Back

Zapier MCP + GoodBarber: drive your app with an AI agent

on 

What if some of the tasks involved in managing your GoodBarber project could be carried out simply by talking to an artificial intelligence?

That's what Model Context Protocol (MCP) is all about: an exchange protocol that gives conversational agents like Claude or ChatGPT the ability to act directly on your app, by interpreting natural language instructions.

Zapier offers a turnkey solution, called Zapier MCP , that makes this technology immediately accessible. No need to deploy a complex infrastructure. Thanks to Zapier MCP, you can take advantage of this power in just a few clicks.

In this article, you'll discover how to send a push notification, or update your eCommerce app's order tracking URLs en masse with a simple instruction to Claude .

Welcome to a new kind of no-code integration, designed to boost your productivity.

GoodBarber + Zapier, it already exists

Before we go any further, a quick reminder: GoodBarber already offers two official integrations with Zapier: These connectors make it possible to integrate a GoodBarber app with thousands of other services, without writing a single line of code.
=> To find out how Zapier works in general, read this article .
=> And for a concrete example on the content app side, this one shows how to automatically add a user to a group via Zapier.

These integrations are generally used in scenarios that you've designed from start to finish: a command is placed → an event is triggered (Trigger) → a Zap is executed (Action). Simple and effective.

With Zapier MCP, the logic changes: the human becomes the trigger, and the interface... is natural language. No need to create a Zap manually: a simple instruction given to your AI assistant can be enough to launch the desired action.

Test #1 - Send a push notification via Claude

To test how Zapier MCP works with GoodBarber, we started with a simple case: sending a push notification via a conversational assistant.

To carry out this test, you'll need :
  • A Claude account with the option to connect to an MCP server (available with the Pro package),
  • An MCP server created from Zapier MCP. This is activated in just a few clicks from your Zapier account,
  • GoodBarber eCommerce integration activated in your Zapier MCP server (including a "Send push notification" action).
Once this configuration is in place, simply ask Claude:
"Send a push notification to my users to tell them that the sales end tonight at midnight. "

Claude recognizes the intention, identifies the action available via Zapier thanks to the MCP protocol, and then executes the sending of the notification through the GoodBarber eCommerce integration.

This simple test validates the overall operation of the chain:
Claude → Zapier MCP → GoodBarber.

It's a clear example of the value of the MCP protocol: an AI agent understands an instruction, detects a compatible action via the MCP server, then executes it.

That said, in the context of GoodBarber, this use case remains limited, as push notifications can already be sent very simply via the My GoodBarber app, available on mobile.

Let's now look at a much more relevant use case for a merchant.

Test #2 - Updating tracking URLs from an Excel file

Our second test illustrates a use case that is far more relevant to the daily life of a merchant using GoodBarber eCommerce: the mass updating of order tracking URLs.
 
When you prepare your orders for shipment, you need to obtain tracking links from your carrier (Colissimo, UPS, DHL, etc.). Once these links have been obtained, each order must be manually updated in the back-office, so that your customer can consult the tracking directly in your application.
 
This operation can quickly become tedious - and a source of errors - when several dozen or even hundreds of parcels have to be processed.
 
To explore solutions to this problem, we prepared the following test:
  • An Excel file containing two columns:
    • the IDs of the orders to be updated,
    • the corresponding tracking URLs.
  • A Claude account connected to Zapier MCP
  • The "Update tracking URL" action activated in the GoodBarber eCommerce integration on the Zapier side.
 
Once the file was loaded into Claude, we simply gave him the following instruction:
"From this table, update each order with the corresponding URL."
 
Claude understood the structure of the table, identified the "Update tracking URL" action available via Zapier MCP, and used it to update each order, one by one.
 
All orders were updated correctly - without writing a single line of code, without using an API, and without having to manually open each order form in the GoodBarber back-office.

A real gain in productivity... and reliability.

A new way to manage your app

What these tests show is that it's now possible to interact with your GoodBarber project in a totally new way: in natural language, via an AI assistant, without needing to master APIs or manually build automations.
 
Thanks to the Model Context Protocol and out-of-the-box solutions like Zapier MCP, a conversational assistant like Claude can execute actions in your app by relying on existing GoodBarber integrations.
 
Today, this makes it possible, for example, to :
  • update information in your back office from a simple Excel spreadsheet,
  • execute targeted actions without opening the administration interface,
  • save precious time on repetitive tasks, with less risk of error.
 
But that's just the beginning.
 
This type of integration opens the way to new uses in the day-to-day management of your app: a personal eCommerce assistant, capable of taking action on your behalf, simply from a conversational exchange.
 
Already using Zapier MCP with GoodBarber? Do you have ideas for scenarios you'd like to automate with an AI assistant?
Tell us about them in the comments - we look forward to discussing them with you.