Ecom Copilot (AI Sales Assistant)

What is Ecom Copilot?

Ecom Copilot is a ChatGPT-powered conversational sales assistant that you can embed directly on your store. Customers open a chat widget, describe what they are looking for in plain language, and the assistant responds with tailored product recommendations, attribute comparisons, and buying guidance, all drawn from your live catalog.

Unlike the search bar, Ecom Copilot supports multi-turn conversations: the assistant remembers earlier messages in the session, so customers can refine requests (“show me something similar but cheaper”) without starting over.

How to Enable Ecom Copilot

1. Log in to your ExpertRec control panel.
2. Navigate to AI > Ecom Copilot.
3. Toggle the assistant on.
4. Configure the appearance and behavior settings described below.
5. Save your settings. A floating chat button will appear on your storefront at the configured position.

Configuration Options

• Widget position: Bottom-right or bottom-left corner of the page.
• Brand colors and theme: Set the primary color, font, and header text to match your storefront style.
• Greeting message: Customize the first message the assistant displays to new visitors.
• Assistant persona: Give the assistant a name and tone (friendly, professional, concise) that reflects your brand voice.
• Catalog scope: Optionally restrict the assistant to specific product categories or collections so it stays focused on relevant inventory.
• Response length: Choose short (one-liner recommendations) or detailed (full attribute breakdown) response style.

Demo Page

Before going live, use the built-in Demo tab inside AI > Ecom Copilot to preview the chat experience. The demo page loads a sandbox version of your catalog so you can test real queries and verify recommendations without affecting your live storefront.

Use Cases

• Guided product discovery: A customer unsure of which product to buy can describe their needs (“gift for a 10-year-old who likes science”) and receive curated suggestions.
• Attribute-based filtering: “Show me wireless headphones under $80 with noise cancellation” lets the assistant handle complex filtering that the standard search bar might miss.
• Post-search refinement: Customers who found several options but are unsure which to choose can ask the assistant for a direct comparison.
• Upselling and bundling: The assistant can suggest complementary products or higher-tier alternatives based on conversation context.

Best Practices

• Write complete product descriptions: Ecom Copilot retrieves product information to answer questions. Products with thin or missing descriptions will be surfaced less effectively.
• Set a focused greeting: A greeting like “Hi, I can help you find the perfect product. What are you looking for?” performs better than a generic “Hello”.
• Test edge cases in the demo: Try queries for out-of-stock items, very broad categories, and competitor brand names to see how the assistant handles them before launch.
• Review conversation logs: Use the reporting data to identify common intents and gaps in your catalog content that the assistant cannot answer well.