Product feeds are a type of data feed that can be used to generate product recommendations in an email. A product feed will pull product information from your ecommerce integration and display this information in an email template. This feature can be used in flow or campaign emails via the drag-and-drop product block.
Product feeds are available by default for these integrations:
- Magento 2
- Salesforce Commerce Cloud
If you are using a custom integration, or your integration does not appear on this list, contact our customer support team and request to enable the feature.
Create a Product Feed
Go to the Data Feeds tab in Klaviyo, and click on Add Product Feed. If you have one of our built-in ecommerce integrations enabled, popular, and trending product feeds are automatically created for you.
To start, give your feed a descriptive name. If you have more than one catalog in your Klaviyo account, next select the catalog you intend to use. If you only have one catalog, you will not see this option.
Next, you can choose to filter the items in the feed by categories that you may have. To include categories in your feed, check Limit to specific categories, and select the categories you want to include. Likewise, to exclude categories, select Exclude specific categories, and choose the categories from the dropdown that you want to exclude. If there is an item that is in both a category that is included and excluded, the product will not show to customers.
Next, configure the product ordering for your feed. When there is not enough data on a customer to give a personalized recommendation, the ordering selected will determine what the customer sees. There are three options for you to choose from:
Products in the feed are randomly selected at recommendation time.
- Popular products first
Products are ordered in descending popularity — those that are purchased most frequently in the past 90 days are recommended first.
- Trending products first
Products are ordered not only by the frequency but also by the recency of their purchases. Under this option, a product purchased many times today will be ranked higher than a product that has not been purchased frequently in the last week.
Items without associated images, items have already been purchased by the recipient, as well as out of stock items will not appear in the recommendations.
Next, select the metric you'd like to use to determine an item's popularity and generate personalized recommendations. The variety of product recommendation types available creates an opportunity for you to experiment and learn what works best for your specific use case. Select one of the following options from the dropdown menu:
- Ordered Product
Ordered products and purchases send a strong signal about user preferences and likeliness to buy a particular product. Recommendations will be based on previous purchases and what similar customers have purchased. This is a good metric to choose if you have customer profiles with deep purchase history.
- Viewed Product
Viewed products are another good indicator of the types of products customers are likely to buy. New and smaller businesses can benefit greatly from viewed product-based product feeds when your customer base does not yet have purchase history. However, businesses of all sizes can benefit from leveraging view product data to create recommendations that cover a broader customer base, such as new customers with zero purchases and window shoppers, in addition to a larger variety of products, including newly-cataloged items and higher-priced items.
Blended recommendations take advantage of both the high personalization signal from product orders and the voluminous nature of product views, providing a simple solution to create targeted and personalized recommendations for a broad customer base.
Personalized Recommendation Engine
Klaviyo’s personalization engine tailors product recommendations to each email recipient based on their past behavior, depending on the metric you choose.
If you select Ordered Product: Klaviyo can infer each recipient’s preferences, and recommend products that are purchased by other customers who have similar tastes, styles, or needs based on your customers’ past order data. For example, if we know that a customer bought a green shirt, a green hat, and green shoes, and another customer buys a green shirt and hat, Klaviyo will recommend green shoes.
In order to use the ordered product personalized recommendation engine, a customer must have sufficient purchasing data. If they have never made a purchase before, they will receive recommendations based on the ordering you select.
If you select Viewed Product: Klaviyo will generate recommendations based on the products that customers have browsed in your store. Customers without existing purchase data can receive recommendations of products similar, but not identical, to the ones they have already viewed. With this metric, you will not need to wait for products newly added to your catalog to start gathering purchases before they can appear in product recommendations.
An existing profile is required to make personalized recommendations based on viewed products. It is not currently possible to use anonymous tracking for product recommendations.
If you select Blended: Klaviyo will generate recommendations based on a combination of Ordered Product and Viewed Product metrics. Blended recommendations consider both the customer's prior purchase history and the history of products they like to view.
Using a Product Block
After you’ve created a product feed, you can use your feed and the recommendation engine in your emails. Drag and drop in a product block to get started. For more information on product blocks, head to our article on Inserting a Product Block.