What does this skill do, and when should you use it?
This is a sub-skill of the ai-marketing-claude suite, invoked with /market ads <url>. It fetches the target site, extracts the product, pricing, audience, unique selling proposition, and likely objections, then generates full campaign structures that conform to each platform's specs. Output covers a three-stage retargeting funnel, budget allocation by business type, ROAS/CPA benchmarks, and landing-page message-match checks, all written to AD-CAMPAIGNS.md. It is aimed at agency builders and solo marketers who need client-ready ad plans produced fast.
1) Fetches and analyzes the target URL to extract product, price, audience, USP, conversion action, social proof, and objections; 2) Maps business goals to campaign objectives (awareness, lead gen, trials, e-commerce, app installs); 3) Generates platform-specific ads: Google Responsive Search Ads (15x 30-char headlines, 4x 90-char descriptions, keyword groups with match types and negatives), Meta image/video/carousel variants with creative specs, LinkedIn Sponsored Content and InMail, TikTok vertical video scripts, and Twitter/X promoted tweets; 4) Produces a three-stage retargeting sequence (awareness 40%, consideration 35%, conversion 25% of spend) with sample ads per stage; 5) Recommends budget splits by business type (SaaS, e-commerce, local, agency, creator), industry ROAS and CPA benchmarks, plus a 1-10 landing-page message-match score; 6) Outputs an A/B testing priority framework with kill/scale rules; 7) Writes everything to AD-CAMPAIGNS.md with a terminal summary; 8) Reuses artifacts from sibling skills (COPY-SUGGESTIONS.md, COMPETITOR-REPORT.md, FUNNEL-ANALYSIS.md) when present. The skill is prompt-driven — it runs no scripts and calls no paid APIs.
- An agency builder preparing a pitch generates a full multi-platform ad plan for the prospect's site as a client-ready deliverable
- A SaaS founder starting paid acquisition needs first-draft Google Search headlines, descriptions, and keyword groups
- An e-commerce operator launching a new product wants Meta and TikTok copy plus 9:16 video scripts at platform specs
- A freelancer needs to justify a proposed budget split across platforms and funnel stages with industry ROAS benchmarks
- A media buyer diagnosing weak conversions uses the landing-page alignment checklist and message-match scoring
How do you install this skill?
- Pure prompt-instruction skill with no tests or execution evidence; actual output quality is independently unverified.
- CPA/ROAS benchmarks are unsourced and possibly outdated; verify against current platform data before spend decisions.
- Targets Google/Meta/LinkedIn/TikTok; misaligned with mainland-China ad ecosystems and no Chinese-language support.
- No declared data-handling or confirmation mechanism when fetching target sites; review URL sources before use.
- Publisher unverified, no versioning or changelog; maintenance responsibility and update path unclear.
- Network access
- Local filesystem
Install the full suite (which includes this skill): curl -fsSL https://raw.githubusercontent.com/zubair-trabzada/ai-marketing-claude/main/install.sh | bash, or manually git clone https://github.com/zubair-trabzada/ai-marketing-claude.git and run ./install.sh. Optionally pip install reportlab for PDF report support (not needed by this skill). Uninstall with ./uninstall.sh. MIT licensed.
tmp="$(mktemp -d)"
git clone --depth 1 https://github.com/zubair-trabzada/ai-marketing-claude.git "$tmp"
mkdir -p ~/.claude/skills
cp -R "$tmp/skills/market-ads" ~/.claude/skills/
rm -rf "$tmp"Generated from the source repository and skill path; it copies only this skill's folder. If the author's install steps above differ, follow those first. To scope it to one project, replace ~/.claude/skills with that project's .claude/skills.
How do you use this skill?
In Claude Code, run /market ads <url> (e.g., /market ads https://example.com). The skill fetches the site, analyzes the business and audience, and generates the full campaign plan. Results are saved to AD-CAMPAIGNS.md in the working directory; the terminal prints a summary of platform structure, variation counts, recommended monthly budget, expected CPA, and target ROAS.
What are this skill's strengths and limitations?
- Covers five major ad platforms while respecting real character limits and asset specs (Google 30-char headlines, Meta 1080x1080, TikTok 9:16)
- Goes beyond copy: includes targeting advice, a retargeting funnel, budget allocation by business type, and ROAS/CPA industry tables
- Built-in landing-page alignment checklist and A/B testing priority framework make output nearly media-buyer-ready
- Integrates with sibling skills' outputs (competitor reports, funnel analysis) for a coherent end-to-end plan
- Budget splits and ROAS/CPA benchmarks are heuristic reference values with no cited sources — not validated industry data
- It only produces plans; it cannot create campaigns in ad platforms or access real performance data to calibrate advice
- Retargeting stages assume pixel/audience data (site visitors, cart abandoners) exists; the skill cannot verify this
- No test cases or sample outputs are provided; quality depends on the underlying model and how well the site can be fetched
- Relies on Claude Code slash-command mechanics; triggering needs rework in a plain API or other client
How does this skill compare with similar options?
Side by side with related skills; every score comes from the same FSRS standard.
| Skill | FS score | Stars | Last updated | License |
|---|---|---|---|---|
| Ad Creative & Copy Generation Skill this page | 49 · Use with care | ★ 2.8k | 7mo ago | MIT |
| Ad Creative Builder | 60 · Recommended | ★ 3k | 11d ago | Apache-2.0 |
| Paid Ads Conversion Signal QA | 56 · Use with care | ★ 3k | 11d ago | Apache-2.0 |
| Paid Ads Campaign Architect | 57 · Use with care | ★ 3k | 11d ago | Apache-2.0 |
| Email Sequence Generation (Market Emails) | 49 · Use with care | ★ 2.8k | 7mo ago | MIT |
Compared to drafting ads directly in a general assistant like ChatGPT, this skill's structured edge is built-in platform specs, budget models, retargeting sequences, and benchmark tables — output closer to an agency deliverable. It does not connect to ad platform APIs and cannot replace real campaign management tools.
How did FollowSkills review this skill?
Pure prompt-instruction skill: no script execution, no credential exfiltration, no destructive operations; data flow largely transparent (fetch target URL, write AD-CAMPAIGNS.md). Deducted for: no user-confirmation gates, undeclared fetch scoping, and cross-skill file reads (COPY-SUGGESTIONS.md etc.) without privacy disclosure.
Instructions are self-consistent: trigger, platform specs, output format, and terminal template align; character-limit tables are internally coherent. Deducted for: zero error handling — no failure-feedback guidance for unreachable URLs, non-English sites, or missing pricing pages; static review cannot exceed 10.
Trigger condition is precise (/market ads <url>), platform coverage and output format are clear. Deducted for: no declared non-fit boundaries (complex B2B sales, non-English markets), total misalignment with mainland-China ad ecosystems (WeChat, Ocean Engine) and no Chinese-language support.
Well-layered docs (phases, tables, templates), MIT license. Deducted for: no version number, no changelog, no known-limitations disclosure, install notes live in README not the skill, no maintenance-ownership statement.
AD-CAMPAIGNS.md output template is complete and deliverable to a media buyer; copy-angle library and budget benchmarks have practical value. Deducted for: CPA/ROAS benchmarks unsourced and possibly stale, generated quality unverified; static review cannot exceed 7.
Entirely author-written instructions and marketing claims; README examples are illustrative, not execution evidence; no tests, no third-party corroboration, unsourced benchmark data. Static review cannot exceed 5.
Open a dimension to read why it scored that way
Evidence confidence：Low — Mostly static review, author material or a limited demo; useful for discovery, not high-risk decisions.
See the full review method →