agentleFS
agiea-brainagiea-brain/agency-skills/skills/kayla-bonnette/ad-creative-content-strategy/SKILL.md

ad-creative-content-strategy

Skill1 starsChanged 16 days agoMITView on GitHub

Use this skill when onboarding a new ecommerce client and the ad production team needs an evidence-based content strategy before anything gets shot. Produces internal ad library and competitor analyses, review-based audience personas, tone of voice and key messages, and a client-ready HTML strategy report with content pillars, hooks, funnel mapping and a testing framework. Also use for "content strategy for a new client", "analyse their ad library", "competitor ad analysis", "build personas from reviews", "creative strategy for Meta".

The file

We split this skill into the rules it gives an agent

8 rules. About 0 are common, 8 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈8

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Run all four steps in one conversation, since the strategy needs every prior output in context.

    line 29
  2. Distinctive:

    MUST collect all assets before Step 1.

    line 41
  3. Distinctive:

    MUST keep the Step 1 and Step 2 docs internal.

    line 42
  4. Distinctive:

    MUST build personas from the brand's own reviews, and note how many reviews were used.

    line 43
  5. Distinctive:

    MUST note clearly wherever findings are inferred from organic content rather than paid ads.

    line 44
  6. Distinctive:

    MUST source language guidelines from the brand immersion deck.

    line 45
  7. Distinctive:

    MUST make the HTML report match the brand's look and feel.

    line 46
  8. Distinctive:

    NEVER reuse a previous client's conversation.

    line 47

≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Applies at the start of a new ecommerce engagement, before ad production begins. Produces two internal analyses for review and one client deliverable: an HTML content strategy report.

The play

  1. Set up a Claude Project with the instructions in references/project-setup.md. Start a new chat for every new client. Never reuse a previous client's conversation.
  2. Collect the assets before anything runs: client name and industry, channels in scope, monthly ad budget, production capacity (in-house, UGC or studio), the product review pages, the brand immersion deck, the website, the brand's Meta Ad Library, social profiles, and three competitors or aspirational brands with their ad library links. The full checklist is in references/project-setup.md. Claude flags anything missing before Step 1.
  3. Step 1, client ad library analysis. Formats, partnership versus brand ads, repeated messaging, apparent personas, the longest-running ads, creative velocity, and a gap analysis comparing what the brand says it stands for against what is actually in its ads. Internal doc. Prompt and the no-active-ads alternate are in references/research-prompts.md.
  4. Step 2, competitor analysis. The same framework across three competitor or aspirational ad libraries, ending with what the client can use strategically. Internal doc. Prompt in references/research-prompts.md.
  5. Step 3, audience persona research. Scrape the brand's own product reviews, extract and rank personas with verbatim hook quotes, compare the real audience against the one the brand thinks it has, then define tone of voice and key messages. Four parts, in references/persona-and-voice-prompts.md.
  6. Step 4, content strategy. Synthesise everything into a styled HTML report: content pillars mapped to personas, three hooks per pillar across varied hook types, funnel mapping across four stages, and a testing framework sized to the budget and production capacity. Prompt in references/strategy-prompt.md.

Run all four steps in one conversation, since the strategy needs every prior output in context. If someone picks up mid-session, they read back through the conversation before running the next prompt.

What good looks like

  • First thing I check that most people skip: the gap between what the brand says it stands for and what is actually running in its ads. The immersion deck says one thing; the ad library usually says another.
  • The second is who the reviews say the audience is, compared with who the brand thinks it is. The personas come from real customer reviews, not from the brief, and the verbatim quotes become the hooks.
  • The mediocre version defaults to solution-based hooks only. Every pillar gets three hooks across varied types: problem-first, pattern interrupt, contrarian, before and after, comment format, fear-based, future-self.
  • I know the strategy is right when every content pillar and key message maps back to a named persona, and the testing framework fits what the client can actually produce and spend.
  • Personas describe an actual audience type, like a skin-conscious millennial or a clean beauty buyer, never a funnel stage or a content format.

Rules

  • MUST collect all assets before Step 1.
  • MUST keep the Step 1 and Step 2 docs internal. They are QA documents, not client deliverables.
  • MUST build personas from the brand's own reviews, and note how many reviews were used.
  • MUST note clearly wherever findings are inferred from organic content rather than paid ads.
  • MUST source language guidelines from the brand immersion deck.
  • MUST make the HTML report match the brand's look and feel. It is the only client deliverable.
  • NEVER reuse a previous client's conversation.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.