Abstract: A user interacts with item listing while shopping for a gift for an intended recipient. Based on the user interactions, a recommendation engine selects item listings describing items for sale for presentation to the user, the item listings being for items of interest to the intended recipient. The selection is based on interactions by the user with item listings (e.g., past purchases by the user), attributes of the selected item listings, a wish list of the intended recipient, and behavior of other users. A recommendation may also be presented based on a date of an occasion for which an item may be purchased by the user as a gift for the intended recipient.