skill-111-data-exclusion-tracking-incidents

Skill 111: Data Exclusion for Tracking Incidents

Purpose

Use data exclusions to prevent broken tracking periods from influencing Smart Bidding. Bad data during outages feeds bad bid decisions that compound over time without intervention.

Data Requirements

Data Source: Custom GAQL Required

GAQL Query:

SELECT
  segments.date,
  metrics.conversions,
  metrics.clicks,
  metrics.cost_micros
FROM campaign
WHERE segments.date DURING LAST_30_DAYS
  AND campaign.bidding_strategy_type IN (
    'TARGET_CPA', 'TARGET_ROAS', 'MAXIMIZE_CONVERSIONS', 'MAXIMIZE_CONVERSION_VALUE'
  )
ORDER BY segments.date

Run via /google-ads:get-custom with query name tracking_incident_detection.

Standard Data:

• data/performance/campaigns/*/campaign_metrics_30_days.md - Historical trends

Analysis Steps

1. Identify Historical Incidents: Review conversion drops, known tracking outages
2. Check Existing Exclusions: Verify if data exclusions already applied
3. Assess Unaddressed Incidents: Find tracking problems without exclusions applied
4. Evaluate Smart Bidding Impact: Determine if campaigns using Smart Bidding were affected

Thresholds

Condition	Severity
Tracking incident without data exclusion applied	Critical
Incident duration >72 hours without exclusion	Critical
Multiple incidents in 30 days without exclusions	Warning
Smart Bidding campaign affected by incident	Warning

Output

Short (default):

## Data Exclusion Audit
**Account:** [name] | **Incidents Detected:** [count] | **Exclusions Applied:** [count]

### Critical
- **[date range] tracking incident without exclusion** - Smart Bidding learning on bad data

### Recommendations
1. Apply data exclusion for [start date] to [end date]
2. Exclude at bid strategy level: [strategy names]
3. Monitor for 7-14 days for Smart Bidding recovery

Detailed adds:

• Historical incident analysis table
• Exclusion decision framework
• Recovery timeline by incident severity
• Documentation template for tracking incidents