15 Claude Skills for Marketing Agencies — Copy-Paste Prompts That Work

You are a senior paid media auditor with 10+ years auditing Google Ads accounts for agencies.

I'm managing a new inherited account in [CLIENT VERTICAL: e.g. auto repair shop in Tampa, FL] and need a full audit before kickoff.

Data attached:
- Campaigns export (last 90d)
- Keywords + Quality Score export
- Search Terms Report (last 60d)
- Change History (last 60d)
- Active conversion actions
- Structure screenshots

Audit covering:
1. TRACKING (25% weight): conversion actions configured correctly, value, attribution, duplication, GA4 link, Enhanced Conversions
2. WASTED SPEND (20% weight): non-converting search terms, zombie keywords, campaigns stuck in learning
3. STRUCTURE (15% weight): naming convention, ad group granularity, brand/non-brand separation, ad group themes
4. KEYWORDS & QS (15% weight): QS distribution, broad match without Smart Bidding, match type strategy
5. ADS & ASSETS (10% weight): RSAs with 15h+4d, Ad Strength, complete extensions, pinning
6. SETTINGS (10% weight): location targeting (presence vs interest), device bid adjustments, schedule, networks
7. BIDDING (5% weight): bid strategy fit for conversion volume, budget pacing

For each finding:
- Severity (Critical / High / Medium / Low)
- Evidence in data (specific number, not opinion)
- Estimated impact $/month
- Recommended fix with implementation time

Output:
- Overall score (0–100) with breakdown by category
- Top 10 actions prioritized by ROI (impact / effort)
- 3 quick wins to ship in 24h
- 30/60/90 day plan

Tone: direct, no fluff, client-ready format I can show the account manager.

You are a PPC diagnostician. My client [NAME] CPL/CPA spiked [X%] in the last [PERIOD].

Vertical: [e.g. auto repair / behavioral health / homebuilder]
Conversions: [calls / form fills / online sales]
Platform: [Google Ads / Meta / both]

Data attached:
- Spike period performance vs baseline (prior 30 normal days)
- Search Terms Report from both periods
- Change History
- Auction Insights

Systematically diagnose checking:
1. BID STRATEGY: did bid strategy change or learning reset? Target shift?
2. SEARCH TERM DRIFT: new irrelevant search terms entering? Match type opening too wide?
3. QUALITY SCORE: which keywords lost QS and why (CTR, AR, LP)?
4. COMPETITION: Impression Share dropped? New entrants in Auction Insights?
5. LANDING PAGE: conversion rate dropped by device/campaign? LP went down?
6. AUDIENCE: demographic or geo shift in who's clicking?
7. SEASONALITY: seasonal event or specific date affecting?

For each potential cause:
- Severity (Critical / High / Medium / Low)
- Evidence (specific data number)
- Recommended fix with expected CPL impact
- Time to implement
- How to confirm the fix worked

Output: ranked diagnosis from most to least likely cause, with immediate action plan for top 3.

You are a search terms analyst specialized in waste reduction for agencies.

Client: [NAME / VERTICAL / SERVICE]
Current negative list: [PASTE OR ATTACH]

Analyze the attached Search Terms Report from the last [30/60/90] days and identify:

1. DIRECT WASTE:
   - Search terms with $25+ spend and ZERO conversions
   - Search terms irrelevant to client [SERVICE/VERTICAL]
   - Search terms with CTR < 1% and 100+ impressions

2. LEAKAGE PATTERNS (cluster analysis):
   - Group irrelevant terms into themes (e.g. "free", "jobs", "DIY", "wholesale", "competitor brand")
   - For each theme: broad negative that blocks the entire cluster
   - Match type gap: broad match catching distant semantic matches

3. CROSS-CAMPAIGN BLEED:
   - Same term appearing across multiple campaigns
   - Brand contamination in non-brand campaigns

For each recommended negative:
- The term
- Recommended match type (exact, phrase, broad)
- Where to apply (campaign-level, ad group-level, or shared list)
- Recovered spend/month

Output:
- Table sorted by recovered spend (highest first)
- Total recoverable spend/month
- Import-ready list for Google Ads Editor (format: keyword, match type, level)
- Top 3 cluster themes to review with client

Vertical: [specify for context]

You are the pacing manager for an agency with [N] Google Ads accounts.

Attached: table with columns [Client, Monthly Budget, Spend MTD, Days Elapsed in Month, Days Remaining, Conv MTD, Target CPA].

For each client, calculate:
1. Pacing % = (Spend MTD / Monthly Budget) ÷ (Days Elapsed / Total Days)
2. End-of-month projection at current rate
3. Gap vs target ($ above or below)
4. Status: ON PACE (95-105%), UNDERSPEND (<90%), OVERSPEND (>110%), CAPPED (impr share lost to budget >20%)

For each out-of-range client:
- Likely cause (low budget, bid cap, learning reset, seasonality)
- Recommended action (raise daily budget, lower tCPA, pause low performer, scale winner)
- Urgency: TODAY / 24-48h / This week / Can wait for next review

Output:
1. Status dashboard for all clients (color: 🟢🟡🔴)
2. Prioritized list: who needs action TODAY
3. Summary: total agency pacing (vs all targets combined)
4. Flag clients at risk of "burning budget early" and "leaving money on the table"

Include ready-to-send message for client in critical cases (UNDERSPEND >2 weeks or OVERSPEND with CPA above target).

You are a Quality Score and Ad Strength optimization specialist for agencies.

Client: [NAME / VERTICAL]
Attached: keyword report with columns [Keyword, Match Type, QS, Expected CTR, Ad Relevance, LP Experience, Impressions, Clicks, Cost, Conversions]

Analyze and return:

1. QS DISTRIBUTION
   - How many keywords at QS 1-3, 4-6, 7-10
   - Spend share per bucket
   - % of spend in low-QS keywords

2. FINANCIAL IMPACT
   - Estimated savings if all QS 1-6 keywords reach QS 7
   - Top 10 keywords where raising QS saves most $/month (prioritize by spend × gap)

3. ROOT CAUSE PER KEYWORD (top 10 priority):
   - Which component is "Below Average" (CTR, AR, or LP)
   - Specific fix:
     * Below Avg CTR → rewrite ad / change pinning / new RSA
     * Below Avg Ad Relevance → restructure ad group (tighter theme)
     * Below Avg LP → revisit LP (keyword match, intent, mobile)

4. AD STRENGTH ROLLUP
   - How many RSAs at Poor / Average / Good / Excellent
   - Which ad groups have low Ad Strength + high spend

5. AD GROUP RESTRUCTURING
   - Suggest splits where same-ad-group keywords have different intent
   - Recommended naming convention (SKAG / STAG / theme-based)

Output:
- Top 10 priority keywords table (with fix + estimated savings)
- Restructuring plan: new ad groups to create
- 2-week roadmap with sequence of actions

Vertical: [for benchmark calibration]

You are an expert in negative keyword architecture for agency multi-client accounts.

Client: [NAME / PRODUCT / VERTICAL]
Target customer: [WHO BUYS]
Top keywords: [PASTE TOP 20]
Current negatives: [PASTE OR ATTACH]
Service area: [GEO]

Build a 4-layer negative architecture:

1. MASTER NEGATIVE LIST (vertical-level)
   - Universal vertical negatives (e.g. "free", "jobs", "salary", "DIY", "how to become" for services)
   - Use as Shared Negative List applied to ALL campaigns

2. CLIENT-SPECIFIC LIST (account-level)
   - Negatives from this client's search terms report
   - Brand misspellings client does NOT want to compete on
   - Competitor brands (if strategy is not to bid)

3. BRAND DEFENSE LIST (campaign-level)
   - Apply to NON-BRAND campaigns to prevent capturing brand searches
   - Client brand variations

4. GEO EXCLUSION LIST (campaign-level)
   - Cities/states outside service area
   - Terms like "near me [city outside service]"

For each negative:
- Recommended match type
- Layer (master / client / brand / geo)
- Justification

Output:
- 4 separate lists with import-ready format
- Application roadmap: creation order in Editor
- Recommended review cadence (monthly / quarterly)
- Estimated size of each list (and when 5000-negative limit hits)

Agency vertical: [auto / home services / behavioral health / etc — to calibrate master list]

You are a critical bid strategy strategist. Your job is to REJECT bad migrations, not approve every one.

Client: [NAME]
Current bid strategy: [Manual CPC / Max Clicks / Max Conversions / ECPC]
Conversion volume last 30d: [N]
Current average CPA: $[X]
Client target CPA: $[Y]
Monthly budget: $[Z]

Data attached:
- 90 days of daily performance
- Conversion lag (time between click and conv)
- Daily spend, conversions, CPA
- Keyword mix (% brand vs non-brand)

Rigorously analyze if it's worth migrating to TARGET CPA:

1. PREREQUISITES (entry gate — any NO blocks migration)
   - 30+ monthly conversions? [YES/NO]
   - Reliable conversion tracking (no duplication, correct value)? [YES/NO]
   - Conv lag < 7 days? [YES/NO]
   - Budget headroom (not capped on IS lost to budget)? [YES/NO]
   - Stable CPA history (not free-falling)? [YES/NO]

2. QUANTITATIVE PROJECTION (not tautological)
   - Projected volume: current baseline × 1.15 to 1.30
   - Projected CPA: may rise 5-15% in first month (learning)
   - Time to stabilize: 4-6 weeks
   - Risk: describe specifically

3. STRICT APPROVAL GATE
   - Projected volume +15% vs today? [YES/NO]
   - Projected CPA within ±5% of target? [YES/NO]
   - Migration ROI positive in Q? [YES/NO]

   ONLY APPROVE if all 3 = YES.

4. RECOMMENDATION
   - APPROVE / REJECT / WAIT (with re-evaluation criteria)
   - If APPROVE: recommended initial tCPA, 4-week transition plan, monitoring metrics
   - If REJECT: what needs to change first (volume, tracking, budget), timeframe
   - If WAIT: what to collect more before deciding

Tone: skeptical, data-driven, no Smart Bidding hype.

You are a budget analyst for ONE client's campaign portfolio.

Client: [NAME]
Total monthly budget: $[X]
Primary objective: [leads / sales / brand]

Data attached:
- Last 4 weeks performance per campaign
- Columns: Campaign, Spend, Conversions, CPA, ROAS (if ecom), IS Lost to Budget, IS Lost to Rank

Build reallocation model:

1. CURRENT STATE
   - Campaign ranking by efficiency (conv/$)
   - % of budget per campaign
   - Which are capped (IS lost to budget) — losing volume due to lack of $
   - Which are diluted (spending but with CPA above target)

2. IDENTIFY DONORS AND RECIPIENTS
   - DONORS: campaigns with high CPA, low ROAS, or falling conversion rate
   - RECIPIENTS: campaigns with low CPA, high ROAS, and high IS lost to budget

3. MODEL 3 SCENARIOS
   - CONSERVATIVE: reallocate 10% of total budget
   - MODERATE: reallocate 20%
   - AGGRESSIVE: reallocate 35%

   For each scenario, show:
   - Which campaigns lose budget and how much
   - Which gain budget and how much
   - Total projected conv (vs current)
   - Projected blended CPA
   - Projected blended ROAS
   - Risk assessment (what could go wrong)

4. RECOMMENDATION
   - Which scenario to implement and why
   - 2-week rollout plan (don't shift 100% at once)
   - What to monitor over next 14 days to confirm it worked

Output: before/after table for the 3 scenarios + client-ready final recommendation.

You are a senior Google Ads copywriter for agencies.

Client: [NAME]
Product/service: [DESCRIBE]
Target audience: [WHO BUYS]
Differentiators: [TOP 3]
Top keywords for this ad group: [3-5 KEYWORDS]
Current best-performing RSAs: [PASTE]
Competitor ads: [PASTE SCREENSHOTS OR LINKS]
Vertical: [to calibrate tone and compliance]

Generate 2 COMPLETE RSA variations for A/B test:

VERSION A — DIRECT / RATIONAL
- Focus on features, price, proof
- Factual tone

VERSION B — EMOTIONAL / ASPIRATIONAL
- Focus on outcome, transformation, feeling
- Inspirational tone

Each RSA needs:
- 15 HEADLINES (30 chars max each):
  * 3 with primary keyword embedded
  * 3 with unique value propositions
  * 3 with social proof / numbers
  * 3 with urgency / CTA
  * 3 with benefit-focused
- 4 DESCRIPTIONS (90 chars max each):
  * 1 feature-focused
  * 1 benefit-focused
  * 1 social-proof
  * 1 CTA-focused

For each RSA, return:
- Recommended pinning (pos 1, 2, 3 — which headlines to pin)
- Hypothesis: why this variation should outperform
- Suggested test duration based on volume
- Validation: char count per headline/description
- Compliance flag if any copy has risk (absolute claims, before/after, etc — relevant for health, financial, legal)

Output in XLSX-ready table format to paste directly into Google Ads Editor (columns: Campaign, Ad Group, Headline 1-15, Description 1-4, Pinning).

You are a landing page optimization specialist for Google Ads.

Client: [NAME / VERTICAL]
Analyze alignment between ads and LP:

- Ad group keywords: [LIST]
- Current headlines + descriptions: [PASTE]
- LP URL: [URL]
- LP content: [PASTE FULL TEXT OR PROVIDE URL FOR CLAUDE TO FETCH]

Score each LP on 1-10 scale across 7 dimensions:

1. MESSAGE MATCH (2x weight): does LP headline echo ad and keyword intent?
2. KEYWORD PRESENCE: do target keywords appear in H1, subheads, body?
3. INTENT ALIGNMENT (2x weight): does page satisfy what searcher wanted?
4. CTA CLARITY: is there a clear next step above the fold?
5. TRUST SIGNALS: reviews, testimonials, badges, guarantees?
6. MOBILE EXPERIENCE: load speed, layout, tap targets?
7. FORM FRICTION: number of fields, info requested, perceived effort?

Overall score: weighted average (message match + intent = 2x weight)

For each LP scoring below 7:
- Specific changes (rewrite EXACT headline, add section X)
- Expected QS impact (which component improves)
- Expected conversion rate impact
- Priority: Critical / High / Medium

Output:
- Table with score per dimension
- Prioritized changes list by impact
- Textual mockup of how optimized LP should look (headline, subhead, CTA, section order)
- Friction killer flags (form too long, hidden CTA, etc.)

Vertical: [to calibrate best practices — e.g. behavioral health needs 988 disclosure, healthcare needs visible credentials, etc.]

You are the agency head writing the weekly update for the client.

Client: [NAME]
Vertical: [auto repair / home services / behavioral health / etc.]
Language: [EN / PT / ES]
Tone: optimistic but honest, no hype, no absolute promises
Relationship: [new / 3+ months / long-term]

Data attached:
- Last 14 days performance vs prior 14 days
- Top wins of the week
- Any known issues

Write email following structure:

SUBJECT (max 8 words, no clickbait)

PARAGRAPH 1 — HIGHLIGHT
One sentence highlighting the win of the week in outcome terms (not metric).
Bad: "Conversions up 23%"
Good: "Generated 17 new service calls this week — your best week in 60 days."

PARAGRAPH 2 — NUMBERS THAT MATTER
3-5 metrics in prose (not bullet), comparing week vs prior.
Always include: leads/conversions, CPL, and something showing efficiency (CTR or ROAS).

PARAGRAPH 3 — WHAT WE DID
2-3 specific actions taken (not technical jargon).
Bad: "Optimized bidding strategy"
Good: "Shifted budget toward your top 3 service keywords"

PARAGRAPH 4 — WHAT WE'RE WATCHING (if applicable)
1-2 honest concerns, framed as "watching closely" not "broken".

PARAGRAPH 5 — NEXT WEEK
2 specific actions you'll run.

SIGNATURE
[NAME], [TITLE]
[AGENCY]

Rules:
- Max 250 words
- No em-dashes (—), use comma or period
- No words "definitely", "guaranteed", "best in market"
- No emoji unless client uses them
- Tone: client without PPC background understands without asking

You are the agency head preparing the monthly review PPT for the client.

Client: [NAME / VERTICAL]
Report audience: [owner / CMO / marketing director — calibrates technical vs business]
Month: [MONTH]

Data attached:
- Month performance vs prior month
- Performance vs same month last year (if available)
- Top performing campaigns/ad groups/keywords

Structure PPT content in slides:

SLIDE 1 — TITLE
Client | Month | "Marketing Performance Review"

SLIDE 2 — EXECUTIVE SUMMARY (1 slide, max 4 bullets)
- Headline stat (the MOST important thing of the month in 1 outcome-driven sentence)
- Total spend
- Volume (leads / conv / revenue)
- Efficiency (CPA / ROAS)

SLIDE 3 — KPI DASHBOARD (visual, 4 KPIs)
- Metric | Current Month | Prior Month | % Change | Vs target
- Pick 4: Spend, Conversions, CPA, ROAS (adjust per vertical)

SLIDE 4 — WHAT WORKED
3 points in business language (not PPC jargon)
Each point: WHAT happened, WHY it happened, IMPACT ($)

SLIDE 5 — WHAT NEEDS ATTENTION
1-2 honest points
Framing: "monitoring closely" + "here's the plan"
Include $ impact of inaction

SLIDE 6 — TOP PERFORMERS
Top 3 campaigns or ad groups with:
- Spend
- Conv
- CPA
- Hypothesis on why they're winning

SLIDE 7 — NEXT MONTH — PLAN
3-5 specific actions with expected outcome
Include budget recommendation if applicable

SLIDE 8 — THANK YOU + NEXT MEETING

For each slide give me:
- Slide title
- Bullet points with ready text
- Suggested visual (chart type, or "icon list", or "comparison table")
- Speaker notes if presenting live

Output format: structured to drop straight into agency PPT template.

You are a call handling specialist for paid media leads.

Client: [NAME / VERTICAL]
Attached: spreadsheet or call transcripts from last [14/30] days with:
- Call ID
- Duration
- Source (Google Ads, Meta, organic)
- Outcome (booked / no answer / disqualified / hang up / other)
- Transcription (if available)

Analyze:

1. CALL CLASSIFICATION
   - % Good calls (booked appointment OR qualified lead)
   - % Bad calls (no answer, abandoned, irrelevant, robocall)
   - % Recoverable calls (qualified but didn't convert due to bad handling)

2. GOOD CALL EXAMPLES (pick 2-3)
   - Call snippet
   - What the agent did right
   - Final outcome

3. BAD CALL EXAMPLES (pick 2-3, focusing on RECOVERABLE)
   - Call snippet
   - What went wrong (agent didn't answer, didn't follow up, bad transfer, wrong info)
   - Estimated financial impact of lost call

4. SYSTEMATIC PATTERNS
   - Time of day / day of week with most misses?
   - Specific staff person with worst handling?
   - Type of question that always derails the call?

5. CLIENT RECOMMENDATIONS
   - 3 specific changes in call handling
   - Recommended script or framework for first 30 seconds
   - When to expand hours (with data)

Tone: constructive, not accusatory. Client needs to hear this without taking offense.

Output format:
- Executive summary (2 sentences)
- Classification table
- 3 good calls (with timestamp and quote)
- 3 bad calls (with timestamp and quote + lesson)
- 3 prioritized recommendations with estimated $ impact

You are a senior strategist preparing the new client onboarding brief.

Client: [NAME]
Website: [URL]
Vertical: [VERTICAL]
Service area: [GEO]
Agreed monthly budget: $[X]
Conversion event: [calls / forms / online sales]
Direct competitors identified by client: [LIST 3-5]

Do active research and generate onboarding brief with:

1. CLIENT SNAPSHOT
   - What they sell and to whom
   - Differentiators (from website)
   - Digital maturity (have GA4? Tracking ok? CRM?)

2. COMPETITIVE LANDSCAPE
   - Top 5 real competitors in service area
   - What each is doing in Google Ads (estimate via SimilarWeb, SpyFu, Auction Insights if access)
   - Identified gap: what client can attack where no one else is

3. RECOMMENDED POSITIONING
   - Central message for paid media
   - 3 copy angles to test
   - What NOT to use (risky claims, weak words)

4. CAMPAIGN ARCHITECTURE (initial proposal)
   - How many campaigns, which themes
   - Brand vs non-brand split
   - Match type strategy
   - Starting bid strategy (Manual / Max Conv / tCPA)
   - Budget allocation %

5. PROPOSED KPIS
   - Primary KPI (volume)
   - Secondary KPI (efficiency)
   - Health KPI (quality)
   - Targets for first 30/60/90 days (with justification)

6. RISKS & ASSUMPTIONS
   - What can go wrong in first 30 days (and how to avoid)
   - Assumptions to validate (tracking, LP, call handling)

7. 90-DAY ROADMAP
   - Week 1-2: setup
   - Week 3-6: learning
   - Week 7-12: optimization + scale

Output: 2-3 page brief, format for me to present on kickoff call.

You are a competitive intelligence analyst for paid media.

Client: [NAME / VERTICAL / GEO]
Top 5 direct competitors: [LIST]
Platforms to monitor: [Google Ads / Meta / both]

Data attached:
- Client Auction Insights (last 30d)
- Screenshots or exports from Meta Ad Library for competitors
- Current SERP screenshots for key search terms

Generate monthly Competitor Snapshot:

1. AUCTION LANDSCAPE
   - Top 5 advertisers competing with client
   - Comparative impression share
   - Overlap rate + position-above-rate
   - New entrants vs last month (alert)

2. COMPETITIVE COPY MAP
   - For each competitor: 3 recurring copy patterns (most-used headlines)
   - Active promotions (% off, free trial, financing, etc.)
   - Trust signals used (years, reviews, certifications)
   - Most common CTAs

3. POSITIONING ANALYSIS
   - How each competitor positions (price / quality / speed / experience)
   - Where client is overlapping vs where there's white space

4. PROMOTIONAL ENVIRONMENT
   - What type of offer the market is running now
   - Recommendation: should client match? Differentiate? Ignore?

5. STRATEGIC GAPS
   - 3 specific opportunities where competitor is weak
   - 2 threats where competitor is gaining momentum

6. RECOMMENDED ACTIONS NEXT MONTH
   - 3 specific moves for client (copy update, budget shift, new ad group, etc.)
   - Each with estimated investment and expected impact

Output in client-ready format: 2-3 pages, visual when possible (comparison table), analytical but accessible tone.