Product feeds 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 only for these integrations:
- Magento 2
NoteFor Shopify: If you are using Klaviyo to customize your store's default transactional email templates, you won't be able to export one of these transactional templates from Klaviyo if it contains a product block. This is because these blocks only work for emails sent by Klaviyo.
Create a Product Feed
Go to the Data Feeds tab in Klaviyo, and click on the blue button that says Add Product Feed. If you have one of our built-in ecommerce integration enabled, popular and trending product feeds are automatically created for you.
After you have given your new feed a name, you can choose the kind of weighting you'd like for your feed. There are three options: Unweighted, Popular, and Trending.
You can also configure a feed such that it is limited to (and will exclude) certain categories:
- Include: Item must be in at least one listed category to show up in the feed
- Exclude: Item must be in none of the listed categories to show up in the feed
When there is a conflict, exclude takes priority over include, as we err on the side of not showing something that you don't want customers to see.
How Product Feeds Work
There are two components when it comes to how we sort and present product recommendations: a personal recommendation engine and a weighting system.
Any products that do not have images associated with them will get filtered out automatically. For each email recipient, we will automatically also filter out products the recipient has purchased in the past. If products are out of stock, Klaviyo will consider them "unpublished" in your catalog and not show these products in your feeds.
Popularity is calculated by dividing the number of times a specific product was bought by the total number of products sold over the period of last 90 days. This means a new surge in a single product may not impact a product feed immediately.
Personal Recommendation Engine
Based on past purchasing data, we look at how similar the email recipient is compared to the other users in the dataset. Then we come up with a recommendation score for each product and sort the products based on this score.
Personalized recommendations are based entirely on past purchase behavior. For example:
- User A orders Product 1, Product 2, and Product 3
- User B orders Product 1 and Product 2
In this example, recommendations for User B will show Product 3.
If a user has never placed an order, they will see products based on the general setting of the feed rather than a personalized recommendation. Whenever there isn't enough data available to generate customized recommendations, the general settings of the feed will prevail.
The Weighting System
There is no sorting priority. If you have the personal recommendation engine turned off or if you are a free account user, this option means a randomized product mix. We recommend this option if you have a sizable amount of products and a lot of transactional activity. If you have very few products and/or have little transactional activity, you will want to use a different weighting system.
- Popular Products First
Popularity is calculated by dividing the number of times a specific product was bought by the total number of products sold over the period of last 90 days.
- Trending and recently popular products first
Similar to how we calculate popular products, but we give more weight to products that are bought frequently recently. This is also calculated over the period of 90 days. It's possible that if you have two products that are equally popular, but because the purchases for product A surges earlier than product B over the last 90 days, product B will be shown as trending higher.
After we have these two scores, we then assign different weights to them and come up with a composite score. This score determines the final order of products shown in product feeds.
Items are sorted by (recommendation engine score * 80%) + (weighting * 20%).
This is great for if you want to send out the campaign to people who've bought your products and people who just signed up for your mailing list. We will personalize the feeds for those who purchased previously and show the most popular items to those who have not yet bought anything.
Using a Product Block
The Product Block in Klaviyo provides two ways for users to populate buyable items from their product catalog into an email template:
- Populate from Feed: Klaviyo will populate products into your template dynamically at send time, where items are curated from a pre-defined product feed.
- Select from Catalog: Want to choose your own items to feature? We make it easy to browse your entire product catalog and insert up to 9 items with a few clicks.
When you add a product block, and choose the "populate from feed" option, you will have the option to enable the personal recommendation engine and send more targeted emails to your recipients. If you have personal recommendation option turned off, product recommendation will depend completely on the weighting system of the feed. If you have the personal recommendation option turned on, a recommendation is determined by both the more advanced personal recommendation algorithm and the weighting system.
If you are using a feed, and you want each product, when clicked, to redirect to a URL that is different from your integrated store's URL, you will have to manage this from within your ecommerce platform. For Shopify users, you can navigate to Shopify’s Domains panel and force all domains to be redirected to your alternative domain. While links from Klaviyo will still start with the root domain, recipients will be re-directed in real time to the same product pages but with your alternative domain.