The Frequently Viewed Together AI recommendation algorithm predicts products that are frequently viewed together by customers. It is commonly used on e-commerce platforms to recommend products that customers are more likely to purchase together.
How does it work?
The algorithm uses a collaborative filtering technique to analyze customer behavior and identify patterns in their product views and purchases. It then uses these patterns to generate recommendations for frequently viewed products.
Specifically, the algorithm works by first creating a matrix of customer-product interactions, where each row represents a customer and each column represents a product. The matrix is then decomposed into two lower-dimensional matrices representing the customers and products, respectively. These matrices are then used to calculate the similarity between products based on their co-occurrence in customer views and purchases.
Finally, the algorithm generates recommendations by identifying products that are frequently viewed together by customers and have a high similarity score.
Supported Rule Types
- Global
- Home Page
- Product
- Cart
Recommended Placements
While this widget can be added to any page, here are the most recommended placements for it:
- Product Page: FVT for this product
- Cart: FVT for products in the cart
- Checkout: FVT for products being purchased
- Home: Customer Also Bought (generic)
Use Cases of Merchandising Rules in FVT
- Pin products to promote them
- Exclude products
Behavior for Non-Logged-in User
FVT: Behavior for Home Page, Category, Product, Search, Cart, Checkout or Pages
- If the user has done some interactions in the session, the system will provide recommendations based on products viewed within the session for the FVT widget based on the context awareness logic.
- If the user is brand new and has not browsed any products,
- If “FALLBACK” is not enabled, hide the widget when users are not logged in.
- If the user is on the product page, it will show the suggestions that are made to get started with the current product.
- If the user is on the home page, it will not show any suggestions because there is no browsing history.
- If the “FALLBACK” algorithm is enabled, it will show a widget with fallback algorithm products.
- If the user is on the product page, it will show the suggestions that are made to get started with the current product, and it might use the FALLBACK button to fill in the empty spots.
- If the user is on the home page, it will not have any suggestions because there is no browsing history, so it will show the products from FALLBACK only to help with the cold-start problem.
Behavior for Logged-in User
FVT: Behavior for Home Page, Category, Product, Search, Cart, Checkout or Pages
- If the user has done some interactions in the session, the system will provide recommendations based on products viewed within the session for the FVT widget based on the context awareness logic.
- If the user is brand new and has not browsed any products,
- If “FALLBACK” is not enabled, hide the widget when users are not logged in.
- If the user is on the product page, it will show the suggestions that are made to get started with the current product.
- If the user is on the home page, it will not show any suggestions because there is no browsing history.
- If the “FALLBACK” algorithm is enabled, it will show a widget with fallback algorithm products.
- If the user is on the product page, it will show the suggestions that are made to get started with the current product, and it might use the FALLBACK button to fill in the empty spots.
- If the user is on the home page, it will not have any suggestions because there is no browsing history, so it will show the products from FALLBACK only to help with the cold-start problem.
When should you use this algorithm?
The Frequently Viewed Together AI recommendation algorithm is particularly useful for e-commerce platforms that want to increase sales and customer engagement by providing personalized recommendations. It can be used in Experro to provide recommendations across multiple channels, including mobile apps, websites, and email.
Example
Here is an example of how the Frequently Viewed Together algorithm could be used in Experro:
- A shopper visits an e-commerce website and views a product page for a designer dress.
- Experro uses the Frequently Viewed Together algorithm that recommends a matching dress that other customers have frequently purchased with those dresses.
- This would increase the likelihood of the customer making additional purchases and improve their overall shopping experience.
- It displays a list of Frequently Viewed Together products for the shopper on the product page.
To create or learn more about widgets, go to our Widget Configurator article and get started.