A Product Recommendations for Email experience relies on the same recommendation algorithms and underlying integrations as a traditional Web-based recommendations experience, which means you can use many of the same algorithms, lookup keys, and filters to build recommendation strategies that you can use across channels.
Due to fundamental differences among channels, you must keep in mind certain considerations to account for how marketers use Monetate to power recommendations. Specifically, you cannot use any recommendation strategy that leverages current session logic in a Product Recommendations for Email experience. These strategies don't appear as options when you're building the experience.
When building a new recommendation strategy specifically for Product Recommendations for Email experiences, you must ensure the following:
- If you select collaborative recommendation algorithm, then you must select a From session scope option that's based on past sessions. You cannot select a session scope option based on the current session.
- If you create a recommendation strategy that uses a Recommendations dataset, you must set Base Recommendations on Items to either No Lookup Key — Show All Items in Dataset or Item group ID(s) in run-time parameter (for email). Otherwise, the strategy doesn't appear as an option when you configure the WHAT action in a Product Recommendations for Email experience. Even if the Recommendations dataset that you select has
none
as the value for thelookup_key
attribute for all products in the dataset, you still must select No Lookup Key — Show All Items in Dataset. - If you build a recommendation filter in the strategy that uses a dynamic value and choose not to use an option that references a value passed at run time, then you must select a dynamic value option that relies on previous session behavior (for example, ID of a Previously Purchased Product) and not one that uses current session behavior (for example, ID of a Product Currently Being Viewed).