anomaly-detector

Detect spend spikes and performance drops by applying statistical anomaly detection using Z-score and IQR methods to your ad campaign metrics. This identifies outliers like sudden CTR or CVR drops in real-time.

Z-score and IQR anomaly detection works for monitoring ad metrics by calculating standard deviations and interquartile ranges from historical data. Values falling outside configured thresholds are flagged as outliers for automated alerting.

Yes, you can configure automated alert thresholds for Google Ads and Meta Ads performance monitoring. The system provides cross-platform query patterns to flag metrics exceeding projected run rates or standard deviation limits.

The best way to set up real-time monitoring for advertising spend anomalies is using Python algorithms for composite scoring. This allows you to automatically flag campaigns when daily spend exceeds 150% of projected run rates.

A limitation of using statistical anomaly detection for ad metrics is that Z-score and IQR methods require sufficient historical data to establish accurate baselines. Without enough data, normal fluctuations might trigger false outlier alerts.

Yes, you need Python to analyze hourly spend data for campaign anomalies. The Skill provides Python algorithms that perform real-time monitoring and composite scoring to identify unusual fluctuations in your advertising data.