Recommendations generate big revenue among the market leaders such as Amazon, Netflix, and eBay.
Product recommendations by Bloomreach Engagement could be used in various placements:
- Homepage: To show your online store sells awesome products (or to reactivate, personalize, etc.).
- Category detail: To show the most popular items within the category that are currently viewed by a user.
- Product detail: To show alternatives for a currently viewed product by a user.
- Basket: To suggest other relevant products to those already in the basket.
- Email campaigns: To personalize email marketing strategies or just to show which products your online store sells.
- Web layers, push notifications...
Bloomreach Engagement Recommendations use customer and catalog data to personalize item suggestions for each shopper, tailored to the situation. Thanks to artificial intelligence, these recommendations improve the user experience and boost key performance indicators (KPIs).
Watch this introductory video about Product recommendations:
With Bloomreach Engagement, you can easily access AI benefits through ready-made recommendation templates for various uses with just a few clicks. These are customizable — not a mystery box — so you can mix engines to meet your exact needs. Plus, everything updates in real time across the entire platform.
To create a new RS in Bloomreach Engagement, you first need to build an engine using predefined templates in Bloomreach Engagement. Go to
Campaigns > Recommendations > + New recommendation and pick a suitable template for your Use Case. Now you can create a new recommender system.
Each template is discussed in the next chapters with examples of ideal setup. You can choose from the following list:
Two distinctions between various RS are recognized:
- Rule-based mode - A family of methods based on some simple heuristics (e.g., purely statistical approaches). This computation is very quick and the model is ready to be used immediately.
Example: top-selling products, most viewed products...
- AI model - Computed by a machine learning algorithm based on the historical data (the process of learning is often called “training”). Each training of an AI model takes several hours.
Example: TF-IDF, Matrix factorization...
- Recommendations based on their past events.
- Recommendations based on the context of the customer's last actions regarding the product they are currently viewing. There was not enough time to load these actions as events, so they are taken into account by the recommendation through Recommendations for customer API call. An example of such actions is viewing a product or adding it to the cart.
Example: Recommend alternative products for a product currently viewed on the product detail page.
To create recommendations, go to
Data & Assets > Catalogs, where you first import the catalog and then make sure:
- To drag the
item_idlabel to the column that indicates a unique product ID and matches the same ID as defined in the events, e.g.,
- To index (mark as searchable) all of the columns that will be used in the recommendations catalog filter, e.g., active, category (note that you cannot change indexing after you import the catalog).
- To set labels for columns image, URL, title, and price to see recommendations results in the single customer view and test tab of each recommender engine.
Most recommendation engines require event tracking to understand and make use of customer behavior.
Make sure that event tracking is set up correctly, especially when it comes to:
view_itemwhich tracks which items customers view,
cart_updatewhich tracks which items are added or removed from the cart,
purchase_itemwhich tracks which items were purchased.
Data Tracking Reference
For more information on event tracking visit our Data Manager article.
For all the recommender systems to work correctly, go to
Data & Assets > Data manager > Mapping and fill in the following events:
- Purchase item
- Add to cart
- View item
- View category
Updated 22 days ago