Skip to main content

Documentation Index

Fetch the complete documentation index at: https://amethystapps.com/llms.txt

Use this file to discover all available pages before exploring further.

Install and configure OpticMatch to boost your Average Order Value (AOV) using AI-powered visual product recommendations and high-conversion bundle discounts. OpticMatch is built for high-performance scale on both Shopify and BigCommerce.

Installation Guide

1. Find the App

Search for “OpticMatch — AI Visual Upsell” on the Shopify App Store or the BigCommerce App Marketplace. Look for the Amethyst purple icon in the Upsell or Marketing categories. View App Details →

2. Install the App

Click “Install” or “Add app”. Review the permissions — OpticMatch requires access to your product catalog to analyze imagery and the ability to create discount codes for your bundle offers.

3. Open the App

Navigate to your store admin apps section and launch OpticMatch. Our unified interface ensures a seamless experience whether you are managing a Shopify Plus or BigCommerce Enterprise store.

How it Works: The AI Vision Engine

Unlike traditional apps that rely on simple “Customers Also Bought” data, OpticMatch uses advanced AI to see your products just like a human merchandiser would.
  • Visual Analysis: We use Replicate (CLIP) to analyze the aesthetic and style of your product images.
  • Semantic Matching: GPT-4o analyzes titles and descriptions to ensure recommendations make logical sense (e.g., pairing a snowboard with the correct bindings).
  • Automated Sync: Every time you add a new product, OpticMatch automatically generates new pairings in the background across all your connected storefronts.

Setting Up Your Storefront

1. Choose Your Recommendation Style

Navigate to the Recommendations tab. Here you can see the AI-generated pairings. You can:
  • Accept AI Pairs: High-confidence matches are active by default.
  • Manual Overrides: Manually delete pairings or create your own “Power Pairs.”
  • Match Types: Look for the “AI-Visual” or “Rule-Based” badges to understand the logic behind the match.

2. Add the “Complete the Set” Widget

  • For Shopify: Open your Theme Editor, navigate to a Product Page, and add the OpticMatch — Complete the Set App Block.
  • For BigCommerce: Use our native Page Builder component or inject the widget via the Script Manager for custom Stencil themes.
  • Customization: Fully customize colors, fonts, and “Add to Cart” button styles to match your brand identity.

3. Create Bundle Discounts

Go to the Bundle Discounts tab to create “Buy Both & Save” offers. Set a percentage discount (e.g., 15% off) that is automatically applied at checkout when the bundle button is clicked, leveraging the native checkout APIs of your respective platform.

Monitoring Performance

The Analytics Dashboard

OpticMatch provides a real-time view of your ROI:
  • Impressions: How many times your recommendations were seen.
  • Click-Through Rate (CTR): Which pairings are catching your customers’ eyes.
  • Conversion Lift: The actual dollar amount OpticMatch has added to your total revenue.

Frequently Asked Questions

We use CLIP (Contrastive Language-Image Pre-training) to convert your product images into mathematical vectors. This allows the app to find products that ‘look’ good together based on color, shape, and style.
No. OpticMatch is built on a serverless edge architecture. The storefront widget loads asynchronously after your main content is interactive, ensuring zero impact on your Core Web Vitals or BigCommerce PageSpeed scores.
Yes. If you are on BigCommerce, OpticMatch respects Customer Group pricing and visibility settings, making it the perfect choice for complex B2B catalogs.
Yes. Since we are built on Gadget, we offer a robust API for headless builds (Next.js, Hydrogen, etc.). Contact our engineering team for documentation.

Need a Custom AI Model?

For high-volume Plus and Enterprise stores, we can fine-tune recommendation rules based on your specific industry niche. Contact Amethyst Engineering

Start Selling More with AI →