib's The 1-Page Marketing Plan splits a campaign into three phases with different jobs, different audiences and different metrics. Most failed campaigns are a During-phase tactic aimed at a Before-phase audience.

PhaseAudienceJobPrimary metricTypical channels
BeforeThey do not know youEarn attention and permissionReach, new subscribersContent, paid awareness, PR, partnerships
DuringThey know you, have not boughtConvert interest into a decisionConversion rate, pipelineEmail, retargeting, demos, sales enablement
AfterThey boughtRetain, expand, get referralsRetention, expansion, referral rateOnboarding, community, case studies

One campaign, one phase. A brief covering all three is a strategy document, not a campaign, and it produces channel plans nobody can execute. Pick the phase, then choose channels that serve it.

Decision thresholds

SignalThresholdWhat to do
Phases in scopeMore than 1Split into separate briefs, sequenced
Primary KPIsMore than 1Pick one. The rest are supporting metrics
Verified proof pointsFewer than 3Ship fewer and move the gap to Open questions. Never generate one
Baseline for the targetMissing or [UNVERIFIED]Tag the target, and state which decisions change if the baseline is wrong
Channels plannedMore than 4 for one phaseCut. Thin coverage across many channels beats nothing, but not by much

Related skills

  • /brand-context owns knowledge/company.md, the source of verified proof points
  • /social-calendar turns the content line of this brief into a dated schedule
  • /kpi-review reads the results against the target set here
  • /retro runs after the campaign and writes reusable learnings back
  • /growth-experiment when a channel in the plan is unproven for this audience

What this skill cannot know

  • Whether the team has capacity to produce the assets this brief plans. Ask before finalising
  • Whether the budget figure is approved or aspirational
  • Whether a channel worked for this audience before, unless it is in knowledge/learnings.md
  • Whether the target derived from an [UNVERIFIED] baseline is reachable

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Rate any product, feature, or experience on the 11-star scale (Brian Chesky's Airbnb thought experiment). Use when user says "rate this experience", "11-star", "star rating", "experience audit", "how good is this", "experience rating", "product audit", "quality assessment", or wants to evaluate product quality and identify improvement paths. Also trigger when user wants to benchmark a feature, assess where a product stands, or map out what "great" looks like - even if they don't explicitly say "11-star".

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Generate A/B copy variants grounded in Eugene Schwartz's awareness and market-sophistication stages and John Caples' tested-headline discipline. Every variant declares its angle, the awareness stage it targets, and a falsifiable hypothesis, and the skill recommends which pair to test first and what delta counts as signal. Use when the user says "A/B test this", "write variants", "give me options", "test different angles", "copy variants", "ab copy", "test this headline", "multiple versions", "which version is better", or wants to test copy before publishing. For reviewing copy that already exists, see copy-review. For sample size and stopping rules, see growth-experiment. For diagnosing a page rather than its words, see page-cro. For full paid ad sets by platform, see ad-campaign-writer.

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Write measurable acceptance criteria for requirements or user stories. Use when the user says "write acceptance criteria", "definition of done", "how do we test this", "fit criteria", "given when then", "Gherkin scenarios", "what does done look like", "testable conditions", "how will we know this works" - even if they don't explicitly say "acceptance criteria".

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Build a strategic account plan for winning, retaining, or expanding a named account. Use when the user says "build an account plan", "strategic account planning", "how do I grow this account", "account expansion strategy", "whitespace analysis", "QBR prep", "key account review", "map the stakeholders in this account", "land and expand plan", "how do I break into [company]", or wants a structured strategy for a specific account.

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Write paid ad copy for LinkedIn, Google, Meta, and YouTube. Produces multiple variants per platform, each targeting a different Eugene Schwartz awareness level and Cialdini persuasion principle. Use when the user asks for ad copy, ad variants, "write LinkedIn ads", Google ads, Meta ads, paid social copy, ad headlines, or wants creative for a paid campaign. Reads brand voice, ICP, and positioning from knowledge/. For organic social posts, see linkedin-post or social-calendar. For the landing page the ad points at, see landing-page-writer. For testing variants, see ab-copy-writer.

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Design clean and consistent APIs. Use when the user says "design an API", "API design review", "design these endpoints", "REST API for X", "GraphQL schema", "API contract", "how should this endpoint work", "review my API design", or wants to create or improve an API interface - even if they don't explicitly say "API design".

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

qa-aman のスキルをすべて見る

このスキルの問題を報告する