Advanced Features

Historical Data Upload

The Strands Recommender system supports uploading historical data in order to weight certain items in your catalog that have historical behavioral data. Use the format below to create a file and then submit the file to support so we can process it.

Parameter List for Historical Data Upload

  • type : the type of event ( purchased, visited, addshoppingcart, etc.. )
  • itemId : the item ID of the given product
  • userId : the user ID related to the event
  • date : date of activity

Types of Events for Historical Data Upload

  • visited : User has visited that item
  • purchased : User has purchased that item
  • addshoppingcart : User has added this item to their cart
  • addwishlist : User has added this item to their wish list
  • addtofavorites : User has added this item to their list of favorites

Sample CSV format for Historical Data Upload

visited,101,,2011/02/15 14:12:00
visited,102,,2011/02/15 15:12:50
purchased,204,,2011/02/18 15:13:50

Email Recommendations

Email Recommendations work very similar to website based ones because they consist of HTML code that you copy and paste from our dashboard and into your email campaign editor.

Styling your Email Recommendations

In order to style you email recommendations, log in to the dashboard and select Recommendations then Mail. Please note, this feature is only available with the Advanced and Enterprise Plans, if you don’t see the link active in your dashboard, contact support. Select a Mail Widget Look from the available list of email recommendations or select Add new recommendations. Similar to the website recommendations, you’ll select the exact look and feel of the recommendations. Click save and you’re ready to integrate the recommendations.

Integrating your Email Recommendations

Under the Integration column header there is a button for Mail. Once you click that link, you’ll get a popup that will help generate the source code you’ll need to integrate the recommendations you want. Each email recommendation has different inputs; for example a recently viewed based template requires the users unique ID and a most popular items this week one doesn’t require any inputs. If you want to experiment with how these email recommendations will render, just copy and paste the HTML into a sample web page and render it on a local browser.