momo is in early access: help shape the open-source AI support desk. Get started
momo

AI customer support for Shopify stores

Shopify is in preview with momo. Store teams can use it to answer questions from their own content and send questions it cannot answer confidently to a shared human inbox; Shopify order lookup is not verified.

What connects

Shopify AI customer support: what connects

Shopify AI customer support uses a storefront chat widget and store information to answer common questions. With momo, Shopify connects store knowledge to the widget, where the AI can answer from that knowledge and route questions it cannot answer confidently to a human inbox. The Shopify order-status action helps with WISMO questions: "Where is my order?" See order-status actions for how an order-status action fits into the support workflow.

For example, a shopper asks, "Where is my order?" This WISMO question calls for an order-status action that retrieves the shopper's specific order information, rather than a general answer about delivery times. A useful AI agent guide for ecommerce explains how support teams can distinguish a customer-specific request from a content question.

The documented product workflow is different: the AI retrieves passages from the store's knowledge, drafts a reply and checks that draft against those sources. When confident, it can send the answer with citations. When unsure, it tells the visitor it is not sure and opens a ticket for the team. This is useful for questions the store has already documented, such as where to find delivery information or how a product is described.

Store knowledge and the widget

Knowledge can come from a website crawl, PDFs, Word files, plain text and Q&A pairs. Add the chat widget to the Shopify theme using the widget snippet and the installation instructions for your setup. The chat widget guide covers the widget's role on a site.

Keep the distinction clear in a WISMO flow: a general answer about delivery information is different from an order-status action that retrieves a particular order's status. Use content-grounded answers for documented questions and route anything the AI cannot answer confidently to a human.

Check the limits before launch

The documented workflow supports answers from a website crawl, PDFs, Word files, plain text and Q&A pairs. When a reply is confident, the AI can cite the knowledge passages it used; when the content does not support an answer, it can say it is unsure and open a ticket in the shared inbox. Test both outcomes with questions based on your own store material. Shopify order lookup is not verified, so do not treat a general delivery answer or a support ticket as confirmation of a shopper's specific order status.

Prepare and check the support workflow

Treat connecting Shopify and preparing support content as separate tasks. Follow the connection and theme instructions for your setup, then check that the widget answers from store knowledge and routes uncertain questions to the shared inbox.

  1. 1

    Gather the answers your team already uses

    Collect current product information, delivery guidance and other common support answers in the material your team maintains. Knowledge can be supplied as a website crawl, PDF, Word file, plain text or Q&A pair. Review the content for outdated details before relying on it.

    A clearer knowledge set for content questions.

  2. 2

    Use the verified preview instructions

    Follow the Shopify connection instructions available for your setup, including any required permissions. Use the provided steps rather than assuming that generic theme instructions apply.

    A connection path based on actual preview instructions.

  3. 3

    Treat the widget snippet as installation-specific

    Add the widget to your Shopify theme with the widget snippet and placement instructions provided for your setup. Check the storefront to confirm the widget appears.

    A widget setup that does not rely on guessed code.

Keep routine content questions in view

For a Shopify support team, distinguish an answer grounded in store material from a customer-specific action that needs verified order access. This helps set expectations for shoppers and gives agents a clear route for questions the available content cannot resolve.

Answers grounded in business content

The AI retrieves passages from the business's own knowledge, drafts a reply and checks it against those sources before responding. This works for documented questions such as delivery guidance or product details, not as proof of an individual order's status. Keep source material current and test representative questions so the team can spot gaps before shoppers rely on the widget.

Answers tied to store material.

Uncertainty becomes a ticket

When the AI is not confident, it tells the visitor it is not sure and opens a ticket in the shared helpdesk inbox for the team.

A route to a human for questions that need judgment.

Human replies can improve knowledge

A human answer can be saved as approved knowledge through the Teach step, helping the AI respond to similar questions from the store's approved information.

Useful answers can be kept for future support.

Plans for different support volumes and budgets

Plans use a flat fee with included AI conversations, not a per-resolution fee; the helpdesk inbox is included on every plan. Compare the included conversation volume with your expected demand before choosing.

Included: The shared helpdesk inbox is included on every plan, including Free, so the team can review tickets when the AI is unsure. · Pricing is in euros, with a flat plan fee and included AI conversations rather than per-resolution charges or credit packs. Check the plan details for the conversation allowance that fits your support volume.

Free

Try momo with real visitors.

€0 forever

  • 30 AI conversations, one-time
  • 1 agent, 1 team member
  • Website widget
  • Helpdesk inbox
  • Basic analytics
  • 2 MB of knowledge, 10 sources

Starter

One agent, steady traffic.

€29per month

  • 500 AI conversations / month
  • 1 agent, 2 team members
  • Website widget and email
  • Helpdesk inbox and Help Center
  • Basic analytics
  • No momo badge
  • 20 MB of knowledge, 50 sources

Growth

More agents, more channels.

€129per month

  • 2,000 AI conversations / month
  • 3 agents, 5 team members
  • Slack, Shopify, WhatsApp, Instagram, Messenger
  • API, Zapier and webhooks
  • Advanced analytics
  • 100 MB of knowledge, 200 sources

Scale

High volume, every channel.

€399per month

  • 6,000 AI conversations / month
  • 10 agents, 10 team members
  • Every channel
  • API, Zapier and webhooks
  • Advanced analytics
  • 500 MB of knowledge, 1,000 sources

Need more resources or custom solutions? Contact us for Enterprise plans

FAQ: Shopify AI support and order questions

Answers about what the knowledge workflow and human handoff establish today.

Can a Shopify customer service chatbot answer WISMO questions with an order-status action?

momo's verified product facts do not establish Shopify order lookup or an order-status action. It can answer from business knowledge and open a ticket when it is unsure, but that is different from retrieving a particular customer's order.

What should a customer provide to ask about an order?

No specific order information or verification steps are established for a Shopify order lookup. A shopper asking where an order is should be treated as an example of a WISMO question, not as a confirmed order-status workflow.

Can customers reach a human through the support widget?

When the AI is not confident, it tells the visitor it is not sure and opens a ticket for the team in the shared inbox. The team can review the ticket and respond to the customer.

Does Shopify have an AI agent for customer support?

The Shopify support workflow described here answers from business knowledge, cites sources when confident and hands questions the AI cannot answer confidently to a human. An order-status action is a specific way to handle WISMO questions.

How does the AI handle a Shopify question it cannot answer confidently?

It tells the visitor it is not sure and opens a ticket in the shared helpdesk inbox. A teammate can take over the conversation, and a human answer can be saved as approved knowledge through the Teach step.

Start with the questions your store can answer

Set up store knowledge, test common content questions and review uncertain cases in the shared inbox.