Campaign Bot: Chatbot in a WhatsApp Campaign

TL;DR

A good WhatsApp marketing campaign creates an immediate flood of replies. If 500 customers respond to an offer at the same time, your team breaks. The solution is to connect the campaign to a chatbot that handles the first steps automatically and passes only certain cases to a human agent. This article covers the concept, how to set up a Campaign Bot, the common patterns, and the advantages over sending a regular campaign.

What a Campaign Bot is

In a regular campaign, you send a message template to a list of customers. If a customer taps a button or replies, the conversation lands in the inbox and an agent has to handle it. If 1,000 customers reply, that’s 1,000 conversations to handle manually.

With a Campaign Bot, the customer’s first reply is automatically routed to a chatbot that gives a relevant response. The chatbot can:

Answer common questions expected to follow the campaign.
Collect information from the customer (name, business type, need).
Classify the lead (hot / cold / not relevant).
Pass only qualified customers to an agent.

How it works technically

The pairing is done through two mechanisms that already exist in Mumble:

A campaign with a post-send action. When you create a campaign, you can set what happens after the campaign is sent. One of the options is triggering a chatbot.
A chatbot with the right trigger. The chatbot is set up with a trigger that fires when a customer replies to the campaign.

Note: the pairing is between two separate features, not a single feature called “Campaign Bot.” Think of it as a usage pattern, not a single button in the interface. See dynamic variables and post-send actions and what triggers the chatbot.

Setup steps

1. Build the chatbot

Create a new chatbot in the Mumble Chatbot Builder. Plan the flow around the specific campaign you’ll send:

An opening message that acknowledges what the customer tapped on.
2-3 buttons for initial classification (for example: “I’m interested,” “I’d like more details,” “Not relevant right now”).
Different flows for each choice.
A clear path to a human agent at the right points.

2. Approve the campaign’s message template

The campaign must use a message template that includes buttons (Quick Reply) or text that invites a reply. That way, when the customer taps or writes, the chatbot can catch the reply. See creating a message template.

3. Set up a post-send action

When you create the campaign, in the “post-send actions” step, choose:

Assign a bot: the chatbot you created in step 1.

This means every customer who receives the campaign is attached to the bot. The moment they reply, the bot starts working.

4. Test with a small list

Test the campaign on an internal test number or on a small opt-in list before a wide send. Make sure the replies are handled as expected in the bot.

5. Send to the full audience

After a clean test, send the campaign to the full list. As the campaign goes out, the bot receives the customers who reply and handles them automatically.

Common Campaign Bot patterns

Promotional campaign with lead filtering

Campaign: a message template with an offer and two buttons: “Interested” and “Not today.”
Bot: whoever taps “Interested” gets qualifying questions (what’s the business, what’s the budget). After 3 questions, they’re classified as a hot lead and passed to a sales agent.
Result: a sales agent gets only qualified leads, not everyone who tapped.

Survey campaign with automatic tagging

Campaign: a single question with 3 buttons (for example, “How big is your organization?”: “1-10,” “11-50,” “51+”).
Bot: each tap tags the customer with the matching list, and sometimes also sends a relevant message.
Result: your list suddenly has useful segmentation data, without a single person having to tag anything.

Appointment confirmation campaign with automatic recording

Campaign: an appointment reminder with buttons: “I’ll be there,” “Cancel.”
Bot: “I’ll be there” tags as confirmed. “Cancel” offers 3 alternatives and fires a webhook to your calendar system.
Result: fewer no-show appointments, less manual work.

Remarketing campaign with a re-onboarding flow

Campaign: reaching out to customers who haven’t bought recently (“We’re back with an offer”).
Bot: whoever taps enters a re-onboarding flow that includes an explanation of changes to the product.
Result: customers get a personal experience, not just “head to the website.”

Why it’s worth using

Significant savings in staffing. A campaign of 1,000 recipients with a 20% reply rate produces 200 conversations. A team that can handle 50 conversations a day is overwhelmed. A bot that finishes 60% of the conversations on its own leaves only 80 for an agent.
Short response time. A bot replies within a second. A customer who gets a fast response stays warm. A customer who waits an hour for a reply has already left the chat.
Organized information. A bot collects structured information (name, business type, budget) instead of an agent having to decode a free-form conversation.
Automatic classification. Instead of an agent reading 100 conversations and deciding who’s a hot lead, the bot does the classification during the conversation.

When not to use it

A small, very personal audience. If the campaign is to 20 VIP customers, a human agent is better. The personal touch matters.
Very complex content. Products that need long explanations, or complex B2B services, don’t always fit a bot.
No resources to maintain the bot. A bot that doesn’t get updated goes stale. If there’s no one to maintain it, a consistent human agent is better.

Best practices

Always give a way to reach a human agent. Every bot flow must have a “Talk to an agent” button that works immediately. Customers who feel stuck with a bot leave.
Keep the bot short. 3-5 nodes, not 20. The longer the flow, the more customers drop off in the middle.
Capture information in a structured way. Use “save reply to a variable” to store the information you collect. Otherwise the information is lost.
Watch the first campaign closely. After the first campaign, go through the conversations manually and see where the bot fell short. Improve and update.
Don’t send the same campaign over and over. If 50% of customers didn’t reply to the first campaign, the content probably needs to change, not the cadence.
Bring in Dolores AI at points that require judgment. A regular bot is right for collecting structured information. Dolores is suited to free-form questions that require understanding. Combine them. See Dolores AI, the getting-started guide.

Common issues

I sent a campaign with a bot assigned, but the bot isn’t triggering

Check:

The bot is in an active state (the Enable AI / Enable bot toggle).
The bot’s trigger matches the customer’s reply.
The customer actually replied to the campaign (if they only received it and didn’t tap, there’s nothing for the bot to do).

The bot replies but doesn’t pass to a human agent

Check the nodes on the canvas. There needs to be an “agent” node with a specific agent or team selected. Without that node, the bot will offer replies but won’t hand off.

I got a lot of customers into the bot, but the agents don’t see them

A bot handles a conversation on its own and doesn’t always surface it to the front of the main team. If you want the conversations the bot is handling to be visible too, make sure their status is set appropriately (for example, “In bot handling”) and that there’s an agent or manager watching the conversations in that status.

How do I know which customers were passed to a human?

On the Campaigns page, in the responding-customers tab, you can filter by status. Customers the bot passed to an agent will have a status of “In progress” or “Pending.”

Bottom line

A Campaign Bot is the pattern that makes the difference between a campaign that creates a workload and a campaign that creates leads. The setup requires connecting two existing features: a campaign with a post-send action, and a chatbot that handles the replies. An hour of planning ahead saves days of manual handling. Always make sure there’s a clear path to a human agent, and improve the bot after each campaign based on the conversations you had.

Want to Use WhastApp API?

Campaign Bot: Pairing a Chatbot with a WhatsApp Campaign to Handle Replies Automatically

Mumble Team