System Messaging

What is System Messaging?

System messaging is where you build out the roadmap for your users' chatbot conversations.  

How does System Messaging work?

How it works: All users will be presented with the Welcome Message when they first open a chatbot (or restart the chatbot)

When a user asks the concierge chatbot a question: 

  1. the chatbot will first look through the knowledge base dialogues (exchanges/FAQs and guided conversations) to see if there's a direct match already created for the question.  
  2. If it finds the answer, it will present it to the user, and then present a thumbs up or down.  If the thumbs down is selected, the feedback message will present itself.
  3. When the chatbot doesn't find the EXACT match to the question:
    1. Optional: a Clarifier message provides "Did you mean?" options to the user
    2. It looks for Fallback solutions (i.e. searches your documents within the Knowledge Base, Sites & Articles, websites, etc.) to provide an answer
  4. If the chatbot can't find an exact match to a question or isn't confident in its answer when searching documents, articles, and sites, the chatbot leverages the No-Match Message flow (create a Helpdesk (HD) ticket, talk with a Live Agent, send an email, etc.) to escalate the user's question to the HD team

What are the types of System Messages?


Welcome Message

The Welcome Message is triggered when the chat bot opens for the first time in a session. You can develop this conversation to:

  • Welcome the user and preview the type of information available in the chat bot
  • Invite the user to see a response to a frequently asked question
  • Let the user file a ticket, start a live chat, or search the cloud drive
  1. Navigate to Knowledge Base.
  2. Click System Messaging under Knowledge Library.
  3. Select Welcome Message.
  4. Click the concierge you want to edit.
  5. Click Edit Conversation in the top right corner.
  6. Build the Guided Conversation, and click Save in the top right corner.
Third Party Interfaces (Slack, Teams, etc.) do not allow for the use of a Welcome Message.

Fallback & No-Match Handling

Fallback, No- Match, and Clarifier messages allow you to control how our system will handle questions that do not directly match an FAQ (dialogue/exchange) in the Knowledge Base. 

Clarifier Match Settings

The clarifier match settings allow you to configure thresholds of question/answer match confidence for the chatbot.  The confidence percentages are indicators of the system's confidence in its suggestions (matching or understanding the user's intent) based off of the question.  

Using the clarifier prompt, you're able to add an additional layer of clarification to the user if the chatbot's confidence in it's answer is within the thresholds. 

Clarifier On:

If you toggle on the "Enable Clarifier" button, and a user asks a question that falls in between the "Clarify Above" and "And Below" match percentages, the chatbot will display a "Did you mean?" prompt.  Under the prompt, there will be a display of up to three suggestions to the user for matches based off of their inquiry.  If the user selects "None of the above", then our system will move onto the Fallback System Message.  

"Did you mean" is the default message, but this can be edited within the clarifier section of your General Settings.  

Clarifier Off:

The Clarifier does not need to be enabled to adjust the match threshold.  If clarifier is disabled and the Confidence percentage falls below the "And below" match threshold, the question will automatically be routed to the Fallback System Message flow


Fallback Message

The Fallback Message is triggered when our system is not able to make a confident direct match or one through our clarifier. Prior to going to the No Match Message, this system message is designed to offer fallback solutions to providing an answer.

This conversation is automatically configured to:

  • Detect if your organization has AI Tools enabled or not to determine if it can use Generative AI search and LLMs to identify an answer.
  • If AI is enabled, but the bot is unauthenticated it can fallback to a web search of the URL specified in the Fallback URL field. Relevant search results can then be provided to the user.
    • For optimal results for AI-enabled organizations that have unauthenticated bots, it is recommended that you supply a Fallback URL. It must be a publicly accessible URL.
  • If no fallback answers can be identified, then it will route to the No Match Message.


  1. Navigate to Knowledge Base.
  2. Click System Messaging under Knowledge Library.
  3. Select Fallback Message.
  4. Click the concierge you want to edit.
  5. Click Edit Conversation in the top right corner.
  6. Build the Guided Conversation, and click Save in the top right corner.


No-Match Message

The No-Match Message is triggered when the chat bot doesn’t know the answer to a question. You can develop this conversation to:

  • Ask the chat user to rephrase the question
  • Search Capacity Cloud Drive
  • Create a Helpdesk ticket
  • Start a Live Chat
  • And much more...


  1. Navigate to Knowledge Base.
  2. Click System Messaging under Knowledge Library.
  3. Select No-Match Message.
  4. Click the concierge you want to edit.
  5. Click Edit Conversation in the top right corner.
  6. Build the Guided Conversation, and click Save in the top right corner.

Feedback Message

The Feedback Message is triggered when the user gives the response from the chat bot negative feedback (the thumbs down icon). You can develop this conversation to:

  • Ask the chat user to rephrase the question
  • Search Capacity Cloud Drive
  • Create a Helpdesk ticket
  • Start a Live Chat
  • And much more...
  1. Navigate to Knowledge Base.
  2. Click System Messaging under Knowledge Library.
  3. Select Feedback Message.
  4. Click the concierge you want to edit.
  5. Click Edit Conversation in the top right corner.
  6. Build the Guided Conversation, and click Save in the top right corner.

Live Chat Fallback

The Offline Chat Message is triggered when the a live chat agent is not available to connect to the user. You can develop this conversation to:

  • Ask the chat user to try again later
  • Search Capacity Cloud Drive
  • Create a Helpdesk ticket
  • And much more...
  1. Navigate to Knowledge Base.
  2. Click System Messaging under Knowledge Library.
  3. Select Offline Chat Message.
  4. Click the concierge you want to edit.
  5. Click Edit Conversation in the top right corner.
  6. Build the Guided Conversation, and click Save in the top right corner.

Was this article helpful?