marketing-campaign

End-to-end marketing campaign planning and execution — audience research, positioning, campaign angle, landing page copy, email sequences, social posts, ad copy, video scripts, and content calendars.

SKILL.md

GitHub

Marketing Campaign

Plan and execute launch campaigns that convert — not just campaigns that ship.

When to use

  • planning a product, collection, or feature launch
  • building a full content suite from a single product brief
  • defining positioning and campaign angle before any copy is written
  • orchestrating multiple content types across channels
  • reviewing an existing copy set for conversion quality and brand consistency

When NOT to use:

  • the ask is one post, one thread, or one article — use content-engine / article-writing
  • the ask is publishing already-written copy — use crosspost
  • positioning is not settled and the user will not settle it. Copy written before the angle is approved gets thrown away; that is the most expensive failure in this skill

Inputs

Required before Phase 3 — absent input = stop, never assume:

  1. Product truth. For SkyyRose: SKU, name, price, and availability come from wordpress-theme/skyyrose-flagship/data/skyyrose-catalog.csv; imagery resolves only via data/sot-images.json / skyyrose.core.sot_images (filenames are not identity); collection language only from docs/brand/collection-stories.md. Never a product fact from memory.
  2. Audience research — from market-research: jobs-to-be-done, fears, language, alternatives.
  3. Approved positioning and angle (Phase 2 output). No approval → no copy.
  4. Voice profile — from brand-voice, so every channel sounds like one author.
  5. A real deadline if urgency will be used. Fake urgency is banned; without a real date the urgency beat is cut, not invented.

Procedure

Phase 1 — Research

Use market-research to profile the audience (jobs-to-be-done, fears, language, alternatives), map 3+ direct or adjacent competitors (positioning, gaps, messaging weaknesses), and identify 1–3 insights the angle will exploit. Deliverable: a research brief — audience profile + competitive summary + key insights.

Phase 2 — Positioning

Produce: core benefit statement (one sentence, no feature list, no jargon) · positioning formula [Product] helps [audience] [achieve outcome] by [mechanism] · campaign angle (the specific tension, insight, or moment the campaign lives in) · tone profile (delegate to brand-voice). Do not write any copy until positioning and angle are approved.

Phase 3 — Content production

In this order — each layer informs the next:

  1. Landing page copy — hero, problem, solution, features, how it works, proof, CTA
  2. Email sequence — one purpose per email; arc: problem → education → agitation → solution → proof → urgency → final CTA
  3. Social posts — platform-native via content-engine (LinkedIn and X are different formats, not the same copy resized)
  4. Short-form video scripts — timestamp-blocked, written for screen and ear
  5. Ad copy variants — 3–4 testing different angles or audience segments
  6. Content calendar — day-by-day: channel, type, timing, dependencies

Write each deliverable to its own file so the Verification checks have real targets.

Phase 4 — Review, then hand off

Gate every deliverable: 5-second test on all hero / above-fold copy · CTA audit (one per piece, specific, earned) · tone consistency across channels · claim audit (every claim specific and supportable) · cross-channel consistency (ad claims match landing page; email body matches subject). Anything paid, published, or written to production goes through the owning skill's STOP-AND-SHOW gate — this skill produces copy, it does not spend money or publish.

Output contract

  1. Positioning brief — angle, core benefit statement, tone profile
  2. Landing page copy — all sections
  3. Email sequence — subject + preview + body + CTA per email, labelled by day and purpose
  4. LinkedIn posts — 3+ platform-native, distinct angles
  5. X posts — 5+ standalone + 1 thread
  6. Short-form video scripts — 2+ timestamp-blocked with visual direction
  7. Ad copy variants — short headline / long headline / body per variant
  8. Content calendar — day-by-day with channel, type, timing, dependencies
  9. Copy review summary — flagged issues and open questions before anything goes live

Verification

Run across the whole deliverable set, not one file. A check that errors on a missing file exits 2 — a dead gate, not a pass (bug-230).

  1. No verbatim reuse across channels — the consistency rule made falsifiable:
sort landing-hero.md email-01.md x.txt linkedin.txt | uniq -d | grep -v '^[[:space:]]*$'

PASS: no output (final grep exits 1). Any returned line is copy-pasted between two channels; rewrite one. [repro]

  1. Hard-banned language is absent everywhere:
grep -rnEi 'game.chang|revolutionar|world.class|cutting.edge|competitive landscape|learn more|click here|thousands (of )?(people |customers )?trust' ./campaign/

PASS: exit 1, zero matches across the whole campaign directory. Every hit is deleted and rewritten, not softened. [repro]

  1. Every claimed number is sourced — stops invented traction and invented pricing:
grep -rnE '[0-9]+(\.[0-9]+)?%|\$[0-9]' ./campaign/ | grep -v 'Source:'

PASS: exit 1, no lines returned. (Source: catalog CSV) or (Source: operator-supplied; unverified) are acceptable labels; silence is not. [repro]

  1. Price/SKU claims match the catalog — the SOT, not memory:
grep -n '<sku>' wordpress-theme/skyyrose-flagship/data/skyyrose-catalog.csv

PASS: the row exists and its name/price match every mention in the campaign copy. No row → the product claim is deleted, not approximated. [repo]

Prove check 1 can fail before trusting it (rule 3): copy one hero line verbatim into an email, confirm uniq -d returns it, restore. Observed in the worked example.

A SKIP is not a PASS. The 5-second test and tone-consistency read are human judgments — no command grades them. State explicitly that they are open and name the reviewer who closes them, rather than letting the mechanical greens imply the whole gate passed.

Worked example

Real run, 2026-07-28, a Black Rose one-run drop campaign. Two deliverables written to the session scratchpad: landing-hero.md ("Armor for the people the concrete raised. / Black Rose. One run. No restock.") and email-01.md ("You stood up first. The coat is the receipt. / Black Rose goes live tonight at 7pm PT."). The hero line is derived from Black Rose canon in docs/brand/collection-stories.md [repo].

$ sort landing-hero.md email-01.md | uniq -d | grep -v '^[[:space:]]*$'; echo "exit=$?"
exit=1

Zero verbatim overlap [repro] — same angle, different sentences, which is the contract.

The number-sourcing gate was proven red first in the same run, on a sibling research file:

$ grep -nE '[0-9]+(\.[0-9]+)?%|\$[0-9]' research-brief.md | grep -v 'Source:'
3:- Average luxury-streetwear AOV is $240

After labelling that line (Source: operator-supplied; unverified), the same command exits 1 [repro]. Open items from this run: the 5-second test and tone read on the hero are unclosed and owned by the requester — mechanical greens do not close them.

Failure modes

Failure What it looks like Rule
Copy before approved positioning drafts written, angle changes, everything rewritten Phase 2 gates Phase 3. Do not start copy until the angle is approved
Invented product facts price, drop date, or SKU stated from memory bug-096 shape. Catalog CSV is the only manifest; data/sot-images.json the only imagery source — filenames are not identity
Mixed collection canon Love Hurts "bloodline" language on a Black Rose asset Each collection has its own locked canon; Black Rose = armor / "the concrete answering back"
Verbatim reuse across channels the ad headline is also the email subject and the hero Check 1. Same voice, different sentences
Fake urgency a countdown with no real deadline Banned. No real date → cut the urgency beat
Hollow proof "thousands trust us" Check 2 catches it. Specific and supportable or gone
Generic CTA "learn more", "click here" Check 2 catches it. Every CTA is specific and earned
Interchangeable copy the campaign would work for a competitor unchanged Rewrite. Specificity beats adjectives on every channel
Mechanical greens read as full pass "campaign verified" with no human tone read A SKIP is not a PASS — name the open judgments and their owner

Related skills

brand-voice (run before content production) · content-engine (platform-native production) · crosspost (distribution) · market-research (audience + competitive intelligence) · seo (landing page on-page optimisation)