Einstein Recommendations

Working with Einstein Recommendations in Marketing Cloud Engagement.

Engagement

Originally published

First published 2020-09-27 in the SFMC Cookbook. Ported here for sfmc.guide.

This page aims to make using Einstein recommendations a little easier by adding a few explanations around how to use it in a modern setup.

• Collect Code
  • Note to developers
  • Initialize the library
    • Collect Code on page load
    • Asynchronous Collect Code
    • Asynchronous Collect Code with preloading
  • Debug your tracking solution
  • Tracking and other Collect Code features
    • Disable tracking
    • Identify Business Unit for Tracking
    • Identify current user
      • Attribute Affinity
    • Track Page Views: trackPageView
    • Track Items in Cart: trackCart
    • Track Purchases / Conversions: trackConversion
      • Tracking overhead cost
    • Track Custom Event: trackEvent
    • Track User Wishlist: trackWishList
• Update Catalog
  • Via Collect Code
  • Update Catalog via API
• Predictive Intelligence (PI) Data Extensions
• Einstein Email Recommendations
• Einstein Web Recommendations
  • Embedding Web Recommendations
    • How the recommender knows who the current user is
    • Enhancing recommendation results
    • Embed via JSON
      • JSON Example responses
    • Embed via JavaScript (“HTML”)
      • JavaScript/HTML example code
  • Debugging Web Recommendations
• Embedding Collect Code via Google Tag Manager (GTM)
  • Loading the GTM library
  • Loading collect.js via GTM
    • GTM: Developer Playground
  • Logging events via GTM
    • Identify current user via GTM
  • Debugging / Previewing your GTM setup
  • Publishing your changes
  • Events mapping: Collect Code to GA4 Retail/Ecommerce