AI Recommendations
AI Recommendations is a feature that enables brands to send dynamic, personalized product recommendations to end users. Product recommendations make marketing campaigns more effective by increasing the clicks on the content, which increases the chance of purchase, downloads, and other product-related actions by end users.
These recommendations are specific to each end user and are based on that end user's interaction with products on the brand website, mobile app, and other services. Example: purchase, add to cart, add to favorites, add a review, or add a rating. The recommendations are sent through available channels on Infobip solutions.
The following images show a product recommendation email over Moments
Unlike static recommendations that are generated at the time the message is sent, Infobip's AI recommendations are generated at the time the end user opens the message. Each time the end user opens the message, the recommendations change based on the end user's interaction with your products.
Example: If the end user opens the email containing recommendations, makes a purchase on the product website, and opens the email again, the recommendations would change based on the recent purchase. Also, these recommendations would not include items that the end user already purchased.
Features
User-specific Recommendations
Recommendations are tailored to each end user based on that person's interaction with the product on the product website, mobile app, and other services.
Dynamic Recommendations
Recommendations are generated each time the end user opens the message.
Customized Messages
Design the message, choose horizontal or vertical layout, determine the number of recommendations, number of rows, and number of recommendations in a row.
Available Solutions
Technical Specifications
The following image shows an overview of the process:
Cold Start
The more information that is available about end users' interaction with your products, the better the product recommendations. However, if you do not have any tracking information, you can still send product recommendations. In such cases, Infobip uses content-based models (text and image processing, and other metadata) to generate recommendations. When there is at least one product-related activity by the end user, Infobip uses personalized models to generate recommendations.
Before Getting Started
You must have an Infobip account to start using AI Recommendations. Log on or sign up to an account.
You must also have one or more of the required Infobip solutions enabled.