Recommendation Scenarios and Blocklists
Recommendation scenarios are a set of configuration guidelines that the machine learning algorithm follows while calculating which items to recommend. The guidelines included in a recommendation scenario are:
-
Recommendations based on – whether the recommendations should be based on:
-
Customer purchase history
-
Order or cart items
-
Recently viewed items
-
Customers who bought this also bought
-
Buy again
-
Alternative items
-
-
Freshness – this defines the degree to which the recommendations should prioritize items recently added to the catalog.
-
Novelty – this defines the degree to which the recommendations should prioritize items that customers rarely see.
-
Diversity – this defines the degree to which the recommendations should prioritize items that are different from each other.
NetSuite creates default recommendation scenarios based on the following and automatically sets them to provide recommendations for sales orders, estimates, and opportunities:
-
Customer purchase history
-
Order or cart items
-
Customers who bought this also bought
-
Buy again
-
Alternative items to the selected item
NetSuite creates up to six default scenarios based on "Recently viewed items". These scenarios are available for use on SuiteCommerce websites. The default scenarios are suited to most businesses but you can modify these scenarios or create additional scenarios with custom intelligent recommendations settings for each website. For more information about using intelligent item recommendations on your website, see Intelligent Item Recommendations for Websites.
Blocklists
Blocklists are lists of items or item collections that you do not want to have included in intelligent item recommendations. If you link a blocklist to a recommendation scenario, the items in the blocklist are not displayed as part of that recommendation scenario.
You can link a specific blocklist to more than one recommendation scenario, and you can link multiple blocklists to a single recommendation scenario.
Always Block vs. Block for Item
There are two ways to block items in blocklists:
-
Always Block – items and item collections marked as Always Block never appear as intelligent recommendations.
-
Block for Item – these items are blocked from appearing in intelligent recommendations only when a specific item has been added to the sales order, estimate, or opportunity record. When the blocklist is used in a website, these items are blocked from recommendations only when the shopper is looking at the specific item. In both cases, the specific item is called a context item.
For example, you may not want an older model of equipment to be recommended when the sales order or estimate already contains a newer model, or when the website shopper is looking at a newer model. You can choose the newer model as the context item and the older model as the blocked item.
Block for Item also applies to item collections. For example, if you sell books and music, you may want to block all music recommendations when the sales order, estimate, or opportunity record contains only books. On your website, you may want to prevent the shopper being shown music recommendations when they are looking at books. In these cases, you can select the Books item collection as the context item collection and the Music item collection as the Blocked item collection.
For information about creating blocklists and recommendation scenarios, see Create a Blocklist and Recommendation Scenarios and Blocklists.