How to use product feeds and recommendations

Last updated at:

You will learn

Learn how to create a product feed in Klaviyo in order to use product recommendations in your messaging.

Product feeds can be used to generate product recommendations in an email. A product feed will pull product information from your ecommerce store and display this information in an email template. Once you create a product feed, you can use it in a product block in your email template in order for product recommendations to appear in emails.

Between September 15 and September 22, 2022, Klaviyo released an improved product feeds experience to customers. Pricing and inventory filters are now available, and you can now configure the ordering of products based on a customer’s view history, abandoned cart history, and your newest products. 

It’s important to note that this release is backwards compatible. If you have existing feeds, they will continue to function exactly as before. If you open them in the new editor, your preexisting feed setup will be mapped to the new settings, and you will be able to edit it using the new settings. 

Before you begin

Product feeds are available by default for the following ecommerce integrations:

If you use an ecommerce platform that is not listed above, you'll need to sync your product catalog into Klaviyo by following our guide, Sync a custom catalog feed to Klaviyo. Note that price and inventory filters are not available for custom catalogs. 

Note that the following will not appear in product recommendations:

  • Items without associated images
  • Items that have already been purchased by the recipient
  • Out of stock items
  • Items in the flow trigger event for a given email

Visitors must have existing profiles to receive product recommendations, as it is not currently possible to use anonymous tracking for product recommendations.

Recommendations based on viewed products will not be available if you have yet to set up Viewed Product tracking. Learn how to set up Viewed Product tracking for your ecommerce store

The option to base recommendations on “Products customer has added to cart” is informed by either the Added to Cart or Checkout Started metric in your account. If you have not set up Added to Cart tracking (which is tracked automatically for most ecommerce platforms, but must be manually enabled for Shopify and BigCommerce) those recommendations will only rely on Checkout Started.

Create a product feed

Learn how to create a product feed. You can return and edit your feed at any time. 

  1. In Klaviyo, select Content in the left hand menu.
  2. In the dropdown, click Products
  3. In the upper right, click Manage Product Feeds, then click Create Product Feed
  4. If you have more than one catalog in your Klaviyo account, select the catalog you intend to use. If you only have one catalog, you will not see this option. 
  5. Under What products should the customers view first?, select an option from the dropdown. You can choose from both Classic and Personalized for each customer options.
    1. Depending on what option you select, you may be prompted to choose an additional filter. For instance, Product customers may also like can be informed by a number of metrics, such as a customer’s view or purchase history. For some other options (such as Best-selling products) you can choose the timeframe to be over the last three days, or over the last 90 days. 
    2. If you selected an option under Personalized for each customer, you’ll be prompted with an additional question: If the customer has limited history, what should they view instead? Choose an option from the dropdown for this default. Note that these defaults are across all customers and are not personalized. 
  6. Under What additional filter would you like to apply?, you can choose optional additional filters. 
    1. You can filter based on category includes or category excludes. When selecting category includes or category excludes, you can select one or multiple categories for the input. When multiple categories are selected in a single row, they are combined using OR logic.
    2. You can filter based on stock level and set a minimum and maximum. For both stock level and price you can leave either the minimum or the maximum blank (indicating a minimum of 0 or no maximum, respectively) but you must set either the minimum or maximum. 
    3. You can filter based on price and set a minimum and maximum
    4. You can add multiple filters by clicking Add filter and making another selection. 
  7. When you’re finished configuring your feed, click Save Product Feed
  8. You will be prompted to name your feed. Then, click Save Product Feed


You’ve successfully created a product feed, which you can now use in a product block to provide personalized recommendations in your emails. 

Use product blocks in emails

After you’ve created a product feed, you can use the feed in your emails. Drag and drop a product block to get started. For more information on product blocks, head to our article on how to insert a product block.



Is anything going away? Will any of my existing product feeds change in any way?

No. The new system is 100% backwards compatible with old feed configurations. If you have existing feeds, they will continue to function exactly as before. If you open them in the new editor, your preexisting feed setup will be mapped to the new settings, and you will be able to edit it using the new settings. 

What happened to the personalization toggle on the product block in the template editor?

It’s gone! You don’t need it anymore - personalization can now be configured within the product feed itself, and it’s baked into the sort order you select, rather than being a toggle. 

What integrations are supported for the price and inventory filters?

We support price and inventory filters for the following integrations:

  • Shopify

  • BigCommerce

  • Magento 2

Note that price and inventory filters are not currently available for custom catalogs. 

How does the “Newest products” recommender work? What date does it use to sort products?

We use a “created_at” timestamp that is specific to Klaviyo. This means that when you first sync your catalog, every product in the catalog will have the same “created_at” value, so the newest products recommender will not be useful at first. Once you’ve added some new items to your catalog after the initial sync, you should start to see the expected behavior.

This also means that items will not be prioritized by the recommender just because they’ve been “updated” or have recently come back in stock. This recommender will only prioritize net new products added to your catalog.

I applied a price or inventory filter, but I see an item in the product block that is outside the range I specified - why is this?

Filters apply to variants, but product blocks render items. If an item has even 1 variant that matches the filter criteria, it is eligible to show within a product block. Items have default images and prices, so it’s possible an item will show an image and price of a variant that doesn’t exactly match the price or inventory filter. We aim to update this behavior in the future, such that blocks will only show variants (image and price) that match the filter criteria exactly.

What currency do the price filters use?

It depends on how the item was configured in your ecommerce platform. Klaviyo will use whatever currency value the catalog uses.

Where are web feeds?

If you are looking to set up a web feed (instead of a product feed), click your account name in the lower left corner, then click Settings, then select Web Feeds. From here, you can add a new web feed. 

What happened to “trending?”

The “trending” option is now expressed as a look back window of “3 days.” Any option that uses “over the past 3 days” will be a version of the old “Trending” order. 

I’m using some custom statistics to power product recommendations. Will these be compatible with the new experience?

Yes. For details, please refer to our developer portal article, Sync a custom catalog feed to Klaviyo.

How does the “Recently viewed” recommender work?

We use a model that weighs both recency and frequency of views to order the products. This means that if a customer viewed a product 5 times yesterday, and a different product once today, we may still prioritize the product they viewed 5 times because we take view frequency into account.

Additional resources

Was this article helpful?
444 out of 767 found this helpful