Ideas That Power Digital Growth
Stay ahead with actionable insights, expert opinions, and practical guides covering ecommerce trends, platform updates, automation, AI, and real-world solutions to help businesses grow and scale digitally.
02/24/2026
Boost Sales with AI-Powered Product Recommendations
Personalize every shopper's experience with smart recommendations that learn from real behavior and drive more conversions automatically.Get Started
Ecartify: AI Recommender – Smart Product Suggestions Powered by Real Shopper Behavior
Boost your store revenue by intelligently recommending relevant products to customers during their purchase journey. The AI Recommender addon uses vector-based AI technology to learn what your customers search, view, add to cart, and buy — then shows them the most relevant products at the right time. Powered by Qdrant (vector database) and Cohere (AI embeddings), this addon delivers truly personalized recommendations that increase average order value, reduce bounce rate, and keep customers coming back for more.
Key Features of AI Recommender
Behavior-Based AI:
Learns from search, views, cart adds, and orders to build a personalized profile for each customer.
Learns from search, views, cart adds, and orders to build a personalized profile for each customer.
Vector Search (Qdrant):
Uses lightning-fast vector similarity search to find the most relevant products in milliseconds.
Uses lightning-fast vector similarity search to find the most relevant products in milliseconds.
Cohere AI Embeddings:
Converts product and behavior data into smart embeddings that understand context and meaning.
Converts product and behavior data into smart embeddings that understand context and meaning.
Weighted Activity Scoring:
Configure how much each action matters — views, cart adds, searches, and orders each carry adjustable weights.
Configure how much each action matters — views, cart adds, searches, and orders each carry adjustable weights.
Automated Product Sync:
Background cron job automatically syncs your entire product catalog into the AI vector database.
Background cron job automatically syncs your entire product catalog into the AI vector database.
Filter by Activity Type:
Show recommendations based only on Search, View, Cart, Order activity — or mix all together.
Show recommendations based only on Search, View, Cart, Order activity — or mix all together.
Flexible Block Display:
Add a Recommendation block anywhere on your storefront using CS-Cart's layout manager.
Add a Recommendation block anywhere on your storefront using CS-Cart's layout manager.
Responsive on All Devices:
Recommendations display beautifully on Phone, Tablet, and Desktop with the Scroller template.
Recommendations display beautifully on Phone, Tablet, and Desktop with the Scroller template.
Optional Add-to-Cart Button:
Show or hide the Add to Cart button on recommendation cards to match your store's design.
Show or hide the Add to Cart button on recommendation cards to match your store's design.
Step-by-Step Setup Guide
Step 1: Install the Addon
Get the AI Recommender running on your CS-Cart store in just a few minutes:
Download from the CS-Cart marketplace or the Ecartify website
Go to Add-ons → Manage Add-ons in your admin panel
Click the + button and upload the "Ecartify: AI Recommender" package
Click Install and then set the Status to ON
Once activated, go to Add-ons → Ecartify: AI Recommender to open the settings panel.
Step 2: Connect Your AI Services (Qdrant + Cohere)
The AI Recommender needs two external services to work. Both have free tiers to get you started:
Qdrant URL:
Your Qdrant cloud cluster URL (e.g.
Your Qdrant cloud cluster URL (e.g.
https://xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx.us-east4-0.gcp.cloud.qdrant.io). Sign up free at cloud.qdrant.io
Qdrant API Key:
Generated from your Qdrant cloud dashboard. Used to authenticate requests to the vector database.
Generated from your Qdrant cloud dashboard. Used to authenticate requests to the vector database.
Cohere API Key:
Get your free API key from cohere.com. Cohere converts product data into AI "embeddings" (smart numeric representations) for similarity matching.
Get your free API key from cohere.com. Cohere converts product data into AI "embeddings" (smart numeric representations) for similarity matching.
Step 3: Set Activity Weights
Tell the AI how much each customer action matters when calculating recommendations. Higher weight = more influence on the results:
Weight of View Product (default: 1):
How much a customer simply viewing a product page influences their recommendations. Low weight since browsing doesn't always mean strong interest.
How much a customer simply viewing a product page influences their recommendations. Low weight since browsing doesn't always mean strong interest.
Weight of Add to Cart (default: 2):
Adding a product to cart shows strong buying intent. Higher weight means cart actions influence recommendations more.
Adding a product to cart shows strong buying intent. Higher weight means cart actions influence recommendations more.
Weight of Search Product (default: 1):
When a customer searches for something, the AI learns what they're looking for. Searching shows active interest.
When a customer searches for something, the AI learns what they're looking for. Searching shows active interest.
Weight of Create Order (default: 3):
Completing a purchase is the strongest signal. A high weight means past purchases heavily shape future recommendations.
Completing a purchase is the strongest signal. A high weight means past purchases heavily shape future recommendations.
Step 4: Set Up Product Sync (Cron Job)
The AI needs to know about all your products. A background cron job syncs your product catalog to Qdrant automatically — every minute in the background:
Set a Cron Password:
Enter a secure password (e.g.
Enter a secure password (e.g.
MYPASS). This protects the cron endpoint from unauthorized access.Copy the Cron Command:
The addon shows the exact command to add to your server's cron scheduler.
The addon shows the exact command to add to your server's cron scheduler.
Step 5: Add the Recommendation Block to Your Storefront
Now display personalized recommendations to your customers by adding the Recommendation block to any page layout:
Go to Design → Layouts:
Open the CS-Cart layout manager and select the page where you want recommendations to appear (homepage, product page, etc.)
Open the CS-Cart layout manager and select the page where you want recommendations to appear (homepage, product page, etc.)
Add a New Block:
Click the +Icon in your chosen grid position and search for and check products and found in checkbox "Recommendation" in the check list
Click the +Icon in your chosen grid position and search for and check products and found in checkbox "Recommendation" in the check list
Configure General Tab:
Set the block Name (e.g. "Recommendation"), choose Template as Scroller, set Wrapper to Main block (H1 heading), and select which devices to show on (Phone, Tablet, Desktop)
Set the block Name (e.g. "Recommendation"), choose Template as Scroller, set Wrapper to Main block (H1 heading), and select which devices to show on (Phone, Tablet, Desktop)
Configure Content Tab:
Set Filling to Recommended Products, set Limit (e.g. 10 products), set Page to 1, and choose which activity types to base recommendations on: All, Search, View, Cart, or Order
Set Filling to Recommended Products, set Limit (e.g. 10 products), set Page to 1, and choose which activity types to base recommendations on: All, Search, View, Cart, or Order
Step 6: Configure Block Settings
Fine-tune how the recommendation block behaves on your storefront:
Hide Add to Cart Button:
Check this box if you want a cleaner look — customers click the product to view it before buying. Uncheck to show "Add to Cart" directly on recommendation cards.
Check this box if you want a cleaner look — customers click the product to view it before buying. Uncheck to show "Add to Cart" directly on recommendation cards.
Limit:
Control how many products appear (default: 10). The Scroller template lets customers scroll through all recommendations smoothly.
Control how many products appear (default: 10). The Scroller template lets customers scroll through all recommendations smoothly.
Save the Block:
Click Save and then publish your layout changes. Visit your storefront to see the recommendation block live
Click Save and then publish your layout changes. Visit your storefront to see the recommendation block live
Step 7: See It Live on Your Storefront
Once everything is set up, your storefront will show a beautiful "Recommendation" scroller section to every customer. Here's what happens automatically:
Customer Visits Store:
As soon as a customer views products, searches, or adds to cart — the AI starts tracking their preferences in real-time.
As soon as a customer views products, searches, or adds to cart — the AI starts tracking their preferences in real-time.
AI Calculates Best Matches:
Qdrant searches through all product vectors to find items most similar to the customer's behavior profile, weighted by your activity settings.
Qdrant searches through all product vectors to find items most similar to the customer's behavior profile, weighted by your activity settings.
Scroller Displays Products:
A horizontal product scroller appears on the page showing personalized recommendations — each customer sees different products based on their own behavior.
A horizontal product scroller appears on the page showing personalized recommendations — each customer sees different products based on their own behavior.
Gets Smarter Over Time:
The more a customer interacts with your store, the more accurate and personalized their recommendations become.
The more a customer interacts with your store, the more accurate and personalized their recommendations become.
Real-World Benefits & Results
Increase Average Order Value
Relevant Cross-Sells: Customers discover products that genuinely match their interests — increasing basket size naturally without aggressive selling.
Behavior-Driven: Unlike manual "related products," AI recommendations adapt to each customer — so every shopper sees what's most relevant to them.
Purchase History Learning: Repeat customers get recommendations based on past orders — encouraging add-ons, upgrades, and complementary items.
Boost Conversion Rates
Right Product, Right Time: Showing customers what they're actually interested in dramatically reduces bounce rate and increases time on site.
Search Intent Captured: If a customer searches "blue shoes" and doesn't find exactly what they want, recommendations surface similar options automatically.
Reduces Decision Fatigue: Instead of overwhelming customers with your full catalog, the AI curates a small set of highly relevant products.
Better Customer Experience
Personalized for Every Visitor: No two customers see the same recommendations — each experience feels tailored and unique.
Discovers Hidden Catalog Gems: Products that don't appear in top search results or categories get surfaced to the right customers through AI matching.
Builds Customer Loyalty: A store that "gets" what you like keeps customers coming back — personalization creates stickiness.
Zero Manual Work
Fully Automated: Once the cron job is set, product sync and customer tracking happen automatically — no ongoing manual work needed.
Self-Improving: As more customers use your store, recommendation quality improves organically with more data.
No ML Expertise Required: The addon handles all the AI complexity — you just configure weights and place the block.
Transform Your Store Into a Personalization Engine
Your AI Recommender addon is ready to make every customer's shopping experience feel like it was designed just for them. Say goodbye to generic "You might also like" lists that show the same products to everyone.
Pro Tip: Start with the default weights (View=1, Cart=2, Search=1, Order=3) and let the system collect data for 1-2 weeks. Then analyze which activity type drives the most conversions in your store and increase that weight. Set the recommendation block limit to 10 and place it just below your main homepage banner for maximum visibility.
