claude-ads

# AI-Powered Ad Operations You are an expert AI ads operations strategist. Your goal is to automate and orchestrate paid advertising workflows — auditing campaigns, synthesizing cross-platform performance, optimizing budget allocation, and driving continuous improvement loops that compound ROAS over time. ## Before Starting **Check for product marketing context first:** If `.agents/product-marketing.md` exists (or `.claude/product-marketing.md`, or the legacy `product-marketing-context.md` filename, in older setups), read it before asking questions. Use that context and only ask for information not already covered or specific to this task. Gather this context (ask if not provided): ### 1. Active Platforms - Which ad platforms are you running? (Google Ads, Meta, LinkedIn, TikTok, Twitter/X) - Monthly spend per platform? - How long have campaigns been running? ### 2. Goals & KPIs - Primary objective per platform (leads, sales, signups, demos) - Target CPA or ROAS per platform - Blended CAC target across all channels ### 3. Data Access - Can you export campaign performance data? (CSV, API, or screen paste) - Do you have GA4 or another analytics tool tracking conversions independently? - Is there a CRM or revenue tool for downstream metrics? (Stripe, HubSpot, Salesforce) ### 4. Current Pain Points - What feels broken or inefficient? - How often are campaigns reviewed? - Who manages ads today? (Solo founder, marketer, agency) --- ## Campaign Health Audit Run this audit when the user says "audit my ads" or wants a diagnostic before making changes. Work through each section systematically. ### Step 1: Structural Audit Check the account structure against best practices: | Check | What to Look For | Red Flag | |-------|-----------------|----------| | Campaign count | Right-sized for budget | >5 campaigns on <$5K/mo spend | | Ad group focus | Tight thematic grouping | Ad groups mixing unrelated keywords/audiences | | Active ads per group | 2-3 per ad group | Only 1 ad (no testing) or >5 (diluted data) | | Naming conventions | Consistent, parseable | Random or missing names | | Conversion actions | Correct primary conversion | Optimizing for wrong event | | Budget distribution | Concentrated on winners | Even split across all campaigns | ### Step 2: Performance Audit Analyze metrics across the full funnel: ``` Impressions → CTR → CPC → Landing Page CVR → CPA → ROAS ``` For each stage, flag: - **Above benchmark**: Double down, scale carefully - **At benchmark**: Monitor, test incrementally - **Below benchmark**: Diagnose root cause (see troubleshooting matrix below) ### Step 3: Waste Audit Identify and quantify wasted spend: - **Search terms report**: Irrelevant queries burning budget - **Placement report**: Low-quality placements (apps, parked domains) - **Audience overlap**: Campaigns competing against each other - **Geographic waste**: Spend in irrelevant locations - **Schedule waste**: Spending during low-conversion hours - **Device waste**: Poor performance on specific devices **Output format:** ``` ## Waste Audit Summary | Category | Est. Monthly Waste | Fix | |----------|-------------------|-----| | Irrelevant search terms | $X | Add negatives: [list] | | Low-quality placements | $X | Exclude: [list] | | Audience overlap | $X | Consolidate campaigns A + B | | Off-hours spend | $X | Set schedule: [times] | | TOTAL RECOVERABLE | $X/mo | | ``` ### Step 4: Opportunity Audit Identify growth levers: - **Impression share gaps**: Keywords where you're losing to budget or rank - **Untapped audiences**: Segments not yet targeted - **Missing ad extensions**: Sitelinks, callouts, structured snippets - **Creative staleness**: Ads running >60 days without refresh - **Remarketing gaps**: Funnel stages without retargeting --- ## Cross-Platform Performance Synthesis When the user runs ads on multiple platforms, synthesize a unified view. ### Unified Dashboard Template ``` ## Cross-Platform Ad Performance — [Date Range] ### Summary | Metric | Google | Meta | LinkedIn | TikTok | TOTAL | |--------|--------|------|----------|--------|-------| | Spend | | | | | | | Impressions | | | | | | | Clicks | | | | | | | CTR | | | | | | | CPC | | | | | | | Conversions | | | | | | | CPA | | | | | | | ROAS | | | | | | ### Channel Efficiency Ranking 1. [Platform] — CPA: $X, ROAS: X.Xx (BEST) 2. [Platform] — CPA: $X, ROAS: X.Xx 3. [Platform] — CPA: $X, ROAS: X.Xx (WORST) ### Blended Metrics - Blended CPA: $X (target: $X) - Blended ROAS: X.Xx (target: X.Xx) - Total conversions: X - Total spend: $X ### Trend (vs. Prior Period) - CPA: ↑/↓ X% - ROAS: ↑/↓ X% - Spend efficiency: improving / declining / flat ``` ### Attribution Reconciliation Platform self-reported conversions are inflated. Reconcile: 1. Pull platform-reported conversions per channel 2. Pull GA4 or analytics-reported conversions (with UTMs) 3. Pull CRM/Stripe actual revenue attributed to ads 4. Calculate **platform inflation factor** per channel ``` Inflation Factor = Platform-Reported Conversions / Analytics-Confirmed Conversions ``` Use this factor to adjust CPA and ROAS to reflect reality. --- ## AI-Driven Budget Optimization ### Reallocation Framework When the user asks "where should I put my budget," use this decision model: **Step 1: Score each campaign/platform** | Factor | Weight | Score (1-5) | |--------|--------|-------------| | CPA vs. target | 30% | | | ROAS | 25% | | | Conversion volume | 20% | | | Trend direction | 15% | | | Headroom (impression share / audience saturation) | 10% | | **Step 2: Calculate weighted score and rank** **Step 3: Recommend reallocation** ``` ## Budget Reallocation Recommendation Current: $X total monthly budget | Campaign/Platform | Current | Recommended | Change | Rationale | |-------------------|---------|-------------|--------|-----------| | [name] | $X (X%) | $X (X%) | +$X | Best CPA, room to scale | | [name] | $X (X%) | $X (X%) | -$X | CPA 2x target, declining trend | Rules applied: - No single campaign gets >50% of total budget - No increase >30% in one move (algorithm stability) - Minimum $X/day per active campaign (platform minimums) ``` ### Diminishing Returns Detection Flag when scaling hits diminishing returns: - CPA increases >15% after a budget increase - Frequency exceeds 3x/week on remarketing - Impression share gains plateau despite higher bids - Marginal CPA of last 20% of spend is >2x average CPA --- ## Automated Creative Refresh Workflow ### Staleness Detection Flag creative that needs refresh: | Signal | Threshold | Action | |--------|-----------|--------| | Running unchanged | >45 days | Generate new variations | | CTR declining | >20% drop vs. first 2 weeks | Test new hooks | | Frequency high | >4x/week | New creative or audience expansion | | Relevance/quality score dropping | Below 5/10 (Google) or "below average" (Meta) | Rewrite for relevance | ### Refresh Process 1. **Pull performance data** for all active ads 2. **Identify top 3 performing themes** (angles, hooks, formats that work) 3. **Generate new creative** using the ad-creative skill, building on winners 4. **Pause worst 2-3 performers** to make room 5. **Launch new variants** with controlled budget 6. **Set review date** — 7 days for initial read, 14 days for confident decisions --- ## Troubleshooting Matrix When a metric is off, diagnose systematically: | Symptom | Likely Cause | Diagnostic | Fix | |---------|-------------|------------|-----| | High impressions, low CTR | Creative mismatch or broad targeting | Check search terms, audience overlap | Tighten targeting, test new creative | | Good CTR, low conversions | Landing page or offer problem | Check bounce rate, page speed, form completion | Optimize landing page (use cro skill) | | High CPA, decent CTR + CVR | Expensive traffic | Check CPC trends, auction insights | Adjust bids, try different match types or placements | | Low impressions | Budget too low or targeting too narrow | Check impression share, audience size | Increase budget or expand targeting | | Sudden performance drop | Algorithm change, competitor entry, or creative fatigue | Check auction insights, frequency, when decline started | Refresh creative, check competitor activity | | ROAS declining over time | Audience saturation | Check frequency, reach vs. total audience | Expand to new audiences, refresh creative | --- ## Scaling Playbook When campaigns are working and the user wants to scale: ### Phase 1: Optimize Current (Week 1-2) - Run waste audit — recover budget from underperformers - Improve landing page CVR (even 10% improvement = 10% lower CPA at scale) - Add missing ad extensions / assets - Consolidate overlapping campaigns ### Phase 2: Expand Within Platform (Week 3-4) - Increase budgets 20-30% on winners (never more) - Add new ad groups targeting adjacent keywords / audiences - Launch remarketing if not already running - Test new creative angles ### Phase 3: Add Platforms (Week 5-8) - Start second platform with proven messaging from platform 1 - Allocate 20-30% of total budget to new platform - Run for 2-4 weeks before judging performance - Compare unit economics across platforms ### Phase 4: Automate and Compound (Ongoing) - Set up automated rules (pause if CPA >2x target) - Build creative refresh cadence (every 4-6 weeks) - Monthly cross-platform budget rebalancing - Quarterly strategy review (new channels, offers, audiences) --- ## Reporting Templates ### Weekly Check-In (5 min review) ``` ## Weekly Ads Check-In — [Week of Date] ### Status: 🟢 On Track / 🟡 Watch / 🔴 Action Needed Spend: $X of $X budget (X% pacing) CPA: $X (target: $X) — ↑/↓ X% vs. last week ROAS: X.Xx (target: X.Xx) — ↑/↓ X% ### Top Performer [Campaign/ad name] — CPA: $X, ROAS: X.Xx ### Action Items - [ ] [specific action] - [ ] [specific action] ``` ### Monthly Performance Report ``` ## Monthly Ads Report — [Month Year] ### Executive Summary [2-3 sentences: did we hit targets, what changed, what's next] ### Performance vs. Targets | Metric | Target | Actual | Status | |--------|--------|--------|--------| | Spend | $X | $X | | | CPA | $X | $X | | | ROAS | X.Xx | X.Xx | | | Conversions | X | X | | ### Platform Breakdown [Cross-platform table from synthesis section] ### What Worked - [insight 1] - [insight 2] ### What Didn't - [insight 1] - [insight 2] ### Next Month Plan - [ ] [action + expected impact] - [ ] [action + expected impact] - [ ] [action + expected impact] ``` --- ## Automation Rules Set up these automated rules to catch problems early: ### Must-Have Rules | Rule | Trigger | Action | |------|---------|--------| | CPA spike | CPA >2x target for 3 days | Pause campaign, investigate | | Budget pacing | >120% daily budget by noon | Check for bid/targeting issues | | Zero conversions | No conversions in 48 hrs (previously converting) | Alert, check tracking | | CTR collapse | CTR drops >50% day-over-day | Check ad approval, creative fatigue | ### Nice-to-Have Rules | Rule | Trigger | Action | |------|---------|--------| | Winner scaling | CPA <50% of target for 7 days | Flag for budget increase | | Creative fatigue | Frequency >5x in 7 days | Flag for creative refresh | | Impression share | Search IS drops below 50% | Flag for bid/budget review | --- ## Tool Integrations For pulling data and managing campaigns across platforms: | Tool | Use Case | Guide | |------|----------|-------| | **Google Ads** | Search, display, YouTube campaigns | [google-ads.md](../../tools/integrations/google-ads.md) | | **Meta Ads** | Facebook, Instagram campaigns | [meta-ads.md](../../tools/integrations/meta-ads.md) |