Skip to main content

How to Track Sessions, Analyze Data & Customize Your Assistant for AI Shopper

Mikaela Lasig avatar
Written by Mikaela Lasig
Updated this week

AI Shopper Overview

The AI Shopper is your store’s smart assistant that helps shoppers in real time. It’s not just a chatbot, it learns about your brand, your products, and your customers to give helpful, personalized support.

It can answer questions, suggest products, help with fit, and guide customers to the right purchase—all while collecting useful insights for you.

This guide will walk you through how to use and customize your AI Shopper so you can boost sales and improve your customer experience.


Why Use AI Shopper?

Here’s how the AI helps both your customers and your business:

For Customers

  • Fast Answers: Instant replies about size, materials, care tips, and more

  • Style Advice: Recommends matching items or styling tips

  • Virtual Try-On: Helps shoppers see how products look on them

  • Easy Support: Routes customers to your team if needed

For Your Store

  • More Sales: Helps remove buying hesitation by answering questions

  • Customer Insights: See what people ask most and where they need help

  • Stronger Brand: Customize the AI to sound just like your brand

  • Less Work for Support: The AI handles common questions automatically


How to Fine-tune Your AI Shopper

You can personalize your AI so it fits your brand and does what you need.

Agent Settings

We auto-generate a starter profile from your brand, you can adjust anything.
Go to the Settings tab:

Agent Name – Give your assistant a brand-reflective name (default: “Shopping Assistant”).

Agent Image – Upload a profile picture or avatar to create a friendly presence.

Agent Profile – Describe background/expertise to help the AI give context-aware responses.

Agent Tone – Choose how your agent sounds: Professional, Friendly, or Enthusiastic.

Communication Style – Set response detail: Concise, Balanced, or Detailed.

Custom Prompt – Add specific instructions to control behavior, knowledge, or conversational style.

Key Features

  • Virtual Try-On

Allow customers to see what products look like on a model before purchasing. This is especially powerful for apparel and beauty brands.

  • Weekly Try-on Limit per Customer: Set the maximum number of try-ons allowed per customer each week using a slider (default is 50) to manage usage.

  • Default Silhouette: Choose the default model customers will see for try-ons: Male, Female, Beauty, or Custom. The preview image updates instantly to reflect your choice.

  • Advanced Settings:

    • Sign-in Options: Choose whether to ask for customer details Sign in first (before the try-on) or Suggest after try on.

    • Sign in prompt text: Customize the message that prompts users to sign in to save their try-on.

    • Button Label: Change the text on the primary sign-in button (e.g., "Sign in & Save").

  • Sizing: Enable this to provide sizing recommendations and fit guidance. You can also toggle Initial Size Recommendation to show size charts when recommending sizes.

  • Shade Finder: A must-have for cosmetics, this feature helps customers find their perfect foundation shade using AI analysis. You can customize the Body Text and set the Default Silhouette to Beauty.

Add to Cart: Allow customers to add products directly to their cart during conversations. You can also Blacklist by Tag or Blacklist by Collection to prevent certain items from being added to the cart via the chat.

Connect to Support: When a customer needs human assistance, this feature directs them to your support team. Simply provide a contact page URL or email address in the Support Contact field.


Klaviyo Integration

Sync customer data and create targeted email campaigns based on AI interactions. This integration automatically syncs:

  • Email addresses from signed-in users

  • Try-on images and product interactions

  • Customer segments based on shopping behavior

  • Chat conversation insights and conversion data

To set this up, simply enter your Klaviyo Private API Key and click Connect Klaviyo.


Chat Settings

Fine-tune the appearance of your chat interface to match your site's design. A preview pane on the right shows your changes in real time.

  • Background Color: Sets the overall background of the chat window.

  • User Message Color: Sets the background color of the customer's message bubbles.

  • User Message Text Color: Sets the color of the text within the customer's message bubbles.

  • Font Style: Choose a font that matches your website's typography.


Coupons

How to Create a Coupon

  1. Go to Coupons – From the dashboard, select Coupon in the left-hand menu.

  2. Create a New Coupon

    • First coupon: click Create Your First Coupon.

    • Subsequent coupons: click + Add Coupon at the top-right.

  3. Configure Coupon Details – Fill out:

    • Promo Code – Unique code for checkout (e.g., SAVE15).

    • Discount – Set percentage (type or use arrows).

    • Display Text – Message shown to customers (e.g., “Save 15% on your order!”).

    • Tip: You can reuse promo codes from your e-commerce platform for consistency.

  4. Set Status and Save – Toggle Active or Disabled, then click Save. Your coupon is now ready to use.


Tracking and Improving Performance

Review Conversations

In the Sessions tab, see what customers are asking and how the AI is responding.

You can view:

  • Dates

  • Questions asked

  • Whether they tried on products

  • Whether it led to a sale

This helps you:

  • Spot common sales-driving questions

  • Find gaps in your knowledge base

  • See how shoppers move through your site

Analytics

Use the Analytics tab to track performance and trends. The page is organized like this:

Date range

Pick a time window at the top right to update all metrics.

Overview cards

  • All Sessions, anyone who engaged with Tolstoy

    • Generated Revenue with a breakdown for Direct and Indirect

    • Conversion Rate

    • Sessions

  • Signed In Customers, shoppers who engaged and signed in

    • Generated Revenue with Direct and Indirect

    • Conversion Rate plus an Avg. signed-in CVR reference

    • Sessions

Generated Revenue Over Time, totals for All Customers vs Signed In


Conversion Rate Over Time, average CVR for All Sessions and Signed In


Sessions Over Time, total sessions and signed-in sessions

Most Popular Queries

  • A ranked list of top shopper questions from the AI chat

  • Each item shows share of all queries and the exact count

Features

  • Total Try-ons

  • CTR on Upsell Offers

Tips

  • Compare All Sessions vs Signed In Customers to see the lift from sign-ins.

  • Use Most Popular Queries to spot gaps in knowledge, then add links or files in the Knowledge tab.

  • If Try-on or Upsell CTR is low, review widget placement and prompts, and consider enabling coupons to encourage sign-ins.

Did this answer your question?