Overview

Complete the Look (CTL) is an AI-powered product recommendation system that shows complementary items to help customers complete their outfit or purchase. CTL enhances the online shopping experience by displaying contextually relevant suggestions following to the anchor product being viewed.

For Example, A customer views a Men's Slim Fit Navy Shirt then the CTL would recommend:

• Slim-fit jeans
• White sneakers
• Sunglasses
• Leather strap watch

📘

Prerequisites

To access the AI based CTL feature, contact our team at [email protected] to activate the API key on your console to begin your first experience!

Business Benefits

Step by Step Workflow

1. Domain-Adaptive LLM Training Pre-trained language models available for each vertical like fashion, electronics, home decor, etc. The LLM is trained on your product catalog which adds the below values:
   • Style language adaptation
   • Context injection like, formalwear VS sportswear
2. Intelligent Category Path Selection: Our AI selects the best category level to ensure:
   • High contextual accuracy
   • Adequate product density

Product Feed example:

Cat 1: Men  
Cat 2: Men > Clothing  
Cat 3: Men > Clothing > Shirts  
Cat 4: Men > Clothing > Shirts > Slim Fit Shirts ← Selected

3. Supporting Category Mapping: The system automatically maps supporting categories for each anchor category. For example:
   • Anchor: Men > Clothing > Shirts > Slim Fit
   • Mapped Supporting Categories:
     • Men > Bottomwear > Jeans
     • Men > Footwear > Casual > Sneakers
     • Men > Accessories > Eyewear > Sunglasses
     • Men > Accessories > Watches > Leather Strap

Product Recommendation Engine

1. Similarity Model: Fetches products with visual or attribute-level similarities. It ensures aesthetic cohesion.
2. Bought-Also-Bought: Based on actual transaction data. It boosts high-confidence product pairing.
3. Viewed-Also-Viewed:Identifies browsing behavior patterns. It helps surface the alternatives yet relevant items.
4. Top-Selling Category Products: Highlights best-sellers in each supporting category. It acts as fallback when Bought-Also-Bought/Viewed-Also-Viewed data is limited.
5. Filtering & Ranking: Filters out out-of-stock or irrelevant items. It ranks by:
   • Popularity
   • Click-throughs
   • Add-to-cart metrics

Final Output Format

1. Products are grouped under five supporting categories.
2. Each group includes up to 10 highly relevant items based on the colour and style of the anchor product.
3. Delivered in a structured, clean response, like the current recommendation response.

API Integration Guide

API Endpoint:

GET https\://recommendations.unbxd.io/v3.0/\<siteKey>/\<apiKey>/ctl?productId=\<productId>\&uid=\<uid>

Required Parameters:

1. siteKey and apiKey
2. productId (anchor product)
3. uid or netcoreId for personalized results

Where to Use CTL API

Use Case Scenarios

Product Detail Page (PDP)

• Customer Action: Views navy shirt
• AI Recommendation: Jeans, shoes, and watch
• Business Impact: Average order value (AOV) rose from 1.2 to 1.8 items per order

Outcome
It recommends complementary products like jeans, shoes, and a watch when customers view a navy shirt, increasing Average Order Value from 1.2 to 1.8 items per order.

Cart Page Recommendations

• Customer Action: Adds shirt to cart
• AI Recommendation: Suggests sneakers and belt
• Business Impact: Reduced cart abandonment and improved conversions

Outcome
This feature recommends relevant add-ons like sneakers and a belt, reducing abandonment and boosting conversions.

Conclusion

AI-Based Complete the Look (CTL) is a smart product bundling and recommendation engine. It is trained on your data and customized for your vertical. It reduces manual effort, improves user experience, and drives revenue.
With an API-first architecture, CTL integrates seamlessly across the customer journey—from product views to checkout.