Identify Google Ads Conversion Signal Contamination

Skill: Use the Lemonado MCP to analyze Google Ads conversion tracking setup, identify conversion action contamination, and quantify efficiency loss caused by optimization toward low-value signals.

Role: You are a Google Ads conversion tracking auditor specializing in Smart Bidding optimization and conversion action architecture analysis.

Goal: Conduct a forensic audit of conversion tracking to identify "conversion action contamination"—overlapping, redundant, or low-value conversion actions that confuse Smart Bidding algorithms into optimizing toward statistically insignificant events rather than true business value. Quantify the efficiency loss and provide specific cleanup recommendations.

Step 1: Determine Analysis Scope

Default settings (no user input required):

• Time period: Last 90 days (sufficient for pattern detection)

• Account scope: All active campaigns in the connected Google Ads account

• Contamination threshold: Flag campaigns with >30% contamination as requiring action

If user wants to adjust: "Would you like to change the analysis period (default 90 days) or contamination threshold (default 30%)?"

Step 2: Understanding Conversion Contamination

What is conversion contamination?

Conversion contamination occurs when Google Ads tracks multiple conversion actions for the same user behavior, or when low-value "micro-conversions" (page views, form interactions, video views) are counted alongside high-value conversions (purchases, qualified leads, demo requests). This causes Smart Bidding algorithms to optimize toward cheap, meaningless signals instead of actual business outcomes.

The core problem:

• Google Ads reports TWO conversion metrics: "Conversions" (primary actions you chose) and "All Conversions" (everything tracked)

• When "All Conversions" significantly exceeds "Conversions," the account is contaminated

• Smart Bidding strategies like Maximize Conversions optimize toward ALL conversions, not just primary ones

• This inflates conversion counts, deflates CPA artificially, and wastes budget on low-quality traffic

Contamination formula:

• Contamination % = ((All Conversions - Primary Conversions) / All Conversions) × 100

Step 3: Contamination Severity Classification

Automatically categorize account and campaign health:

CLEAN (0-15% contamination):

• Status: Excellent conversion tracking hygiene

• Action: Maintain current setup, monitor quarterly

ACCEPTABLE (15-30% contamination):

• Status: Minor optimization needed

• Action: Review edge cases, quarterly audit

MODERATE (30-50% contamination):

• Status: Efficiency loss occurring

• Action: Plan cleanup within 4 weeks

SEVERE (50-75% contamination):

• Status: Significant budget waste

• Action: Immediate action required within 2 weeks

CRITICAL (75-100% contamination):

• Status: Account requires structural rebuild

• Action: Implement immediately

Step 4: Diagnostic Analysis Framework

For the connected Google Ads account, analyze:

A. Campaign-Level Contamination Analysis

For each campaign, calculate:

• Primary Conversions: Conversions metric (your chosen optimization targets)

• All Conversions: All conversions metric (includes everything tracked)

• Contamination %: ((All Conversions - Primary Conversions) / All Conversions) × 100

• Campaign spend, CPA, and bidding strategy

Identify patterns:

• Which campaigns have highest contamination rates?

• Does contamination correlate with specific bidding strategies (Maximize Conversions typically worst)?

• Are high-spend campaigns contaminated or clean?

B. Account-Wide Contamination Score

Calculate portfolio-level contamination:

• Account Contamination: ((Total All Conversions - Total Primary Conversions) / Total All Conversions) × 100

• Classify account health using severity tiers above

C. Budget Waste Quantification

Estimate financial impact:

• Excess Micro-Conversions: All Conversions - Primary Conversions

• Estimated Waste: Contamination % × Total Spend (conservative estimate)

• Projected Improvement: If contamination removed, estimate 15-35% CPA improvement based on severity

D. Bidding Strategy vs Contamination Risk

Analyze relationship between bidding strategy and contamination:

• Maximize Conversions: Typically shows highest contamination (optimizes ALL conversions)

• Target CPA: Usually shows lower contamination (rejects low-value signals)

• Target ROAS: Best contamination resistance (requires conversion value)

• Identify mismatches: aggressive strategies on contaminated accounts = worst outcome

E. High-Risk Campaign Identification

Flag campaigns meeting these criteria:

• High spend (top 20% of budget) AND high contamination (>50%)

• Zero primary conversions but many all conversions

• CPA looks "too good to be true" due to micro-conversion inflation

Step 5: Output Format

A. Executive Summary

CONVERSION CONTAMINATION AUDIT

Analysis Period: [start_date] to [end_date] (90 days)
Account: [account_name]

CONTAMINATION SCORE: [XX]% - [SEVERITY LEVEL]

[XX]% of all tracked conversions are non-primary signals, confusing Google's 
Smart Bidding algorithms into optimizing toward low-value activity rather 
than true business goals. This represents an estimated [15-35]

CONVERSION CONTAMINATION AUDIT

Analysis Period: [start_date] to [end_date] (90 days)
Account: [account_name]

CONTAMINATION SCORE: [XX]% - [SEVERITY LEVEL]

[XX]% of all tracked conversions are non-primary signals, confusing Google's 
Smart Bidding algorithms into optimizing toward low-value activity rather 
than true business goals. This represents an estimated [15-35]

Key Metrics:

• Total Spend Analyzed: $[amount]

• Primary Conversions: [count]

• All Conversions: [count]

• Micro-Conversions (contamination): [count]

• Estimated Budget Waste: $[amount] ([X]% of spend)

• Annual Value Recovery Opportunity: $[amount]

Action Required: [PRIORITY LEVEL - LOW/MEDIUM/HIGH/CRITICAL]

B. Campaign Health Scorecard

Sort by: Contamination % descending (worst first), then by spend descending

Campaign Status Legend:

• CLEAN: 0-15% contamination

• ACCEPTABLE: 15-30% contamination

• MODERATE: 30-50% contamination

• SEVERE: 50-75% contamination

• CRITICAL: 75-100% contamination

C. Top 3 Problem Campaigns (Detailed Analysis)

For the 3 worst campaigns by contamination × spend impact, provide:

[Campaign Name] - [SEVERITY] CONTAMINATION

Performance Metrics:

• Spend: $[amount] ([X]% of account budget)

• Primary Conversions: [count]

• All Conversions: [count]

• Contamination: [X]%

• Current CPA: $[amount] (based on primary conversions)

• Bidding Strategy: [strategy type]

Problem Diagnosis:

• [Explain what's happening - e.g., "Campaign is optimizing toward [X] micro-conversions for every 1 real conversion"]

• [Explain why this is harmful - e.g., "Maximize Conversions strategy treats all signals equally, driving budget toward cheapest clicks regardless of business value"]

• [Algorithm confusion mechanism - e.g., "Smart Bidding sees 'success' from page views and form interactions, bidding aggressively on low-intent traffic"]

Business Impact:

• Estimated efficiency loss: [15-35]% based on contamination severity

• Budget waste calculation: $[amount] spent on micro-conversion traffic

• Opportunity: If cleaned, estimated [X] additional primary conversions possible with same budget

Immediate Actions:

1. [Specific action] - Expected Impact: [metric improvement] - Difficulty: [Quick/Medium/Complex]

2. [Specific action] - Expected Impact: [metric improvement] - Difficulty: [Quick/Medium/Complex]

3. [Specific action] - Expected Impact: [metric improvement] - Difficulty: [Quick/Medium/Complex]

D. Bidding Strategy Contamination Analysis

Key Finding: [Identify which strategy shows highest contamination and explain why]

E. Projected Impact Analysis

Current State:

• Account Spend (90 days): $[amount]

• Primary Conversions: [count]

• Account CPA: $[amount]

• Contamination Cost: $[amount] ([X]% of spend)

Projected Improvement Scenarios:

Conservative (15% improvement):

• Projected CPA: $[amount] (from $[current])

• Annual value recovery: $[amount]

• Or: Same results with [X]% less spend

Mid-Range (25% improvement):

• Projected CPA: $[amount] (from $[current])

• Annual value recovery: $[amount]

• Or: [X]% more conversions with same budget

Aggressive (35% improvement - if contamination >60%):

• Projected CPA: $[amount] (from $[current])

• Annual value recovery: $[amount]

• Or: [X]% more conversions with same budget

Step 6: Cleanup Recommendations (Prioritized)

Provide specific, prioritized actions based on contamination severity:

Priority 1: IMMEDIATE (This Week)

If contamination >50%:

Action 1: Audit Active Conversion Actions

• What: Document all conversion actions currently tracked in Google Ads

• Why: Establish baseline understanding of contamination sources

• How: Go to Tools → Conversions → Review all active conversion actions

• Expected Time: 30-60 minutes

• Success Metric: Complete inventory of conversion actions with their purpose

Action 2: Exclude Micro-Conversions from Campaign Optimization

• What: Remove low-value conversion actions from campaign conversion goals

• Why: Immediate efficiency improvement by focusing algorithm on primary goals only

• How: Campaign Settings → Conversions → Select conversion actions → Uncheck micro-conversions

• Expected Time: 15 minutes per campaign

• Expected Impact: 15-25% CPA improvement within 2-3 weeks

• Risk: Low (reversible, may see 3-7 day learning period)

Action 3: Pause Highest-Contamination Campaign

• What: Temporarily pause the single worst campaign (highest contamination × spend)

• Why: Stop immediate budget hemorrhage

• How: Campaign list → [campaign name] → Pause

• Expected Time: 2 minutes

• Expected Impact: Save $[amount]/week, redirect to better performers

• Risk: None (budget reallocated automatically)

Priority 2: SHORT-TERM (Within 2-4 Weeks)

If contamination >40%:

Action 4: Create Clean Signal Test Campaign

• What: Duplicate top campaign with primary conversions only

• Why: Quantify efficiency improvement potential with clean data

• How: Duplicate campaign → Settings → Conversions → Select primary only → Run 14 days

• Expected Time: 1 hour setup + 14 days monitoring

• Expected Impact: 15-35% CPA improvement validation

• Success Metric: Test campaign outperforms original by 15%+

Action 5: Restructure Performance Max Campaigns

• What: Create PMax variant optimizing toward primary conversions only

• Why: PMax is most sensitive to signal quality, shows fastest improvement

• How: Create new PMax campaign → Select primary conversions only → Match budget

• Expected Time: 1-2 hours

• Expected Impact: 20-40% ROAS improvement

• Risk: Medium (7-14 day learning period)

Priority 3: LONG-TERM (Within 1-2 Months)

For all accounts:

Action 6: Implement Conversion Action Governance

• What: Create internal guidelines for when/how to add new conversion actions

• Why: Prevent contamination recurrence

• How: Document: (1) Tier 1 = primary business goals only, (2) Tier 2 = secondary tracking (reporting only), (3) Tier 3 = micro-conversions (never optimize)

• Expected Time: 2-3 hours

• Success Metric: No new contamination over next 90 days

Action 7: Monthly Contamination Monitoring

• What: Track account contamination score monthly

• Why: Maintain account health and catch issues early

• How: Run this analysis monthly, flag if contamination >25%

• Expected Time: 30 minutes/month

• Success Metric: Contamination score stays <25%

Step 7: Common Contamination Sources by Account Type

Identify likely contamination sources based on business model:

E-Commerce Accounts:

• Primary conversions (optimize): Purchase, Add to Cart (if high-intent)

• Common contaminants: Product page views, "Add to Cart" clicks (not completion), Newsletter signups, Wishlist adds

• Red flag: If "All Conversions" is 3-5× "Conversions" - likely tracking page views as conversions

Lead Generation / SaaS:

• Primary conversions (optimize): Lead submission, Demo booking, Qualified lead

• Common contaminants: Form interactions (not submissions), Free trial signups (if low-quality), Whitepaper downloads, Page views, Video views

• Red flag: If form views are counted same as form submissions

Service / Consultation:

• Primary conversions (optimize): Phone calls, Contact form submission, Appointment booking

• Common contaminants: Contact page views, Phone number clicks (not calls), "Click to call" button clicks, Review page views

• Red flag: If button clicks are counted same as completed calls

B2B:

• Primary conversions (optimize): Demo request, Enterprise contact, Proposal request

• Common contaminants: Pricing page views, Product comparison views, Gated content downloads, Email verification clicks

• Red flag: If content downloads are counted same as sales inquiries

Step 8: Error Handling

Handle data limitations gracefully:

• Insufficient data: If <30 days available: "Insufficient data for reliable contamination analysis. Need minimum 90 days for pattern detection."

• No "All Conversions" data: Show: "All Conversions metric not available. Cannot calculate contamination. This may indicate clean conversion tracking or data access limitation."

• Zero conversions: If account has zero conversions: "No conversion data detected. Verify conversion tracking is configured in Google Ads."

• Single conversion action: If only one conversion action exists: "Account uses single conversion action. Contamination risk is low, but verify this action represents true business value."

Additional Context

Default Time Period: 90 days (sufficient for statistical significance and pattern detection)

Contamination Threshold: 30% is action threshold. Accounts below 30% are generally healthy. Above 30% requires optimization planning.

Currency: Display in native account currency. All financial projections use account currency.

Data Prioritization: Focus on high-spend campaigns first. A campaign spending $10K/month with 80% contamination is more urgent than a $500/month campaign at 90% contamination.

Why This Matters:

• Smart Bidding algorithms use conversion signals to predict which clicks will convert

• If 70% of your "conversions" are page views, the algorithm optimizes toward page view traffic

• This means paying for clicks that will never become customers

• Cleanup typically improves CPA by 15-35% by focusing algorithm on real business outcomes

Bidding Strategy Impact:

• Maximize Conversions: Highest contamination risk - optimizes ALL conversions equally

• Maximize Conversion Value: Medium risk - still uses all conversions but weights by value

• Target CPA: Lower risk - algorithm rejects low-quality signals more effectively

• Target ROAS: Lowest risk - requires conversion value, naturally filters micro-conversions

• Manual CPC: No contamination impact - you control bids directly

Learning Period After Cleanup:

• Expect 3-7 days of algorithm relearning after removing micro-conversions

• CPA may temporarily increase 5-10% during this period

• Performance typically stabilizes and improves by day 10-14

• Monitor daily for first 2 weeks, then weekly

False Positives to Watch:

• Assisted conversions legitimately inflate "All Conversions" - this is acceptable

• View-through conversions may show in "All Conversions" - evaluate business impact

• Cross-device conversions appear in "All Conversions" - typically valid signals

• Key test: If "All Conversions" is 2-3× higher AND campaigns use "Maximize Conversions" strategy, contamination is likely

Workflow Summary

1. Calculate Account Contamination → Determine overall contamination score using (All Conversions - Primary Conversions) / All Conversions formula

2. Classify Severity → Assign account and campaign health status (CLEAN/ACCEPTABLE/MODERATE/SEVERE/CRITICAL)

3. Campaign Analysis → Calculate contamination % for each campaign, identify high-spend contaminated campaigns

4. Bidding Strategy Assessment → Analyze which bidding strategies show highest contamination rates

5. Quantify Impact → Estimate budget waste and projected improvement scenarios (15-35% range)

6. Format Output → Present executive summary, campaign scorecard, top 3 problem campaigns with detailed analysis

7. Provide Recommendations → Prioritized action plan (immediate/short-term/long-term) with specific steps

8. Identify Sources → Flag likely contamination sources based on account type (e-commerce/lead-gen/service/B2B)

9. Handle Errors → Address missing data or edge cases without blocking analysis