The Recommendation API provides a range of personalization capabilities that tailor product recommendations to individual users. By leveraging user behavior, such as browsing history, purchase activity, and real-time interactions, the API delivers more relevant product suggestions that increase engagement and conversion rates.
To enable personalization, you must include the Insider ID in your Recommendation API requests. This allows the system to associate incoming requests with existing user profiles and apply behavior-driven logic.
Currently, the Recommendation API supports the following personalization features:
Personalized Recommendation Algorithms
The following recommendation algorithms inherently deliver personalized product suggestions based on user behavior and interaction data.
User-Based Recommendations
Generate personalized recommendations by analyzing a user’s historical interactions and the behavior patterns of similar users.
Real-Time User Engagement Recommendations
Provide real-time personalized suggestions based on the user’s current session activity and live interactions.
Recently Viewed Products
Recommend products based on the user’s most recent product page views, helping reinforce recent browsing intent.
Purchased with Last Purchased
Suggest products that are frequently purchased together with the user’s most recent purchase, supporting effective cross-sell scenarios.
These algorithms leverage both historical and real-time data to ensure recommendations remain relevant, timely, and aligned with individual user preferences.
Personalization with the User’s Last Visited Item
The Recommendation API can personalize results using the last product page a user visited. This behavior applies to both User-Based Recommendations and Real-Time User Engagement algorithms.
When the API does not have sufficient user-level data to generate a fully personalized recommendation, it automatically falls back to Viewed Together recommendations based on the user’s most recently viewed product.
This fallback mechanism ensures that recommendations remain relevant and context-aware, even when historical user data is limited.
Personalization with Users Recent Interactions
The Recommendation API can automatically exclude products a user has already interacted with, so they won't see recommendations for items they've viewed, purchased, or otherwise engaged with.
Use the following parameters to enable these exclusions:
API Endpoint Parameter | Description |
|---|---|
excludeViewItem | Excludes the last X Product Visits of the user from the API response |
excludeViewDay | Excludes Product Page views of the user in the last X days from the API response |
excludePurchaseItem | Excludes the last X Product Purchases of the user from the API response |
excludePurchaseDay | Excludes Product Purchases of the user in the last X days from the API response |
The following Recommendation API requests illustrate these personalization features:
The request below instructs the Recommendation API to exclude the last ten products the user viewed in Product Detail pages from the API response:
https://recommendationv2.api.useinsider.com/v2/most-popular?details=true¤cy=TRY&locale=tr_TR&partnerName=dataforceapi&userId=testUser&excludeViewItem=10The request below instructs the Recommendation API to exclude the products that the user purchased in the last three days.
https://recommendationv2.api.useinsider.com/v2/most-popular?details=true¤cy=TRY&locale=tr_TR&partnerName=dataforceapi&userId=testUser&excludePurchaseDay=3Attribute Affinity
Attribute affinity scores represent a user’s preference for specific product attributes based on their interaction behavior, including product views, add-to-cart actions, and purchases. Purchases carry a higher weight than views, as they indicate stronger intent. Affinity scores are normalized by a user’s total activity. As a result, a user with a high purchase-to-view ratio demonstrates a stronger affinity than a user who browses frequently but converts rarely.
The Recommendation API uses attribute affinity to personalize product recommendations according to each user’s demonstrated interests. Affinity data is refreshed daily, ensuring recommendations reflect recent and relevant user behavior.
Within the Recommendation API, the hp endpoint parameter controls whether the Attribute Affinity feature is applied. When enabled, the API incorporates a user’s attribute affinity scores into the recommendation logic, prioritizing products that align with the user’s strongest preferences.
The following Recommendation API request demonstrates how the Attribute Affinity feature is applied when the hp parameter is enabled:
https://recommendationv2.api.useinsider.com/v2/most-popular?details=true¤cy=TRY&locale=tr_TR&partnerName=yourPartnerName&userId=testUser&hp=1