About Product Recommendations (PR)
About Product Recommendations (PR)
This is for:
System AdministratorCoveo Machine Learning (Coveo ML) Product Recommendation (PR) models take advantage of Coveo Analytics to suggest relevant products to end users based on their past and present interactions with your Coveo-powered commerce implementation.
Coveo ML PR enhances the user’s experience by offering products that suit their profile, context, and buying behaviors. To provide relevant suggestions, the model continuously learns from your end users' feedback by scoping their buyer profile and analyzing their interactions with different products. PR offers several recommendation strategies, each suited to a different storefront placement. Seeded strategies recommend products related to the seed products passed in the recommendation query, unseeded strategies recommend products based on general popularity, and user history strategies personalize recommendations from a visitor's browsing and purchase history.
Depending on your context, you can leverage one or more of the available PR strategies.
This article describes the prerequisites needed to create and deploy Coveo ML PR in a Coveo for Commerce interface.
|
A new version of Coveo Machine Learning (Coveo ML) Product Recommendation (PR) models was released in February 2026. This version is currently in open beta, and a migration path is available for Coveo organizations using the previous version. |
How PR compares to other recommendation models
Coveo offers several recommendation models. They differ in what they recommend and which signals they learn from:
-
PR recommends catalog products, learning product relationships from historical interaction patterns aggregated across visitors and sessions. Choose PR for the recommendation placements in a commerce storefront, such as product detail pages, cart pages, and the home page.
-
SBPR also recommends catalog products, but it enriches those historical signals with the visitor’s current shopping session, so recommendations reflect in-the-moment intent. Choose SBPR when intent shifts within a single visit, or when many of your visitors are unauthenticated and have no interaction history to learn from.
-
CR recommends content rather than products, such as articles, blog posts, videos, and community posts. Choose CR for content properties such as a support portal or a community, not a commerce storefront.
Prerequisites
-
Coveo Machine Learning (Coveo ML) Product Recommendation (PR) models use Coveo Analytics events to relevantly target and suggest products to your visitors. Therefore, you must log commerce events to ensure that your commerce interfaces correctly track user interactions. More specifically, you must log the following event types:
-
Product views
-
Purchase events
-
Cart events (add/remove)
-
Click events
Notes-
PR models will work without cart and click events, but you should still log them for accurate reporting and attribution.
-
To serve relevant recommendations, a PR model needs at least 10,000 view and/or purchase events to learn from.
-
-
Have configured a catalog entity and catalog configuration in your Coveo organization.
-
Your catalog data for items of the Product catalog object contains data for the
ec_categoryfield. Otherwise, category-based PR strategies won’t function correctly. We also recommend that you populate the other commerce standard fields to enhance recommendation precision and diversity.
Create the PR model
After you confirm that your ecommerce storefronts track the proper usage analytics events, you can create a PR model.
Associate the PR model with a query pipeline
Once your model has been created, you must associate your model with a query pipeline and select an appropriate strategy.
What’s next?
Once your model is associated with a query pipeline, you can then build recommendation interfaces to query the model and display the recommendations in your commerce storefronts.