On this page:

See also:


Overview

Product recommendations predict and display products your contacts might be interested in purchasing. 

Ometria has 7 recommendation engines for use in your campaigns. 

Be aware that some recommendation engines are better suited to automation campaigns and others are better for broadcast campaigns. 

Recommendations can be restricted by category or attribute, or filtered by price range, and you can blacklist certain products or even remove products your contact has already purchased.

Go to: Campaigns > Product Recommendations

The Product Recommendations list displays:

View your active or archived campaigns by selecting from the dropdown list on the right:

If you select ‘SHOW ARCHIVED’, you have the option to UNARCHIVE old product recommendations:

...and if you select SHOW ACTIVE, you have to option to EDIT or ‘Archive’ your product recommendations:

Both lists display the following:

HeadingDescription
TitleThe name of the recommendation (as defined by the creator).
EngineThe type of recommendation engine used.
Date createdThe date the product recommendation was created.
SettingsEdit, archive or unarchive your recommendation. 

You can also click the CREATE NEW button to begin working on a new product recommendation. 


Note: To edit an existing product recommendation, see Create New Product Recommendation, as the steps are similar.


Product recommendation criteria

Not all products are available for display in product recommendations. 

This section lists the criteria for a product to be used in a product recommendation instance.

Key criteria

  • Ometria only pulls parent products in recommendations not variants. 

Note: A ‘parent’ product, also known as a top-level product, prime product or configurable product (in Magento).


  • The product must have been sold in the last 28 days. We don’t recommend products without sales, as they are unlikely to convert. The exceptions to this rule are:
    • Recently added products - this only looks for the newest products.
    • Personalised products - in this case, if no sales are found for that set of attributes the recommendation falls back to products most recently added to your website.
  • The product must be active, with the following details defined at parent level:
    • Title
    • Price
    • URL
  • The ‘is in stock’ checkbox must be selected at parent level (in the back end).

Product data requirements for ecommerce importers

To recommend the right products to the right contacts, the products passed from your ecommerce platform to Ometria must contain the following fields with valid values (i.e. no null values).

Field Notes
attributeAttributes to include.
attribute_blacklistingAttributes to exclude.
attribute_type
blacklist_productsProduct IDs to exclude.
bought_productsProduct IDs of products that were purchased.
context_productsProduct IDs from triggering event - sometimes used, depending on the engine.
priceMandatory.
storeAll products must be associated with a store, usually from the triggering event. 

Additionally, if a product is out of stock and you don’t want to include it in any recommendations, both the global and store level ‘is_in_stock’ fields in the Product API should be set to ‘False’.

Out of stock products

Product recommendations generate when your contact opens the email, which reduces the likelihood of contacts receiving recommendations for out of stock products. 

However, the recommendations are calculated on a nightly basis, and this calculation includes whether or not the product is ‘out of stock’. 

This means that if a product sells out during the day, e.g. at 2:00pm in the afternoon, and your contact opens their email at 5:00pm, they may still see out of stock products.

Product recommendation rules

Provided that the key criteria and product data requirements listed above are met, Ometria generates product recommendations based on the following rules.

If...then...
no store ID parameter is passed inOmetria uses product details based on the top level product record, e.g. title, URL, active status, price, currency, image, etc.
a store ID parameter is passed inThe store ID is converted to a 'product listing'.
  • If the listing is found, the title, URL, active status, price, currency, image etc. are merged in from the listing into the record before it is displayed.
  • After the merge, if the active status is false (i.e. inactive), the product is removed from the returned products and not displayed

Store ID

Which store ID is passed in depends on context.

  • Product recommendations - it depends on the  selected store.
  • Automation campaigns - emails corresponding to an event that occurred in a store (e.g, order, abandoned basket), the store ID is the store that triggered the event.
  • All other sends - the ‘default store’ for the contact (if set).

If none of these are set, the store ID falls back to the default defined in Settings > Stores.

Product listings and product variants

Product listings are overridden product fields for individual stores. These listings allow us to generate localised product information.  

For example, a product might cost £45 in the UK store, and €59 in the French store, or it may also have a different URL, image, title, etc. 

Product listings only apply to parent products.

Product variants are for shippable SKUs, e.g. a shoe available in 8 sizes would have 1 parent product and 8 variants.

Edge cases

  • If ‘is_active’ is set to FALSE on the parent product, but TRUE on the store listing, then the product will not display.

  • If the store ID is passed in where there is no product listing, Ometria uses the default product details, and the product will display. This is because some integrations send listings and others don’t.