Skip to main content
The same two questions as BaseLinker answers, for shops that keep everything in Shopify. The tools are the same, so the agent behaves identically; only the shop it asks changes.

Setting it up

  1. In Shopify admin go to Settings → Apps and sales channels → Develop apps and create a custom app.
  2. Under Configuration → Admin API grant read_orders, read_customers and read_products.
  3. Install the app and copy the Admin API access token (it starts with shpat_).
  4. In the app: Integrations → Shopify → paste the shop address (shop.myshopify.com) and the token.
  5. On the agent: Integrations tab → Order status (Shopify) and Availability and prices (Shopify) → pick the connection.
Building a customer support agent in the wizard, switching on Check orders and availability in Shopify and pasting the shop address and token is enough: the wizard creates the connection and turns both tools on.
This is a custom app in your own admin, not an app from the Shopify App Store. A shop connects in two minutes, the token is yours, and you can revoke it whenever you like.
Press Test connection after pasting both. It reads your shop’s name back, so a wrong address or a rejected token fails there rather than in front of a customer.

Who is allowed to hear what

The rule is the same as for BaseLinker. On an inbound call the agent knows the caller’s number, finds them in Shopify and reads their orders without asking for anything. An order found by its number alone needs confirming: the agent asks for the e-mail address or phone number the order was placed with. A customer says “order 1042” where Shopify says “#1042”, and the agent looks both up. Names, addresses and e-mail addresses are never returned to the agent. An order that is only partly paid, a deposit for instance, is not reported as paid - otherwise the customer calls back.