Empathic Recommendations is an adaptive approach that applies the optimal recommendation algorithm for each user based on their state of mind, journey stage, and location on the site. This built‑in logic personalizes each experience without requiring you to manually create different experiences with different algorithms for different users.
Using Empathic Recommendations in your campaigns
When you apply the Empathic Recommendation algorithm to your campaigns, you are allowing our decision engine to determine your site visitor's state of mind, and apply the most optimal of the available algorithms based on that and the page context. Each of our defined states of mind relates to a different stage in the user journey and accounts for what they are most likely interested in.
States of mind and associated algorithms
| User state of mind | Algorithms applied |
|---|---|
| Curious: The user is new to the brand and has no history. | Popularity or Popular in Location, to surface trending products based on other user behavior. |
| Interested: The user began browsing and indicated an interest in a category or a product. | A mix of algorithms, depending on the page context, to surface trending products related to the user's recent interest:
|
| Focused: The user indicated a strong interest in a specific product by adding it to the cart or wishlist. | Prioritizes algorithms like User Affinity and NextML to surface similar or complementary products. |
| Satisfied: The user recently made a purchase on your site. | Purchased with recently purchased, in addition to the algorithms applied for the focused state of mind. |
Good to know:
- The Empathic Recommendations algorithm also considers the site’s industry when recommending products. For example, it incorporates recently purchased items into the recommendation for packaged goods and groceries.
- When a site visitor browses a specific category, Empathic Recommendations automatically recommends popular items from that category.
- You can add custom rules to the Empathic Recommendations algorithm. For example, you can prevent recommending products with a specific attribute.