Contextual Product Recommendation

This guide explains all configuration options available in the Contextual Product Recommendation module. By adjusting these settings, you can create smart product recommendations that show customers relevant items based on what they're currently viewing or have in their cart. This module generates either complementary products (items that go well together) or substitute products (alternative options) for any given item.

What is Contextual Product Recommendation?

Contextual Product Recommendation analyzes relationships between products to suggest relevant items when customers view specific products. Unlike general personalized recommendations, these are based on the specific context of what the customer is currently viewing.

Common Use Cases:

E-commerce Product Pages: Show "Customers who bought this also bought" or "Complete the look" suggestions
Shopping Cart Recommendations: Suggest complementary items before checkout
Cross-selling: Recommend accessories or related products that enhance the main item
Upselling: Show substitute products with better features or higher value
Bundle Creation: Identify products that are frequently purchased together

Setup Properties

These are fundamental system settings that must be configured during initial implementation.

Recommender Schedule:

Purpose: Set when recommendation updates occur using a cron schedule
Options: Cron expression format
Use Case: Update relationships daily at midnight, or weekly to capture changing trends

Recommendation Type:

Purpose: Type of recommendation to generate for each item
Options:

Complement: Items that work well together
Substitute: Alternative items that can replace the current item
Use Case: Use "complement" for accessories, "substitute" for showing alternative products

Number of Recommendations

Purpose: Number of recommendations to return per item
Options: 1-50 (integer)
Use Case: Show 4-8 complementary items on product pages, or 3-5 substitute options

Configuration Properties

These are tuning settings that control the business logic and behavior of your recommendations. These can be adjusted and refined based on performance and business needs.

Fallback configuration: configure the fallback recommendations to add to the existing set of recommendations.

Fallback Configuration

Fallback Method

Purpose: Method to use when contextual recommendations are insufficient
Options:
- Popularity: Show popular items from the catalog
- Popularity Attribute: Show popular items based on specific attributes

Product Selection Rules: configure the selection of products to recommend

Item Filter

Purpose: Specify criteria for filtering items in recommendations
Options:

Enable/Disable: Turn this filter on or off
Inclusion Criteria: SQL expression builder defining which items to include
Use Case: Only show items from specific categories as contextual recommendations (e.g., only show "Accessories" when viewing electronics)

Item Exclusion Criteria

Purpose: Specify criteria for excluding certain items in recommendations
Options:

Enable/Disable: Turn this filter on or off
Exclusion Criteria: SQL expression builder defining which items to exclude
Use Case: Exclude out-of-stock items, discontinued products, or items with low ratings from contextual recommendations

Product Boosting: Boost products from being recommended more frequently based on numerical or categorical attributes.

Boost by Numerical Attribute

Purpose: Set up boosts for items based on their numeric attributes
Configuration Options:

Enable/Disable: Turn this boosting on or off
Boost Strength: Value between 0-10 that determines boost intensity
Boost Factor Expression: Select a numerical product attribute to boost by
Use Case: Boost high-margin accessories, or prioritize items with better compatibility scores

Boost by Categorical Attribute

Purpose: Define how item groups based on specific conditions are boosted in recommendations
Configuration Options:

Enable/Disable: Turn this boosting on or off
Boost Condition: SQL expression builder defining which items to boost
Boost Factor: Value between 0-10 that determines boost intensity
Use Case: Boost items from the same brand, or prioritize new arrivals in contextual recommendations

Fixed Position Recommendations

Purpose: Place specific items at guaranteed positions in recommendation lists
Configuration Options:

Enable/Disable: Turn this positioning on or off
Item Condition: SQL expression builder identifying which item to position
Target Position: Fixed position (1-15) for the chosen item
How It Works: Rules are applied in order, with each rule affecting only one item
Use Case: Ensure high-margin accessories always appear in the first position when viewing main products

Diversity Control: Configure the diversity spread of recommendations for each product.

Maximum Recommendations per Category

Purpose: Define the maximum number of recommendations per category to ensure variety
Configuration Options:

Enable/Disable: Turn this diversity control on or off
Grouping Feature: Select the feature to group similar products (e.g., Color, Category, Brand)
Maximum Similar Products: Maximum number of products allowed with same feature value (1-15)
Use Case: When viewing a laptop, limit to maximum 2 laptop bags and show other accessories like mice, keyboards

Maximum Recommendations per Characteristic

Purpose: Control the variety of products by limiting recommendations with specific feature values
Configuration Options:

Enable/Disable: Turn this feature limit on or off
Product Feature Condition: SQL expression builder defining product feature values to limit
Maximum Items: Maximum number of items to include with this feature (1-15)
Use Case: Limit products where category = 'memory_cards' to maximum 1 item when viewing cameras.

Delivery via API Integration

To integrate contextual product recommendations into your website or application via API, technical setup and configuration is required.

For API integration assistance, please contact your CM Customer Success Manager (CSM). They will coordinate with technical specialists to help you: