Einstein Recipe Ingredients

Any eligible item

Considers any eligible item as a recommendation. All built-in requirements still apply, such as products must have an explicit price, but any item can be a recommendation without special conditions. This ingredient doesn’t require an anchor item and follows any recipe exclusions. Weight isn't adjustable for this ingredient because it’s the fallback ingredient for other ingredients in the recipe.

None.

Co-Browse

This ingredient considers items and content that customers browsed with the anchor item at least 10 times during the specified lookback period.

For example, if Article A is the anchor item and customers browse Article B with Article A hundreds of times during the lookback period, Personalization recommends Article B when customers browse Article A.

However, if article C is the anchor item and customers browse Article C with Article A only nine times during the lookback period, Personalization doesn’t recommend Article C when customers browse Article A.

Weight—Select how to weight this ingredient against others in the same recipe.

Lookback—Enter the number of days to look back in the history of a customer’s views and purchases.To look at today only, enter a value of 0. Specify up to a maximum number of n days (the maximum allowed in the Einstein Recipe tool).

Anchor—Select an option to further define the anchor item: On page, In cart, Last purchased, From explicit selection, Recent purchases, Recent views, or Recently spent time viewing.

Min. View Time—Enter the number of seconds that a customer must view an item to establish their interest in it.

Co-Buy

This ingredient considers items and content that customers purchased or downloaded along with the anchor item at least five times during the specified lookback period.

For example, if a red shirt is the anchor item and customers purchase it with red socks hundreds of times during the lookback period, Personalization recommends red socks when customers view or add the red shirt to the cart.

However, if customers purchase the red shirt with a specific pair of jeans only four times during the lookback period, Personalization doesn’t recommend the jeans when customers view or add the red shirt to the cart.

Weight—Select how to weight this ingredient against others in the same recipe.

Lookback—Enter the number of days to look back in the history of a customer’s views and purchases. To look at today only, enter a value of 0. Specify up to a maximum number of n days (the maximum allowed in the Einstein Recipe tool).

Anchor—Select an option to further define the anchor item: On page, In cart, Last purchased, From explicit selection, Recent purchases, Recent views, or Recently spent time viewing.

Require Purchases in Same Basket—Recommends items that customers buy together in a single purchase.

Collaborative filtering

This ingredient considers recommendations based on a customer's affinities during their browsing and purchasing.

Collaborative filtering recommends content and products based on a given customer's cluster, which is defined by their behavior and interests. Basically, a customer sees products or content based on a "people like me" algorithm.

Weight—Select how to weight this ingredient against others in the same recipe.

Lookback—Enter the number of days to look back in the history of a customer’s views and purchases. To look at today only, enter a value of 0. Specify up to a maximum number of n days (the maximum allowed in the Einstein Recipe tool).

Skip serving trending fallback recs for users with no activity—Select this option so that Personalization doesn’t recommend trending items when a customer has no activity.

Most recently published

This ingredient uses the published dates from the catalog to prioritize items you’ve recently added to your site.

Weight—Select how to weight this ingredient against others in the same recipe.

Decay Age—Enter the length of time to consider an item as a new item. You can relate this setting to how often you introduce new products. For example, to promote products introduced in the past 10 days, set the decay age to 240 hours.

Relevancy Retained—Enter a percentage (from 10% to 95%) to indicate how to score items as they reach their decay age. For example, if you enter 80%, items that reach the decay age are 80% less relevant than items that haven’t reached the decay age.

Similar items

This ingredient considers content or items that:

• Are similar in name or description to the anchor item
• Are in the same category as the anchor item
• Have the same tag as the anchor item

Weight—Select how to weight this ingredient against others in the same recipe.

Anchor—Select an option to further define the anchor item: On page, In cart, Last purchased, From explicit selection, Recent purchases, Recent views, or Recently spent time viewing.

Include Results Matching—Select options to limit the recommendations that return: Name, Description, Categories, Tags.

SmartBundle

This ingredient considers items from SmartBundle relationships that customers bought together in the same cart. The ingredient only considers items or content viewed a minimum of 2 times with the anchor item or content for the lookback period.

To use the SmartBundle ingredient, Define a SmartBundle Relationship.

Weight—Select how to weight this ingredient against others in the same recipe.

Lookback—Enter the number of days to look back in the history of a customer’s views and purchases. To look at today only, enter a value of 0. Specify up to a maximum number of n days (the maximum allowed in the Einstein Recipe tool).

Anchor—Select an option to further define the anchor item: On page, In cart, Last purchased, From explicit selection, Recent purchases, Recent views, or Recently spent time viewing.

Soon to expire

This ingredient uses the catalog expiration date to prioritize items with an upcoming expiration.

If a recipe only has this ingredient, Personalization presents the products with the closest expiration dates first.

Weight—Select how to weight this ingredient against others in the same recipe.

Decay Age—Enter the length of time to consider an item as a new item. You can relate this setting to how often you introduce new products. For example, to promote products introduced in the past 10 days, set the decay age to 240 hours.

Relevancy Retained—Enter a percentage (from 10% to 95%) to indicate how to score items as they reach their decay age. For example, if you enter 80%, items that reach the decay age are 80% less relevant than items that haven’t reached the decay age.

Static relationship

This ingredient uses the static relationships you define to recommend other products. After you associate products with each other in a static relationship, Personalization can recommend products according to those relationships. This ingredient is available only if you have Static Relationships enabled for your account, and you’re tracking products on the dataset.

To use the Static Relationships ingredient, Define Static Relationships.

Weight—Select how to weight this ingredient against others in the same recipe.

Anchor—Select an option to further define the anchor item: On page, In cart, Last purchased, From explicit selection, Recent purchases, Recent views, or Recently spent time viewing.

Relationship Type—Select the type of static relationship to use when recommending products related to the anchor item.

Trending

This ingredient recommends products or content trending on your site during the past 24 hours, and orders them according to your selection.

Weight—Select how to weight this ingredient against others in the same recipe.

Order by—Select how you want to order the trending items within the list of recommendations:

• Most revenue generated
• Most time spent viewing
• Most purchased
• Most viewed