Product Recommendations Enrichment

The Product Recommendations Enrichment creates a personalised set of recommended products for each Customer. These recommendation sets can then be applied to your newsletters, on-platform personalisation, follow-up purchase messaging etc. 

This Enrichment requires Customer and Product data as well as either Tracking and/or Purchase history data to be added to Distil

This Enrichment creates the following attributes for each recommended Product for each Customer:

  • Product 1 ID
  • Product 1 Image Url
  • Product 1 Name
  • Product 1 Precis
  • Product 1 Price (Ex. Tax)
  • Product 1 Price (Inc Tax)
  • Product 1 Score
  • Product 1 Shop Url
  • Product 1 Thumbnail Url
  • Product 2 ID etc……..for up to 5 Products per Customer

These attributes can be found in the Product recommendations mix tab in the Data panel of the Customers screen. 

Additionally up to 10 default recommended products can be set which can be used if a personalised recommendation list cannot be calculated for a Customer. These produce the following attributes for every customer:

  • Product 1 ID
  • Product 1 Name
  • Product 1 Score
  • Product 2 ID etc…..for up to 10 Products per Customer

These attributes can be found in the Product recommendations (Default Products) tab in the Data panel of the Customers screen.

Product Recommendations Enrichment Settings

Within the settings for the Product Recommendations Enrichment there are three separate tabs, the Mixing DeckDefault Products and Options.

Mixing Deck

The algorithm that calculates the list of recommended products can be tuned using the Recommendation Mixing Desk. The weighting of each of the variables that are used by the algorithm can be adjusted  between a range of 0 – 11, with 11 being the highest and 0 being switched off. These adjustments determine the Products that will be recommended to each Customer.


The following variables can be adjusted using the sliders in the Mixing Desk:

  • Viewed Not Purchased – Products which an individual Customer has viewed, but not yet purchased.
  • Viewed Together – Products that are commonly viewed together by all Customers.
  • Purchased Together – Products that are commonly purchased together by all Customers.
  • Follow on Purchase – Products that are commonly purchased after another product that the Customer has already purchased.
  • Trending over past n day(s) – Products that have the highest number or orders over the last n day(s). n can be specified in the box below the slider.
  • Default Products – Products that are in the list of Default Products (specified in the Default Products tab) .

An additional checkbox is shown below these sliders that enables already purchased products to be excluded or included from the list of recommended products.

Beside the sliders there is a panel where the recommendations generated for any selected Customer can be easily viewed by entering the customer’s name in the Search Customer box. This enables the user to more effectively tune the Recommendation Engine by immediately seeing the effect of adjustments to the algorithm’s variables.

Default Products

The Default Products tab gives the user the ability to curate a custom list of products that can be used as an input into the algorithm that generates Customer recommendations. For example these products can be recommended to Customers that have no previous order or browsing history, or, if the Default Products is selected as the only variable on the mixing desk, then all Customers will be recommended this list of products.


The product(s) required to be added to the Default Products list can be searched for using the search bar at the top of the panel, and the ‘+’ symbol clicked to add this product to the selected list.

Selected Products can be reordered within the list or removed from the default products list by clicking on the ‘X’ beside the Product name.


The Options tab enables Distil users to add additional recommendation sets to each Customer, in addition to the recommendations generated by the Mixing Deck. So for example customers can have not only the Product recommendations generated by the settings in the mixing deck, but also have a separate set of trending products.


Next Steps

Chronotype Enrichment

Engagement Enrichment

Spend Bracket Enrichment

Building Segments