BigQuery anomaly detection uses BigQuery ML models to identify unusual patterns or deviations in datasets, often with time series or clustering algorithms. While it’s powerful for predictions, forecasting, and business decisions, it requires manual setup, model training, and maintenance, as well as a clear definition of “normal” behavior. With the right configuration, it can surface potential data quality issues, but unlike purpose-built tracking solutions, it does not provide continuous, real-time monitoring of events, pixels, or attribution, leaving live tracking issues potentially undetected.
Trackingplan, on the other hand, is purpose-built for detecting data quality issues in analytics implementations. Instead of relying on custom model configuration, it automatically monitors tracking behavior, validates specifications, and alerts teams to issues such as missing events, unexpected parameter changes, traffic drops, or implementation inconsistencies — before they impact reporting and business decisions.While BigQuery ML can help identify statistical anomalies, Trackingplan focuses directly on tracking integrity, giving marketing and analytics teams visibility into what is breaking, why, and where — without requiring data science expertise.
We had the concept on the back burner since around 2018, soon after BigQuery appeared. But as a development project, it was hard to get moving. New clients, urgent tasks, and shifting priorities kept pushing it further down the roadmap.
Synthetic scans can miss real user issues. Trackingplan monitors real interactions in real time.
Manual UTM and campaign checks are slow and error-prone. Trackingplan detects inconsistencies automatically.
Manual setup, simulated tests, and scheduled executions can miss live issues. Trackingplan monitors real user interactions and tracking quality automatically.
Predefined plans need manual review for changes. Trackingplan validates live tracking automatically.
Manual thresholds and delayed notifications can miss issues. Trackingplan provides real-time alerts for anomalies and broken tracking.
