Basics of Contribution Analysis

Contribution Analysis features begin with Anomaly Detection, where segments are applied, a training period selected to define a norm, and anomalous data returned. This subset of data is then analyzed for all relevant contributing factors based on the selected metric. The Contribution Report publishes this information through reports and interactive visualizations.

Basic Anomaly Detection

See Anomaly Detection for in-depth feature information.

Anomaly Detection displays data points across a time line based on the selected view period. The anomalies are set in a Training Period and visualized as a dotted line in the chart.



Anomalies are then identified using the following formula:

upper_bound

Upper level of the prediction interval. Values above this level are considered anomalous.

Represents a 95% confidence that values will be below this level.

lower_bound

Lower level of the prediction interval. Values below this level are considered anomalous.

Represents a 95% confidence that values will be above this level.

forecast

The predicted value based on the data analysis. This value is also the middle point between the upper and lower bounds.

Basic Contribution Analysis

Contribution Analysis queries tens of millions of data sets and applies advanced statistics and machine-learning to identify items that significantly impacted the anomaly being analyzed.

Dimensions analyzed include:

  • eVars
  • props (including pathing version—enter/exit)
  • out-of-the-box Analytics variables
  • Classifications and customer attributes
  • Mobile, video, Adobe Social/Survey/Target

As soon as you click Analyze in Anomaly Detection, Contribution Analysis performs analysis on the top 50,000 items per dimension.