Shopify Chatbot: Add AI Support to Your Store

Shopify stores deal with the same questions on repeat. What is the shipping time? Does this come in a different size? What is your return policy? Can I track my order? An AI chatbot trained on your store's content answers these questions instantly, around the clock, without adding support headcount.

This guide covers what an AI chatbot does for a Shopify store, how to train it on your products and policies, how to install it in your theme, and how to configure it for the best results.

What a Shopify chatbot handles

An AI chatbot on your Shopify store is not a scripted FAQ bot with canned responses. It is trained on your actual content, meaning it can have real conversations about your products, policies, and brand. Here is what it covers without any custom programming.

The chatbot answers based on what it has been trained on. If a question falls outside its training data, it acknowledges that and can offer to collect the visitor's contact information so your team can follow up.

Why AI support matters for ecommerce

Customer support in ecommerce has specific characteristics that make chatbots particularly effective.

Volume and repetition. Most ecommerce support queries are repetitive. The same 20 questions account for 80% of incoming volume. A chatbot handles those so your team focuses on complex issues.

Timing matters. Shoppers browse evenings and weekends. If they have a question about sizing at 10 PM and nobody is there to answer, they leave. A chatbot is always available.

Pre-purchase influence. A shopper asking "Does this run true to size?" is close to buying. A quick, accurate answer can be the difference between a sale and an abandoned cart.

Post-purchase cost reduction. "Where is my order?" is the most common post-purchase question. A chatbot that points visitors to order tracking saves your team significant time.

Training your chatbot for Shopify

The quality of your chatbot depends entirely on what you train it on. For a Shopify store, focus on these sources.

Recommended training sources

Step 1: Train on your store URL

1. In the Agentkit dashboard, go to your chatbot's Sources.
2. Add your Shopify store URL (e.g., https://yourstore.myshopify.com or your custom domain).
3. The crawler indexes your product pages, collection pages, policy pages, and any other publicly accessible content.
4. For stores with JavaScript-rendered content, the crawler uses Firecrawl to render pages before indexing.

The crawler follows links from your main URL, so it will pick up product pages, collection pages, and navigation-linked pages automatically.

For detailed crawling guidance, see the website sources documentation.

Step 2: Add Q&A pairs for common questions

Some questions deserve precise, carefully crafted answers. Q&A pairs let you control exactly how the chatbot responds to specific topics.

Create Q&A pairs for:

• Shipping times by region. "How long does shipping take?" should give specific timeframes.
• Return process. Walk through the exact steps, not just the policy.
• Contact methods. "How do I reach support?" should list email, hours, and expected response time.
• Current promotions. "Do you have any discounts?" should mention active sales or code eligibility.
• Out-of-stock items. "When will X be back in stock?" should explain your restock process.

Example Q&A pair:

Question: What is your return policy?

Answer: You can return any unworn item within 30 days of delivery for a full refund. Start a return by visiting yourstore.com/returns and entering your order number. We will email you a prepaid shipping label. Refunds are processed within 5 business days after we receive the item. Sale items can be exchanged but not refunded.

Q&A pairs take priority over crawled content, so they are your tool for ensuring the chatbot gives the exact answer you want for critical topics.

Step 3: Upload documents

If you have size guides, product specification sheets, or care instructions in PDF or DOCX format, upload them as document sources. This is particularly useful for:

• Detailed size charts that are formatted as downloadable PDFs on your site
• Product catalogs with specifications not fully listed on product pages
• Brand guidelines that inform the tone of chatbot responses

See the document sources documentation for file type and size limits.

Step 4: Test in the playground

After adding your training sources, use the Agentkit playground to test the chatbot before putting it on your live store.

Test with real questions your customers ask:

• Product-specific questions across different collections
• Policy questions (shipping, returns, exchanges)
• Edge cases ("Can I combine two discount codes?")
• Questions the chatbot should not try to answer ("What's the status of order #12345?" if you have not integrated order lookup)

Refine your training based on the results. Add Q&A pairs for any topics where the response is not good enough.

Installing on Shopify

Once your chatbot is trained and tested, add it to your Shopify store.

Step 1: Get your embed code

In the Agentkit dashboard, go to your chatbot's Settings then Embed. Copy the script tag:

<script
  src="https://cdn.agentkit.ai/widget.js"
  data-chatbot="your-chatbot-id"
  async>
</script>

Step 2: Add to theme.liquid

1. In your Shopify admin, go to Online Store then Themes.
2. Click Actions (or the three dots) on your current theme, then Edit code.
3. In the left sidebar, find and click theme.liquid under the Layout section.
4. Scroll to the bottom of the file and find the closing </body> tag.
5. Paste the embed script immediately before </body>:

    <!-- Agentkit Chatbot -->
    <script
      src="https://cdn.agentkit.ai/widget.js"
      data-chatbot="your-chatbot-id"
      async>
    </script>
  </body>
</html>

6. Click Save.

The chatbot is now live on every page of your Shopify store.

Step 3: Verify

1. Visit your store in a new browser tab or incognito window.
2. Confirm the chat bubble appears.
3. Send a test message and verify the response.
4. Check on mobile to confirm it displays correctly.

If the chatbot does not appear immediately, wait a minute for Shopify's CDN to propagate the theme change, then hard refresh the page.

Shopify-specific configuration

After installing, configure these settings in the Agentkit dashboard for the best experience on your store.

Domain restriction

Go to your chatbot's Settings and set the allowed domain to your Shopify store's domain. This ensures your chatbot only works on your store and cannot be embedded on unauthorized sites.

If you use a custom domain (e.g., yourstore.com) in addition to the default Shopify domain (yourstore.myshopify.com), add both to the allowed list.

Custom instructions for ecommerce

In your chatbot's settings, add custom instructions that guide the AI's behavior for ecommerce conversations. Examples:

• "Always recommend checking the size guide before purchasing clothing items."
• "If a customer asks about order tracking, direct them to the order tracking page at yourstore.com/account."
• "Never make promises about restock dates. Say that the customer can sign up for email notifications."
• "If a customer seems ready to buy, mention current free shipping threshold ($75)."
• "For questions about ingredients or allergens, provide information from training data only. Do not speculate."

These instructions shape how the chatbot behaves within the context of your training data.

Rate limiting

If you are on a free or Hobby plan and want to manage your message quota carefully, enable rate limiting. This sets a maximum number of messages per visitor per session, preventing any single conversation from consuming too many messages.

A reasonable limit for ecommerce is 15-25 messages per session. Most product questions resolve within 5-10 messages.

Welcome message

Set a welcome message that aligns with your store's purpose. Instead of a generic "How can I help you?", try something specific:

• "Hi! Ask me about our products, sizing, shipping, or returns."
• "Welcome! I can help you find the right product or answer questions about your order."

A specific welcome message sets expectations and encourages visitors to ask the kinds of questions the chatbot is trained to handle well.

Lead capture for Shopify

One of the most valuable chatbot features for ecommerce is lead capture. When a visitor engages with the chatbot, you can collect their email for future marketing.

Setting up lead capture

1. In the Agentkit dashboard, go to your chatbot's Actions.
2. Enable the Lead Capture action.
3. Configure which fields to collect (name, email, phone).
4. Set when the form appears (after a certain number of messages, or triggered by the AI when appropriate).

For setup details, see the lead capture documentation.

Sending leads to your marketing tools

Once you are capturing leads, connect them to your marketing stack:

Zapier is the easiest path for connecting leads to your email marketing. See the Zapier integration documentation for setup.

Measuring chatbot impact

After the chatbot has been running for a few weeks, evaluate its performance.

Metrics to track

Using conversation data to improve your store

The questions people ask your chatbot reveal what your store content is missing.

• If visitors frequently ask about sizing, your size guide needs more detail or better placement.
• If shipping questions dominate, your shipping information page may be hard to find.
• If people ask whether products are in stock, consider adding stock status more prominently.

The chatbot becomes a continuous feedback loop that improves your store beyond just answering questions.

Keeping the chatbot up to date

Shopify stores change constantly. New products, seasonal promotions, updated policies. Your chatbot needs to keep pace.

On the Free or Hobby plan, retrain your chatbot after significant changes: new collection launches, policy updates, seasonal promotions, or major product additions.

On the Standard or Pro plan, auto-retrain handles this for you. The chatbot periodically re-crawls your site and updates its knowledge base automatically.

For time-sensitive changes (flash sales, holiday policies), add a Q&A pair or text snippet immediately. These take effect faster than a full re-crawl.

Getting started

Setting up a chatbot on your Shopify store takes under 30 minutes:

1. Sign up for Agentkit (free).
2. Create a chatbot and train it on your store URL plus key Q&A pairs.
3. Test in the playground until you are satisfied with the responses.
4. Add the embed script to your theme.liquid file.
5. Configure domain restriction, custom instructions, and lead capture.
6. Monitor conversations and refine training over the first few weeks.

Your store gets 24/7 AI support that handles the repetitive questions, captures leads, and helps visitors find what they need. Your team gets to focus on complex issues and growth.

Build your chatbot for free →

No credit card required.