Smart Recommendations

Smart Recommendations uses machine learning to suggest the right products to the right customers at the right time. From upsells to cross-sells to personalized bundles, intelligent recommendations drive higher conversion and average order value.

AI-Powered Personalization

Smart Recommendations learns from your store's data — purchase history, browsing behavior, and product relationships — to deliver highly relevant suggestions that convert.

How Smart Recommendations Works

Data collection — Analyzes customer behavior and purchase patterns
Model training — ML models learn product relationships and preferences
Real-time scoring — Products ranked by relevance for each customer
Recommendations served — Best products shown at optimal moments
Continuous learning — Models improve with every interaction

Recommendation Types

Available Algorithms

Frequently bought together — Products commonly purchased in same order
Customers also viewed — Based on browsing patterns
Based on cart — Real-time suggestions from current cart contents
Personalized for you — Individual customer preferences
Trending products — Popular items in your store
Recently viewed — Products the customer has shown interest in
Complete the look — Complementary items (fashion, home)
Upgrade options — Higher-tier alternatives

Where Recommendations Appear

Smart Recommendations can be displayed throughout the customer journey:

Recommendation Placements

{
  "placements": {
    "product_page": {
      "algorithms": ["frequently_bought_together", "customers_also_viewed"],
      "position": "below_add_to_cart",
      "max_products": 4
    },
    "cart_page": {
      "algorithms": ["based_on_cart", "trending"],
      "position": "below_cart_items",
      "max_products": 3
    },
    "checkout": {
      "algorithms": ["complete_the_look", "upgrade_options"],
      "position": "post_purchase",
      "max_products": 2
    },
    "thank_you_page": {
      "algorithms": ["personalized", "trending"],
      "position": "main_content",
      "max_products": 4
    }
  }
}

AI Features

Collaborative Filtering

Recommends products based on what similar customers have purchased:

Identifies customers with similar purchase patterns
Suggests products those similar customers bought
Especially effective for new visitors with limited data

Content-Based Filtering

Recommends products based on product attributes:

Analyzes product descriptions, categories, tags
Finds similar products by attribute matching
Works well for finding alternatives or variants

Hybrid Approach

CheckoutOS combines multiple signals for best results:

Hybrid Recommendation

{
  "hybrid_scoring": {
    "signals": {
      "collaborative_score": 0.35,
      "content_similarity": 0.25,
      "purchase_frequency": 0.20,
      "margin_boost": 0.10,
      "inventory_availability": 0.10
    },
    "personalization_boost": {
      "returning_customer": 1.5,
      "vip_customer": 1.3
    }
  }
}

Performance Metrics

Track recommendation effectiveness:

Key Metrics

Recommendation click-through rate (CTR)
Recommendation conversion rate
Revenue attributed to recommendations
Average items per order from recommendations
Customer engagement with recommendation widgets

Basic Configuration

Recommendations Config

{
  "smart_recommendations": {
    "enabled": true,
    "default_algorithm": "hybrid",
    "fallback_algorithm": "trending",
    "personalization": {
      "enabled": true,
      "min_interactions": 3
    },
    "display": {
      "title": "You might also like",
      "max_products": 4,
      "show_price": true,
      "show_rating": true
    },
    "filtering": {
      "exclude_out_of_stock": true,
      "exclude_current_product": true,
      "exclude_cart_items": true
    }
  }
}

Best Practices

Recommendation Tips

Start with proven algorithms — "Frequently bought together" works universally
Match context to placement — Cart page = cart-based, thank you = personalized
Test titles and positioning — Small changes can have big impact
Monitor quality — Review recommendations regularly for relevance
Balance automation and control — Override AI when you know better

Learn More

AI Engine

Deep dive into the recommendation algorithms

Learn more

Manual Rules

Override AI with custom rules

Learn more