The next phase in setting up AI Personalization is configuring the events. This teaches AI Personalization how to interpret user interactions to provide relevant personalized search results. AI Personalization learns from events you setup in your configuration. It also automatically considers events from the replicas of your personalized index. When the configuration is saved, events from the last 24 hours are validated. The dashboard will display errors if invalid user tokens or invalid objectIDs are found in the selected events. To start configuration, go to the User Events section of the AI Personalization dashboard. Here, you’ll find four different event collections to configure.

​
Event collections

An event collection is a group of events that describe a user signal. It’s used by AI Personalization to make sense of user interactions and fine-tune personalization. The event collections are:
  • A user viewed a product in a listing
  • A user viewed a record’s details
  • A user showed interest in a record
  • A user achieved their goal with a record
The dashboard displays only the events received in the last 24 hours.

​
A user viewed a product in a listing

Events related to a user viewing products on listing pages. The collection includes events for:
  • Products viewed on the home page.
  • Products viewed in a category listing.
  • Products viewed in search results.
This event collection is optional. You can configure AI Personalization without it.

​
Examples

Products viewed in a category listing:
JSON
{
  "eventType": "view",
  "eventName": "Category Page Viewed",
  "index": "prod_ecommerce_EN"
}
Products viewed in search results:
JSON
{
  "eventType": "view",
  "eventName": "Search Results Viewed",
  "index": "prod_ecommerce_EN"
}

​
A user viewed a record’s details

Events related to a user viewing the details of your record. This collection includes events for:
  • A product viewed from its PDP (Product Detail Page). This view event captures the signal that a user landed on a product detail page, coming either from another page of this website or from an external referrer (such as Google or social media).
  • (Optional) A product clicked on a PLP. This click event captures the signal that a user is navigating to a PDP.

​
Examples

A product viewed from its PDP:
JSON
{
  "eventType": "view",
  "eventName": "Product Details Viewed",
  "index": "prod_ecommerce_EN"
}
A product clicked on a PLP:
JSON
{
  "eventType": "click",
  "eventName": "Product Clicked",
  "index": "prod_ecommerce_EN"
}

​
A user showed interest in a record

Events related to a user indicating interest in your record. This collection includes events for:
  • Products added to cart. A conversion event that captures the signal that a user added one or multiple products to their cart from any page (such as search results, category pages, or product detail pages).

​
A user achieved their goal with a record

Events related to a user successfully achieving their goal using your record. This collection includes events for:
  • Products purchased. A conversion event that captures the signal that a user has completed the checkout process.

​
Next step

Once you’ve set up the event collections, you’re ready to set personalization re-ranking.

​
Further reading

Guide

Sending events with user token

Guide

AI Personalization error codes

REST API

Update the configuration with the AI Personalization REST API