Discovers hidden patterns within your data to explain statistical anomalies, and identify correlations behind unexpected customer actions, out-of-bound values, and sudden spikes or dips for metrics across audience segments.
Data anomalies are a real headache. Sometimes they mean things are going better than expected. But too often, it means that someone’s tagged something wrong or a campaign ended prematurely. Or worse. The ability to automatically call out anomalies is one powerful feature of Adobe Analytics, but catching them is only one part of the equation. Just as important is understanding what caused the anomaly. For many analysts, this second step took time and required manually digging through large and varied data sets.
With the Contribution Analysis feature in Adobe Analytics, this time-consuming task can now be done literally with the click of a button. Contribution Analysis works together with Anomaly Detection to help you quickly understand what’s going on in your data. Built into the Analysis Workspace, Contribution Analysis queries tens of millions of data set to replace long and difficult analysis with useful visualizations that help you make the best decisions.
Hide duplicate or uninteresting dimensions, outliers, and junk data in reports to get the most relevant answers.
Contribution scores help you analyze the significance of your data dimensions relevant to the anomaly.
Reveal hidden audience segments based on the contributing factors they have in common.
Contribution analysis uses intensive machine learning to uncover contributors much more quickly than would otherwise be possible.
Watch this video to see how you can see how contribution analysis works in Analysis Workspace.
Read more on our Help pages about the basics of contribution analysis.
Statistical modeling automatically finds unexpected trends in your data by analyzing metrics and determining a lower bound, upper bound, and expected range of values. When an unexpected spike or drop occurs, the system alerts you in the report.
We make it easy to create and manage alerts based on data anomalies and “stacked” alerts that capture multiple metrics in a single alert.
With our Analysis Workspace feature, you get a robust, flexible canvas for building custom analysis projects. Drag and drop any number of data tables, visualizations, and components (channels, dimensions, metrics, segments, and time granularities) to a project.