You can see predictive recommendations on the following pages:
Recommendations are sorted based on the combination of two criteria - probability and potential impact on conversion rates, of which the prior has bigger priority. Sorting gives you a meaningful segmentation of your visitors based on behavioral data. In addition, interactions are grouped based on their impact and attribution, ranked by certainty of conversion and historical data of conversion success.
Probability assessments are highly accurate since they are based on elaborate calculating mechanisms, testing, and machine-learning algorithms for predictive analytics and prediction results. Probability estimates the certainty that a recommendation will result in more conversions. The recommendations on top are classified to have the highest probability, or certainty, to drive more conversions.
NOTE:Recommendations with low probability to result in conversions are hidden by default.
Impact is calculated based on historical data for the conversion window and recognized behavioral patterns, or, segments. The recommendation that is sorted higher in the list is basically what has proven to be the most typical, shortest, and most successful path to conversion for this conversion definition so far. It shows the next best experience that would lead to the highest potential increase in visitor conversions and conversion rate, calculated for the next 30 days.
NOTE: This impact sorting tells you if all things equal, how many more visitors would convert if all of them complete the “next best experience”, based on historical data only.
Back To Top
Copyright © 2019 Progress Software Corporation and/or its subsidiaries or affiliates.
All Rights Reserved.