Evergage has introduced Evergage Decisions, a module designed to enable retailers to apply AI for automatically determining and delivering the optimal promotion or offer to each web site visitor, app user and email recipient.
Contextual Bandit, the first algorithm introduced as part of the module, is designed to complement Evergage’s existing machine learning-driven, one-to-one personalization and recommendation capabilities.
The algorithm factors in situational and attribute criteria (e.g. referral source, browser, device type, lifetime value, geolocation, etc.) and deep customer behavioral data. Using this information, Contextual Bandit can:
- Estimate the probability of each shopper interacting with each available offer or experience on a given channel (web site, web app, mobile app, email) in real time;
- Use advanced machine learning to predict the content for each visitor with the highest-value return — weighing the probability of someone engaging with a particular offer or promotion as well as the business value of that offer to the company; and
- Free up retail marketers to focus on creating powerful messaging and offers, instead of spending time defining rules about which experience to show which audience.