How to Use This Guide
This guide gives you a complete, copy-paste-ready Claude Skills stack for growth marketers. Each skill teaches Claude a repeatable workflow it activates automatically — with built-in methodology, reference logic, and structured outputs you can act on immediately.
What a Claude Skill actually is: A skill is an instruction set you install into Claude once. After installation, Claude activates the full methodology every time you trigger the skill — no re-prompting, no re-explaining context. It's the difference between asking Claude a question and giving Claude a job.
How to install any skill:
- Open Claude (claude.ai or Claude desktop app)
- Navigate to Customize and find the Skills section
- Click “+” sign —> Create Skill —> Upload Skill
- Claude confirms installation
- Trigger the skill by name in any conversation
How to read this guide: Each skill entry includes a one-line description, the category it belongs to, a practical example prompt you can copy, and the output you should expect. Skills are grouped by growth function so you can install just the category you need right now, or the full stack.
Recommended starting point: If you're new to Claude Skills, start with the Analytics category. Data Anomaly Detective and Attribution Modeling Advisor produce immediate, visible value on any live campaign — they're the fastest way to understand what a skill actually does for you.
Skills by Category
Paid Ads
These four skills cover the full paid acquisition stack — from audit to copy to search intent analysis. All are framed for founders and senior operators who oversee campaigns, not run them day to day.
1. Ad Performance Diagnostic
What it does: Analyzes campaign data across any paid channel and surfaces concrete optimization recommendations — quality score factors, bid strategy issues, budget distribution problems, and creative fatigue signals.
When to use it: Weekly campaign review, before a budget increase, or when ROAS is declining and you can't pinpoint why.
Example prompt:
Run an ad performance diagnostic on my Google Ads account.
Here's my campaign data for the last 30 days: [paste CSV or metrics].
I want to know where budget is being wasted and what to fix first.Expected output: A prioritized list of issues by revenue impact, with specific recommendations for each — not generic advice. Includes bid adjustment suggestions, audience exclusion recommendations, and creative swap flags.
2. Performance Max Auditor
What it does: Runs a comprehensive PMax campaign audit covering asset group diversity, signal quality, audience inputs, budget allocation, and cannibalization of brand campaigns.
When to use it: Before scaling a PMax campaign, after a sudden performance drop, or when you suspect PMax is eating your branded search traffic.
Example prompt:
Audit my Performance Max campaign.
Asset groups: [describe or paste].
Monthly spend: [amount]. Target ROAS: [number].
Current issue: [brief description — e.g. "brand CPC spiked 40% after PMax launch"].Expected output: Structured audit report with sections for asset diversity score, signal quality assessment, brand cannibalization check, and a ranked action list with expected impact.
3. Google Ads Copy Generator
What it does: Creates compliant Google Ads copy — headlines, descriptions, and extensions — that hits character limits, avoids policy violations, and is optimized for the keyword intent you specify.
When to use it: New campaign launches, ad refreshes after creative fatigue, or testing new angles.
Example prompt:
Generate Google Ads copy for a B2B SaaS product.
Product: [name and one-line description].
Target keyword: [primary keyword].
USP: [your main differentiator].
Tone: [professional / urgent / benefit-led].
I need 5 headline variants and 3 description variants.Expected output: Headlines (30 chars max each), descriptions (90 chars max each), and a set of sitelink and callout extension suggestions — all labelled with character counts and a policy compliance note.
4. Search Term Analyser
What it does: Analyzes your Google Ads search terms report to separate high-intent winners from wasted spend, identify negative keyword opportunities, discover new keyword angles, and map intent to ad group structure.
When to use it: Monthly search term review, or any time CPAs are rising without a clear reason.
Example prompt:
Analyze this search terms report and identify:
1. High-intent queries to add as exact match keywords
2. Wasted spend queries to add as negatives
3. New keyword angles I haven't targeted yet
[paste search terms report or CSV]Expected output: Three clearly separated lists — winners to promote, negatives to add, and new keyword opportunities — each with spend, conversions, and a one-line rationale.
Analytics
These four skills turn raw marketing data into decisions. They work best when you paste actual numbers — the more specific the input, the more precise the output.
5. Attribution Modeling Advisor
What it does: Compares attribution models (first touch, last touch, linear, time decay, data-driven) across your channel mix and recommends budget moves based on which model most accurately reflects your conversion path.
When to use it: Before a quarterly budget review, when a channel looks underperforming on last-click but you suspect it assists heavily, or when leadership is asking which channel "actually works."
Example prompt:
Compare attribution models for my marketing mix.
Channels: [list channels — e.g. paid search, paid social, email, organic].
Average sales cycle: [number of days].
Primary conversion goal: [e.g. demo booked, free trial signup].
I want to understand which channels are undervalued by last-click.Expected output: A model comparison table, a summary of which channels gain or lose credit under each model, and specific budget reallocation recommendations with supporting rationale.
6. Data Anomaly Detective
What it does: Detects unusual patterns in your marketing metrics, diagnoses root causes, prioritizes alerts by business impact, and recommends immediate actions.
When to use it: When a key metric moves unexpectedly — up or down — and you need to explain it fast.
Example prompt:
Analyze these weekly marketing metrics and identify anomalies.
[paste metrics table or describe the numbers]
Flag anything statistically unusual and tell me the most likely cause for each.Expected output: Anomaly log with severity rating (high, medium, low), probable cause for each anomaly, and a recommended diagnostic step to confirm the diagnosis before acting.
7. Cohort Analysis Builder
What it does: Builds retention and LTV analysis from user or customer data, segmented by acquisition cohort. Surfaces which acquisition channels, campaigns, or time periods produce users who actually stick.
When to use it: Monthly performance review, before making channel budget decisions, or when you suspect a high-volume source is bringing low-quality users.
Example prompt:
Build a cohort analysis from this customer data.
[paste or describe: signup date, acquisition source, activity dates, revenue]
I want to see 30/60/90-day retention by cohort and which acquisition channels have the highest LTV.Expected output: Cohort retention table, LTV by acquisition source, and a channel quality ranking that goes beyond volume to value.
8. Funnel Optimizer
What it does: Maps your marketing or product funnel stage by stage, identifies the biggest drop-off points, compares your conversion rates against benchmarks, and generates specific improvement recommendations for each leaking stage.
When to use it: Funnel audits, pre-launch reviews, or when overall conversion rate is declining without a clear single cause.
Example prompt:
Optimize my conversion funnel.
Stages and conversion rates:
- Impression to click: [%]
- Click to landing page lead: [%]
- Lead to demo booked: [%]
- Demo to closed deal: [%]
What's my biggest drop-off and what should I fix first?Expected output: Drop-off analysis ranked by revenue impact, benchmark comparisons per stage, and a prioritized fix list with expected lift estimates.
CRO and Strategy
Four skills covering conversion optimization and growth planning — from page-level CRO to company-level strategy.
9. Landing Page Optimizer
What it does: Combines analytics data with UX and conversion design principles to produce specific recommendations for improving landing page performance — above-fold content, CTA placement, trust signal implementation, load speed, and mobile experience.
When to use it: Before launching a new page, after a page conversion rate drops, or when running a CRO sprint.
Example prompt:
Optimize this landing page for conversions.
URL or page description: [describe the page or paste key sections]
Current conversion rate: [%]
Traffic source: [e.g. paid social, organic search]
Primary CTA: [what you want visitors to do]
Top drop-off point (if known): [e.g. above fold, form step 2]Expected output: Prioritized recommendation list organized by impact — quick wins (implement this week), structural improvements (this sprint), and longer-term tests to run.
10. A/B Test Analyzer
What it does: Calculates statistical significance, effect size, and confidence intervals for your A/B test results, declares a winner or identifies if the test needs more data, and provides a clear recommended action.
When to use it: Whenever you have A/B test results and need to know if they're real or noise.
Example prompt:
Analyze this A/B test result.
Control: [X] conversions from [N] visitors
Variant: [X] conversions from [N] visitors
Confidence threshold I'm working to: 95%
Tell me if the result is significant and what action to take.Expected output: Statistical significance calculation, confidence interval for both variants, effect size, a clear winner declaration (or "needs more data" with sample size requirement), and a one-line recommended action.
11. Growth Experiment Designer
What it does: Generates ICE-scored growth hypotheses (Impact, Confidence, Ease) across your full AARRR funnel — Acquisition, Activation, Retention, Revenue, Referral — and packages each as a structured experiment brief with success criteria.
When to use it: Sprint planning, quarterly growth reviews, or when your team's test backlog is empty.
Example prompt:
Design 10 growth experiments for my business.
Stage of growth: [early / scaling / optimizing]
Primary bottleneck right now: [e.g. activation, retention, referral]
Product: [one-line description]
I want experiments ICE-scored and sorted by priority.Expected output: 10 experiment briefs, each with hypothesis, ICE score breakdown, success metric, minimum detectable effect, and estimated runtime — formatted for direct import into a growth backlog.
12. Campaign Strategy Builder
What it does: Develops a complete campaign strategy from scratch — positioning statement, audience segmentation, channel mix, messaging framework, and a content calendar outline — based on your product, goal, and market context.
When to use it: New product launches, entering a new market, or when a campaign is underperforming and you need a strategic reset.
Example prompt:
Build a campaign strategy for my product launch.
Product: [name and one-line description]
Target audience: [who they are and their core pain]
Primary goal: [e.g. 500 free trial signups in 60 days]
Budget range: [approximate]
Channels available: [list what you have access to]Expected output: Full campaign brief including positioning statement, audience persona summary, channel strategy with budget split rationale, 3 headline messaging angles, and a 4-week content calendar outline.
Research
Two research skills that replace hours of manual synthesis work with structured outputs you can act on immediately.
13. Customer Research Personas
What it does: Synthesizes customer reviews, competitor websites, sales call notes, G2/Capterra reviews, and market research into detailed buyer personas — including pain points, objections, trigger moments, and messaging that resonates.
When to use it: Before writing any copy, before a campaign strategy session, or when you're entering a new segment and need to understand who you're talking to.
Example prompt:
Build buyer personas from this customer research.
[paste: customer reviews, competitor site copy, sales call notes, or G2 reviews]
I want 2-3 personas with: job title, primary pain, trigger moment, key objection, and the message that would convert them.Expected output: 2-3 fully structured personas, each with demographic profile, pain hierarchy, trigger event that makes them search for a solution, top objection and how to counter it, and 2-3 headline angles tailored to each persona.
14. Content Engagement Analyzer
What it does: Analyzes content performance data across LinkedIn, email, blog, or social to identify which formats, topics, lengths, and posting patterns drive the highest engagement — and produces a recommendation set for your next 30 days.
When to use it: Monthly content reviews, before planning a new content calendar, or when engagement has dropped and you don't know why.
Example prompt:
Analyze my content performance and identify what's working.
[paste: post data with impressions, engagements, dates, formats, topics]
I want to know: what format works best, which topics drive highest engagement, and what I should post more of next month.Expected output: Performance breakdown by format and topic, engagement rate comparison, timing analysis, and a prioritized content recommendation list for the next 30 days.
Installation
Option A: Install via Claude Skills Manager (Recommended)
- Open Claude at claude.ai or in your Claude desktop app
- Navigate to Customize and find the Skills section
- Click “+” sign —> Create Skill —> Upload Skill
- Claude confirms: "Skill installed. I'll activate [Skill Name] automatically when relevant."
- Trigger any skill by name in a new conversation
Option B: Stack Them Together
Many of these skills compound when used in sequence. Here are three proven stacks:
The Paid Ads Audit Stack
- Search Term Analyser (identify waste and winners)
- Ad Performance Diagnostic (cross-reference with campaign-level data)
- Performance Max Auditor (check if PMax is inflating or hiding the numbers)
The CRO Sprint Stack
- Funnel Optimizer (find the biggest leak)
- Landing Page Optimizer (fix the page causing the leak)
- A/B Test Analyzer (validate the fix is actually working)
The Strategy Reset Stack
- Customer Research Personas (understand who you're targeting)
- Campaign Strategy Builder (build the strategy around them)
- Growth Experiment Designer (generate the test backlog to execute it)
Download:
|
Want more Claude breakdowns like this? Follow @rananjayraj on LinkedIn for weekly AI tool guides, marketing automation workflows, and practical Claude tutorials.
LinkedIn | The AI Driven Marketer | Website | Github | Youtube