Intelligent Recommendations
Intelligent Recommendations use artificial intelligence algorithms to calculate and display items your customer may be interested in buying. The algorithms use data such as:
-
what the customer bought in the past
-
what other customers with similar transaction history bought in the past
-
what other items were bought by customers who bought a specific item
-
items that have similar names, descriptions, and categories
Intelligent item recommendations are useful to two different audiences. Sales representatives can use them to identify and add relevant items to sales orders, estimates, and opportunity records. You can also add recommendations to different areas of your SuiteCommerce website to show shoppers items they may want to buy.
Intelligent item recommendations are generated based on their context:
-
Item context – the recommendations are based on data related to items. There are four types:
-
Order or Cart Items – these are related to the items already selected in the sales order, estimate form, or opportunity record.
When used on a SuiteCommerce website, these recommendations are specific to the items in the shopper’s cart.
-
Customers Who Bought This Also Bought – these are specific to an item selected in the sales order, estimate form, or opportunity record, and are based on the purchase history of other customers who bought the selected item.
These recommendations can also be shown on Product Details Pages (PDP) on SuiteCommerce websites.
-
Recently Viewed Items – these are related to the items the SuiteCommerce website shopper has viewed recently while browsing the website.
Note:This applies only to SuiteCommerce websites.
-
Alternative Items – these are specific to an item selected in the sales order, estimate form, or opportunity record and show recommendations of alternatives to the item based on similarity of item name, description, and category.
These recommendations can also be shown on Product Details Pages (PDP) on SuiteCommerce websites. For example, if a website shopper adds small green shirts to an order, and then realizes there aren't enough shirts in stock, the Alternative Items recommendation might show large green shirts or small red shirts.
-
-
Customer/ Shopper context – the recommendations are based on data related to customers. There are two types:
-
Customer Purchase History – often called ‘You might like’ recommendations, these are specific to each customer and are based on the customer's purchase history and items bought by customers with a similar purchase history.
These recommended items can be shown on sales orders, estimate forms, and opportunity records as well as SuiteCommerce websites.
-
Buy Again – these are specific to each customer and are based on whether items are regularly bought by the customer. If a customer bought an item only one time, recommendations are based on whether the item is bought regularly by other customers.
These recommended items can be shown on sales orders, estimates, and opportunity records, as well as SuiteCommerce websites.
-
If Customer Relationship Management, SuiteCommerce, or SuiteCommerce Advanced is enabled in your account, an Intelligent Recommendations button appears on sales orders, estimates, and opportunity records. The number of available item recommendations is displayed on the top-right of the button. Clicking Intelligent Recommendations displays a list of recommended items that apply to the specific combination of customer and items at that point of time. A different combination of customer and items results in different recommendations. You can use the dropdown to change the basis on which items are recommended, by adding or removing options from the Recommendations Based On list.
The exception to this are Alternative Items and Customers Who Bought Also Bought recommendations, which can be seen by selecting an item in a sales order, estimate, or opportunity, and clicking Intelligent Recommendations.
The recommendations shown depend upon the transaction data available in the account. For example, if there are only a few transactions containing the selected item, there may not be any recommendations based on 'Order or cart items.'
The machine learning algorithms process and learn from the transaction data every 24 hours so the recommendations periodically change and improve. As the number of customers, products, and transactions increase in your account, the quality of the intelligent recommendations also improves.
If you want to use intelligent item recommendations on your website, see Intelligent Item Recommendations for Commerce Websites.
To check which types of recommendations are currently available for a subsidiary or a site, go to Commerce > Marketing > Intelligent Recommendations > Recommendation Availability.