iAdvize enables you to use trusted generative artificial intelligence to handle conversations autonomously. This article explains how AI Shopping Assistant work—powered by this generative AI—and presents the related use cases.

Overview

By using trusted generative AI, you allow your Assistant to handle conversations autonomously: it generates its own answers based on visitors’ questions and the information available to it, rather than using pre-written scenario responses. It is also able to understand images sent by visitors, which allows it to handle requests related to visuals (for example, a damaged product) and improve the quality of interactions.

Note: Thanks to generative AI, the Assistant automatically adapts to the visitor’s language during the conversation, saving you from creating multiple Assistants for each language.

The Assistant can also engage a visitor whose browser language differs from the configured one (this applies only to Assistants without transfer to a human agent). The widget and the Assistant messages will be translated into the visitor’s browser language.

If a language is forced in your tag, it will take priority. The visitor will see the Assistant and the first messages in that forced language. If they then write in another language, the Assistant will reply in that language.

However, you still remain in control of what the Assistant says:

by deciding the information sources (Knowledge) it will rely on to answer your visitors.
by configuring custom actions for specific situations (technical error, no available answer, prohibited words or phrases, request to speak with a human).

FR Copilot for Shoppers - Short video.gif

Use Cases

Product information

To address this use case, you can feed a Assistant with the data you already have about your products. It will be able to understand visitors’ questions about products and respond within a conversation, drawing its answers from the data provided.

Example: your visitor wants to buy the SuperAspi2000 vacuum cleaner. They have questions about this specific model (price, technical specifications, repairability score, etc.). Based on the product information made available to it, your Assistant can answer your visitor’s questions autonomously and in natural language.

Product discovery

This discovery and advisory capability allows your Assistant to suggest products suited to the specific needs of each visitor to your site:

It can understand the product type and criteria expressed by visitors in natural language. For example: “I’m looking for an Android tablet.”
It can ask the visitor questions to refine their criteria. For example: “Are you interested in a particular screen size, storage capacity, or any other specific requirements?”
It can gather the visitor’s budget and suggest products within the specified price range.
It can suggest between one and three products that best match the visitor’s criteria.
It can detect when visitors change their criteria and adapt product suggestions accordingly.
It can answer any questions about the suggested products and provide comparisons.
It can answer general questions to offer expert advice on any type of product.

For example: “I’m looking for a new electric bike. What should I consider to choose the right one for me?”. Your visitor does not have a specific model in mind but wants a wireless model with a specific budget. Based on these criteria, Assistant draws on the product information available to offer your visitor a range of options.

FAQ

Your visitors often want a quick answer without having to browse your entire FAQ. You can therefore feed your Assistant with this rich source of information. Thanks to this feature, it can answer autonomously all FAQ-type questions for which you have provided information.

Example: Your visitor has now found the vacuum cleaner that suits them. They still have questions about delivery and after-sales service. Your Assistant will be able to answer autonomously based on the FAQ information you have provided.

Features

Add-to-Cart

The Add-to-Cart feature enhances the online shopping experience by allowing visitors to add products directly to their cart from the conversation. When a Assistant recommends a product, the visitor can simply click the “Add to Cart” button and choose key options such as size or color—without ever leaving the chat.

This feature enriches product discovery by offering a guided and interactive exploration for shoppers. By providing real-time decision support, “Conversational Add to Cart” ensures a smooth and engaging shopping experience.

Customers no longer need to navigate multiple pages or interrupt their journey, reducing friction and simplifying the purchasing process.

Marketing without tag.png

To learn more about configuring this feature, see this article.

Compatibility – Public Beta

Allow your customers to find products compatible with those they already own or plan to purchase.

For example: I’m looking for a case for my iPhone 16 Pro.

Read this article to learn more

Cross selling – Public Beta

Enables your Assistant to recommend complementary products (accessories, related items, etc.) during a conversation.

Read this article to learn more

Quick replies

Quick Replies are clickable buttons generated by your Assistant after a response. They appear after a product-related question. Visitors can click a suggestion or freely type their message.

Up to 4 options are shown and generated dynamically based on the conversation context and your knowledge.

Quick Replies aim to reduce your visitor’s cognitive load and make the conversation smoother.

Requirements:

Shopping Panel enabled
A product-related question is asked by the visitor

Setup

Learn how to set up your Assistant in this article, then how to provide it with your information using AI Knowledge.