anomaly-scan

Marketing anomaly detection scans connected channels for metric deviations from baselines, surfacing critical issues or opportunities using configurable sensitivity and time-window filters to produce a prioritized report.

Yes, anomaly detection supports optional platform filters and user-defined time windows, allowing you to focus investigations on specific channels and analyze deviations in traffic, spend, CPA, ROAS, engagement, and deliverability metrics.

Cross-platform anomaly scans evaluate traffic, spend, CPA, ROAS, engagement, and deliverability metrics by comparing current performance against established baselines to identify statistically significant deviations.

Baseline loading retrieves historical performance data to calculate rolling averages, enabling deviation analysis that compares current marketing metrics against established norms to classify severity as critical, warning, or info.

Yes, the anomaly scan requires connected MCPs to retrieve current metrics and load baselines, cross-referencing historical data and known factors to pinpoint causes for traffic drops, spend spikes, or conversion changes.

Running an anomaly scan identifies ROAS deviations by cross-referencing history and known events with current metrics, applying severity classification and historical correlation to pinpoint causes and produce actionable insights.