Mixpanel machine learning algorithms surface insights and detect anomalies in project data. These automatic insights can be found in the Notifications menu of a project, and will provide links to more details about the surfaced information.
Anomaly Detection, Explains, and Automatic Segmentation are all Mixpanel Machine Learning models that aid in uncovering data that requires attention but may have otherwise been missed.
Automatic insights that are surfaced in notifications will be deleted after 30 days. Reports will leave your notifications if they are marked as "Not Useful" or the "x" is clicked.
Anomaly Detection uncovers data irregularities and will present them in the Notifications center. Clicking on an anomaly will direct you to the specific location of the report where the anomaly was detected. Additionally, Mixpanel Explains will provide an explanation to the anomaly, helping diagnose and address discrepancies.
Detected anomalies will appear in reports as yellow dots.
Anomaly detections are available in the following reports:
Mixpanel Explains are machine learning generated explanations for detected anomalies. Explains are coupled with Anomaly Detection, and will be presented with detected anomalies from the notifications center.
As Explanations are tied to anomaly detections, they are available in the reports that contain anomaly detections, including:
Mixpanel Automatic Segmentation automatically segments each funnel and retention report that is saved or on part of a dashboard by a variety of properties. The machine learning model then analyzes the findings and surfaces insights that deviate from the normal trends.
Automatic Segmentations will appear in reports as green dots.
Mixpanel analyzes and will automatically segment all reports on the My Dashboard report, Funnels & Retention reports saved on Dashboards and viewed in the last 14 days, and saved Funnel & Retention reports viewed in last 90 days.