Integrating with the Raptor recommendations engine

Raptor is a data driven personalization service offering intelligent recommendations, based on patterns and trends in visitor and buying behavior across your web shop or website.

Integrating with Raptor involves:

  • Building the RaptorRecommendation project and copying the dll to the bin folder on your solution
  • Configuring the RaptorRecommendation add-in
  • Rendering recommendations in frontend

Read more about Raptor at their website

To integrate with Raptor, you must build the RaptorRecommendation project and move the RaptorRecommendation dll to the bin folder on your solution.

Get the project from the downloads section.

The project is an implementation of two Dynamicweb concepts – notification subscribers and a RelatedProductListProvider – and it works like this:

  • We use notifications to collect and transfer data about visitor behavior to Raptor
  • We use a RelatedProductListProvider to pull information about recommendations from the Raptor web service

You can then render the related products in frontend using the RaptorRecommendation tags and loops.

By default, the project contains the following types of recommendations:

  • Top selling products
  • Most visited products
  • Related items
  • Similar items
  • User history

Any other types of recommendations require you to customize the RaptorRecommendation project.

Once the RaptorRecommendation dll is in your bin folder you must configure the Raptor recommendation add-in (Figure 3.1).

Figure 3.1 The default RaptorRecommendation add-in

Go to Settings > Integration > Integration Framework Live and:

  • Click the Raptor integration add-in
  • Fill in your customer ID and the Raptor API key
  • Specify how many recommendations you want Raptor to return

You can also download and view a log file and test the connection between your solution and Dynamicweb.

Once the add-in is configured, your solution will start transmitting data to Raptor immediately – but it may take a while before Raptor has enough data to make recommendations from. 

When the integration is up and running you can use the RaptorRecommendation tags and loops in your product catalog templates (list and details).

The loops available to you are:
















You also have access to tags for counting the recommendations, and of course to various data inside each loop.

When a user clicks on a product that's being displayed because of a recommendation from Raptor, you can return an "ItemClick" registration to Raptor. This helps you (and Raptor) keep track of the value you get from the recommendations engine.

To trigger the ItemClick event, your links must contain the query string parameter RaptorRecommendation="YourMethod", where YourMethod is the method used to render the product.

For example:

The tracking link to raptor will have the format below:[customerid].rsa?p1=ItemClick&p2=[ProductId]&p3=[ProductName]&p4=[ProductGroupPath]&p5=[ProductPrice] &p6=[CurrencyCode]&p7=[CurrentUserID]&ruid=[EncodedUserEmail]&coid=[UserIdConvertedToGUID]&am=GetSimilarItems