A computer-implemented method, comprising:
- receiving, by a processor, information related to a visitor and a current activity of the visitor on an online enterprise channel in substantially real-time;
- transforming the received information, by the processor, to generate transformed data corresponding to the visitor and the current activity of the visitor on the online enterprise channel;
- extracting, by the processor, a plurality of features from the transformed data;
- using the plurality of features, determining by the processor, whether a treatment from among a plurality of treatments when rendered to the visitor is capable of increasing a likelihood of the visitor performing a desired action during a current visit of the visitor to the online enterprise channel;
- selecting, by the processor, the treatment for the visitor when it is determined that the treatment is capable of increasing the likelihood of the visitor performing the desired action during the current visit;
- rendering the selected treatment to the visitor during the current visit of the visitor to the online enterprise channel;
- rendering no treatment to the visitor during the current visit when it is determined that no treatment from among the plurality of treatments is capable of increasing the likelihood of the visitor performing the desired action when;
- wherein the treatment corresponds to displaying a banner advertisement to the visitor during the current visit of the visitor to the online enterprise channel, wherein the banner advertisement is to be displayed on the online enterprise channel for a first time;
- extracting, by the processor, one or more features from the banner advertisement;
- adding the one or more features to the plurality of features extracted from the transformed data to configure a set of features;
- generating, by the processor, a first probability score based, at least in part, on the set of features, the first probability score indicative of a probability of the visitor performing the desired action when the banner advertisement is displayed to the visitor during the current visit on the online enterprise channel;
- generating, by the processor, a second probability score based, at least in part, on the set of features, the second probability score indicative of a probability of the visitor performing the desired action when the banner advertisement is not displayed to the visitor during the current visit on the online enterprise channel;
- comparing, by the processor, a difference between the first probability score and the second probability score with a predefined threshold value derived at least in part based on a default banner advertisement;
- determining that displaying the banner advertisement is capable of increasing the likelihood of the visitor performing the desired action on the online enterprise channel when the difference is greater than the predefined threshold value;
- and determining that displaying the banner advertisement is incapable of increasing the likelihood of the visitor performing the desired action on the online enterprise channel when the difference is less than or equal to the predefined threshold value.