If you’re connecting to Shopify for the first time, you’ll see a quick onboarding flow to set up Fin for Ecommerce:

Connect to Shopify: Establish a live connection to power Fin and keep your catalogue up to date.
Install the Messenger: Enable customers to interact with Fin through the Messenger.
Deploy Fin: Preview and deploy Fin to your customers on Shopify.

This video below from our Customer Education team walks through the foundations of Fin for eCommerce and its set up.

Step 1: Connect your Shopify store

Get started by connecting your Shopify store to Intercom. Fin for Ecommerce is built for Shopify merchants, with automatic Shopify customer detection and a tailored onboarding experience to help you go live quickly.

Go to Fin AI Agent > Ecommerce > Train > Get started.
Click Set up Fin for Ecommerce.
Follow the one-click flow to connect your Shopify store.
Link one or more Shopify stores to Intercom.

Screenshot of the Fin AI Agent > Ecommerce > Get started page showing the 'Set up Fin for Ecommerce' button and the Shopify store connection flow.

Fin automatically detects Shopify customers and tailors the onboarding experience. When connected, all store content (products, pages, and articles) is imported into Intercom to power Fin. You can manage all content for Fin from the Train > Content section.

Step 2: Install the Messenger

To install:

Go to your Shopify admin dashboard.
Click Edit theme.
Navigate to App embeds.
Toggle on the Intercom app.
Click Save.

Once enabled, the Messenger appears on your storefront so customers can interact with Fin.

Step 3: Train Fin for Ecommerce

Fin for Ecommerce is powered by your Shopify catalogue, content, guidance, and Procedures.

Manage your store

Go to Fin AI Agent > Ecommerce > Train > Store to manage connected Shopify stores.

At least one store is required
All store content is automatically ingested when connected

Screenshot of Fin AI Agent > Ecommerce > Train > Store showing a connected Shopify store with store management options.

Manage your content

When configuring content for Fin for Ecommerce, note that pricing, availability, and variant information are synced directly from your Shopify catalogue and stay up to date automatically as your catalogue changes (you don't need to manage these manually).

From Fin AI Agent > Ecommerce > Train > Content you can configure which content Fin for Ecommerce can access. Make sure key ecommerce content, especially synced Shopify products is included, as Fin relies on this data to accurately search for and display products.

Screenshot of Fin AI Agent > Ecommerce > Train > Content showing the content management interface with the Shopify product catalog and other content sources available to Fin.

Filter your catalog

To refine what Fin recommends, select the three-dot menu next to your product catalog in Train > Content and select:

Configure product filters – Control which products are included.
Configure stock filter – Exclude out-of-stock items.

These filters ensure Fin only recommends relevant and available products.

Screenshot of the Train > Content interface showing the three-dot menu next to a product catalog with 'Configure product filters' and 'Configure stock filter' options visible.

Add guidance (optional)

For most stores, Fin for Ecommerce works well without additional guidance. You may want to add guidance if you have a complex catalogue, specific business rules, or strict tone requirements.

Go to Fin AI Agent > Ecommerce > Train > Guidance to create Fin for Ecommerce guidance for:

Communication style
Conversation strategy & flow
Situation handling & edge cases
Context and clarification rules

Screenshot of Fin AI Agent > Ecommerce > Train > Guidance showing the guidance creation interface with category options including communication style, conversation strategy, situation handling, and context rules.

Examples include:

How to explain complex product combinations.
Important restrictions or exclusions.
Brand tone of voice.
Rules for edge cases or special scenarios.

Add attributes (optional)

Fin automatically detects and applies attributes to ecommerce conversations, such as:

Product recommended
Cart created
Checkout link clicked

These help with reporting and analysis. To add more custom attributes, go to Fin AI Agent > Ecommerce > Train > Attributes.

Screenshot of Fin AI Agent > Ecommerce > Train > Attributes showing the ecommerce conversation attributes list, including Product recommended, Cart created, and Checkout link clicked.

Add escalation guidance and rules (optional)

Fin for Ecommerce is designed to resolve customer questions autonomously while escalating to a human teammate at the right moment.

You can control when escalation happens, whether Fin offers escalation or escalates immediately, and what Fin says during the handover from Fin AI Agent > Ecommerce > Train > Escalation.

Screenshot of Fin AI Agent > Ecommerce > Train > Escalation showing the escalation configuration interface with controls for when and how Fin escalates to a human teammate.

Fin escalation behavior is powered by:

Default escalation behavior
Escalation rules (data-driven)
Escalation guidance (natural language instructions)
Workflows (post-escalation routing and actions)

Best practice for escalation guidance and rules.

Manage Procedures

Fin for Ecommerce supports post-purchase workflows using Procedures and Shopify APIs.

Common use cases include:

Where is my order
Returns
Refunds
Exchanges
Order updates or cancellations

Go to Fin AI Agent > Ecommerce > Train > Procedures to review all Procedures. Some Procedures may be:

Automatically live if already configured.
Created in draft mode if new or incomplete.

**Note:** To activate a draft Procedure, open it from the Procedures list, review each step to confirm it is correctly configured, then click **Set live**. Once live, the Procedure runs automatically when triggered in a customer conversation.

Screenshot of Fin AI Agent > Ecommerce > Train > Procedures showing the list of Procedures with their current status — live or draft.

Learn more about Fin Procedures.

Step 4: Test Fin for Ecommerce

To test Fin for Ecommerce, go to Fin AI Agent > Ecommerce > Deploy and use the Preview panel to validate Fin's logic and messaging before deployment.

Screenshot of the Fin AI Agent > Ecommerce > Deploy page showing the Preview panel used to run sample conversations and test Fin's responses before going live.

You can:

Run sample shopping conversations.
Validate product recommendations and comparisons.
Test broad and vague discovery questions.
Check cart and checkout actions.
Review how Fin handles support requests.
Confirm how Fin handles shopping and support queries in the same conversation.

This is especially useful for businesses with complex catalogues, product rules, or post-purchase workflows.

Tip: Create a test Fin audience and apply it to specific ecommerce workflows and content so that you can set Fin live and test it in your live environment.

Step 5: Deploy Fin for Ecommerce

Go to Fin AI Agent > Ecommerce > Deploy to launch Fin for Ecommerce. Fin can be deployed via Workflows or Simple deploy.

Screenshot of Fin AI Agent > Ecommerce > Deploy showing the two deployment options: Workflows and Simple deploy.

When deploying via Workflows you can choose to use a pre-configured Workflow – or to configure from scratch.

Pre-configured deployment

Use a ready-made workflow to go live quickly.

This includes:

A Let Fin handle step configured for Ecommerce.
A greeting message.
Default trigger (when a customer opens a new Messenger conversation).
Default assignment if Fin cannot resolve.

Screenshot of the pre-configured Fin for Ecommerce workflow showing a 'Let Fin handle' step, a greeting message, a default trigger, and default assignment fallback.

Configure your deployment

Go to Fin AI Agent > Workflows.
Create or edit a workflow.
Choose a trigger, such as:
- When a customer opens a Messenger conversation
- When a customer visits a page
Select Messenger as the channel.
Set your audience.
Add a Let Fin handle step.
Set Fin’s role to Ecommerce.
Configure fallback actions (e.g., assign to inbox)

Screenshot of the Fin workflow editor showing a configured 'Let Fin handle' step with Fin's role set to Ecommerce and a fallback action to assign to an inbox.

Note: Fin automatically determines whether to provide shopping assistance or support. There is no need to define routing or handover to Fin for Service.

Tips: Before deploying, we recommend reviewing the following: