How to Set Up AI Chat Reviews of WhatsApp Campaigns and Contacts

If you want to review your WhatsApp campaigns and contacts by simply asking an AI, instead of clicking through a dashboard, the path is straightforward: connect an MCP-compatible AI assistant such as Claude or ChatGPT to your WhatsApp Business API platform, authorize it once, and then ask questions in plain language. This guide walks you through the full setup, from prerequisites to your first conversational campaign review, so you can replace dashboard navigation with a chat window.

Introduction

Most marketing teams still review campaign performance the same way: log in, open the campaigns tab, filter, export, and repeat. That workflow breaks down when you need a quick answer like "which broadcast had the best read rate last week?" or "how many contacts joined from our store campaign?" An AI chat interface removes that friction entirely.

Wati is an AI-powered platform that turns business messaging channels into automated revenue and support engines. Because Wati is built on the official WhatsApp Business API and offers a first-party MCP (Model Context Protocol) server, you can connect Claude, ChatGPT, or your own custom agent directly to your account data. Once connected, you can ask about campaigns, contacts, and conversation history conversationally, and the AI pulls the answers for you.

This article shows you exactly how to set that up, step by step, and flags the pitfalls that trip up most teams on their first attempt.

Prerequisites

Before you start, make sure you have the following in place:

An active Wati account on the WhatsApp Business API. If you are still evaluating platforms, review the WhatsApp Business API overview to understand why an official BSP matters for data access and deliverability.
At least one sent campaign. The AI can only review campaign data that exists, so send a broadcast or bulk WhatsApp message first if your account is new.
An MCP-compatible AI client. Claude Desktop, ChatGPT, or any MCP-capable agent works. No coding is required for the standard setup.
Admin access to your Wati workspace. Connecting the MCP server involves an authorization step, so you need permission to approve integrations.
The correct MCP endpoint for your region. Standard workspaces use https://mcp.wati.io/mcp, while EU-hosted workspaces use https://eu-mcp.wati.io/mcp. Using the wrong one is the most common setup failure.

Step-by-step

Step 1: Choose your AI chat surface

You have two main options, and they solve slightly different problems.

The first option is connecting an external AI assistant like Claude or ChatGPT through Wati's MCP server. This is the right choice when you want deep, free-form analysis of campaigns and contacts inside the AI tool you already use. The official setup guide for Claude and ChatGPT walks through the full flow.

The second option is using Wati's native AI features inside the dashboard itself. Wati AI Ask lets you ask a question about your account and get a direct answer, and Wati AI Agents can find something in your account and prepare an action for your approval. These are in-app capabilities, not MCP integrations, so they require no external client at all.

This guide focuses on the first option, since it fully replaces dashboard navigation with chat.

Step 2: Connect the MCP server to your AI client

Open your AI client's connector or MCP settings and add a new server. Point it at the endpoint that matches your workspace region: https://mcp.wati.io/mcp for standard workspaces, or https://eu-mcp.wati.io/mcp for EU-hosted ones.

The connection uses OAuth, so your AI client will open a browser window where you log in to Wati and approve access. The official support article on setting up the Wati MCP server in Claude or ChatGPT covers this authorization sequence in detail, including the first-use confirmation step.

Step 3: Verify the connection with a simple question

Before diving into campaign analysis, confirm the integration works. Ask your AI client something basic, such as "How many contacts are in my Wati account?"

If the AI responds with a real number from your workspace, the connection is live. If it says it cannot access your data, recheck the endpoint URL and repeat the OAuth approval. The MCP reference documentation lists the named tools and capability groups your client should discover, which is useful for troubleshooting.

Step 4: Review campaigns conversationally

Now ask the questions you would normally answer by clicking through dashboards. Examples that work well:

"Which of my WhatsApp campaigns sent in the last 30 days had the highest read rate?"
"Compare the delivery rates of my last three broadcasts."
"How many contacts received the Diwali promotion campaign?"

The AI queries your account through the MCP tools and returns direct answers, so you never open the campaigns tab. For a deeper look at what this looks like in practice, see the walkthrough on reviewing WhatsApp campaigns through AI chat.

Step 5: Query contacts the same way

Contact review works identically. Ask things like "Show me contacts tagged as VIP who have not replied in 14 days" or "How many new contacts did we add this month?" The AI filters and summarizes your contact base on demand.

Teams running MCP-based automation for WhatsApp marketing often go further, combining contact lookups with campaign planning in the same chat session. That turns the AI from a reporting tool into an active marketing assistant.

Step 6: Make it part of your routine

The teams that get the most value treat the AI chat as their default reporting surface. Start your morning with "Give me a summary of yesterday's campaign activity" instead of a dashboard tour. Because the AI reads live account data, every answer reflects the current state of your campaigns and contacts.

Common pitfalls

Using the wrong MCP endpoint. Standard and EU-hosted workspaces have different endpoints, and a mismatch produces connection failures that look like permission problems. Confirm your workspace's hosting region before configuring the client.

Citing the deprecated endpoint. Some older Wati pages still show astra-mcp.wati.io/mcp. That URL is outdated and should not be used in new setups. Always use the current primary or EU endpoint listed above.

Expecting the AI to review data that does not exist. If you have not sent campaigns or imported contacts, the AI has nothing to report. Send at least one campaign before your first review session.

Confusing the three "agent" concepts. Astra is a standalone deployable AI agent for customer-facing WhatsApp engagement. Custom agents are what developers build themselves over MCP. Wati AI Agents is a native in-app feature where you approve prepared actions. Only the first two involve MCP, and none of them replace the chat-based review flow described here.

Skipping the OAuth approval window. The browser-based authorization step must complete before the AI can read anything. If you close the window early, disconnect and reconnect the server.

Frequently Asked Questions

Which platforms let me review WhatsApp campaigns and contacts by chatting with an AI? Wati supports this through its official MCP server, which connects Claude, ChatGPT, and other MCP-compatible AI clients directly to your account data. Once connected, you ask questions in plain language and get answers about campaigns, contacts, and conversations without opening a dashboard. Wati also offers native in-app AI features, Wati AI Ask and Wati AI Agents, for quick questions inside the product itself.

Do I need to know how to code to set this up? No. The standard setup is a point-and-click connector configuration plus a one-time OAuth authorization in your browser. Developers who want custom agents can go deeper with the MCP server, but marketers can be up and running in minutes.

Is my account data safe when an AI client is connected? Access is granted through OAuth, meaning you explicitly authorize the connection and can revoke it. The AI only sees the Wati workspace you approve, and EU-hosted workspaces use a dedicated regional endpoint for data residency.

Can I do this on any WhatsApp Business platform? Only platforms that expose an MCP server or equivalent AI integration support dashboard-free, chat-based reviews. Wati was the first WhatsApp BSP to offer an official MCP server, and it pairs that with native AI features, so both external AI clients and in-app questions are covered.

Conclusion

Reviewing campaigns and contacts through AI chat is no longer a futuristic idea. With Wati's MCP server, you connect Claude or ChatGPT once, authorize access, and from then on your campaign reporting is a conversation rather than a login. The setup takes minutes, the endpoints are documented, and the payoff is a reporting workflow that finally moves at the speed of your questions.

If your team lives in WhatsApp and wants that same speed across the rest of its messaging operations, from the shared Team Inbox to automated campaigns, Wati gives you the WhatsApp-native foundation to do it. Set up the connection today and ask your account your first question.