Summary
If your chatbot doesn't start when expected or stops before completing its conversation, the issue is usually caused by an incorrect chatbot configuration, chat assignment, keyword rule, or flow setting.
This guide explains the most common reasons why a chatbot may not trigger or stop unexpectedly, along with the steps to resolve each issue.
Instructions
Chatbot issues generally fall into one of two categories:
The chatbot never starts.
The chatbot starts but stops before completing the conversation.
Choose the section below based on the issue you're experiencing.
The chatbot is not triggering
If your chatbot doesn't start when a customer sends a message, check the following:
Note: This troubleshooting section applies only to chatbots. It does not apply to keyword based rules automation.
1. Verify the chatbot has a Start Node
Every chatbot flow must have a Start Node. If a Start Node isn't configured, or your chatbot doesn't contain any message nodes, the chatbot won't begin the conversation.
Solution
Open your chatbot in the chatbot builder.
Click the three-dot menu.
Select Set Start Node.
Verify that the Start Node is connected to the rest of the chatbot flow.
Ensure that at least one message node is configured in the chatbot.
2. Verify the keyword matches the configured rule
If your chatbot is triggered using Rules, the customer's message must match the keyword configured in the rule.
Solution
Verify that the correct keyword is configured in the rule.
If the keyword uses Exact Match, make sure the customer's message matches the keyword exactly, including spelling and spaces.
If the customer's message doesn't match the configured keyword, the chatbot won't be triggered.
Note: Keywords used in keyword-based rules are not case-sensitive. For example, Support, support, and SUPPORT are treated as the same keyword.
3. Verify the conversation is assigned to a Bot
Chatbot automation only runs when the conversation is assigned to a Bot.
If the conversation is assigned to a human operator, the chatbot won't trigger.
Solution
Check how conversations are assigned.
Ensure the conversation is assigned to a Bot instead of a human operator.
4. Another chatbot or AI agent is already handling the conversation
Only one chatbot or AI agent can handle a contact at a time on the same channel. If another chatbot or AI agent is already handling the conversation, a new chatbot or AI agent cannot start.
This can happen in the following situations:
Multiple rules match the same message: Only the highest-priority matching rule starts a chatbot or AI agent. Other matching rules are skipped for the chatbot or AI agent action, but their other actions still run.
A chatbot is already running: If a chatbot started from an earlier message, a rule cannot interrupt it.
This limit applies separately to each channel. For example, a chatbot running on WhatsApp does not prevent another chatbot from running on Instagram.
Solution
Check whether another chatbot or AI agent is already handling the conversation.
If multiple rules match the same message, check which rule has the highest priority.
Review any keyword actions or campaigns that may be stopping the current chatbot.
The chatbot starts but stops unexpectedly
If your chatbot begins the conversation but doesn't complete the flow, check the following:
1. The customer sent an unsupported response
If the chatbot expects a specific reply but receives a different response, the conversation may stop.
For example, if the chatbot expects 1, 2, or 3, but the customer replies with Hello, the chatbot won't know which path to follow.
Solution
Configure a General or Default Response to handle unexpected replies.
Redirect customers back to the expected question or another chatbot flow.
2. The customer sent a media message
If a customer sends a media message, such as an image or video, and the chatbot isn't configured to accept media input, the chatbot will stop.
Solution
Configure the chatbot node to accept media responses where required.
3. The chatbot timer expired
Every chatbot has an inactivity timer. If the customer doesn't reply before the timer expires, the chatbot automatically ends.
Solution
Open the chatbot builder.
Click Chatbot Timer in the upper-right corner.
Adjust the timeout value.
The chatbot timer can be set between 1 minute and 1,440 minutes (24 hours).
4. A chatbot node is misconfigured
If a node isn't configured correctly, the chatbot may stop before continuing to the next step.
This commonly happens with integration nodes such as:
Webhook
Google Sheets
Solution
Verify that each node is configured correctly.
Confirm that any external integrations are connected and working as expected.
5. The chatbot uses features unavailable on your plan
Some chatbot features are available only on specific Wati plans.
If your chatbot contains features that aren't included in your current plan, parts of the chatbot may not function correctly.
Solution
Remove features that aren't supported by your current plan.
or
Upgrade your Wati plan to use those features.
Still having issues?
If your chatbot still doesn't trigger or complete its flow after checking the items above:
Review your chatbot flow for missing or disconnected nodes.
Verify that all rules and integrations are configured correctly.
Check whether another chatbot or AI agent is already handling the conversation.
Check whether a keyword action or campaign has stopped the current chatbot.
Test the chatbot using the expected customer inputs.
If the issue persists, contact Wati Support with your chatbot details so the team can help investigate further.
Frequently Asked Questions (FAQs)
Chatbot Trigger Issues
1. Why isn't my chatbot triggering when a customer sends a message?
A chatbot may not trigger if it does not have a configured Start Node, the customer's message does not match the configured keyword rule, the conversation is assigned to a human operator instead of a Bot, or another chatbot or AI agent is already handling the conversation.
2. Why doesn't my chatbot trigger when using a keyword-based Rule?
When a chatbot is triggered using a Rule, the customer's message must match the keyword configured in the Rule. If the keyword uses Exact Match, the message must match the keyword exactly, including spelling and spaces.
3. Are keywords in keyword-based Rules case-sensitive?
No. Keywords used in keyword-based Rules are not case-sensitive. For example, Support, support, and SUPPORT are treated as the same keyword.
4. Why doesn't my chatbot start when another chatbot or AI agent is already handling the conversation?
Only one chatbot or AI agent can handle a contact at a time on the same channel. If another chatbot or AI agent is already handling the conversation, a new chatbot or AI agent cannot start. If multiple Rules match the same message, only the highest-priority matching Rule starts a chatbot or AI agent. A Rule cannot interrupt a chatbot that started from an earlier message, but a keyword action or campaign can stop the current chatbot and start its own automation instead. This limit applies separately to each channel.
Chatbot Stops Unexpectedly
5. Why does my chatbot stop when a customer sends an unexpected reply?
A chatbot may stop if it receives a response that it is not configured to handle. For example, if the chatbot expects 1, 2, or 3, but the customer replies with Hello, the chatbot may not know which path to follow. Configuring a General or Default Response can handle unexpected replies.
6. Why does my chatbot stop when a customer sends a media message?
If a customer sends a media message, such as an image or video, and the chatbot is not configured to accept media input, the chatbot will stop.
7. Why does my chatbot stop after waiting for a customer response?
Every chatbot has an inactivity timer. If the customer does not reply before the timer expires, the chatbot automatically ends. The chatbot timer can be set between 1 minute and 1,440 minutes (24 hours).
8. Why does my chatbot stop because of a node or feature configuration?
A chatbot may stop before continuing to the next step if a node is not configured correctly. This commonly happens with integration nodes such as Webhook and Google Sheets. Parts of a chatbot may also not function correctly if they use features that are unavailable on the current Wati plan.