How to Prevent Lost Conversations in Zendesk Sunshine Conversations
Learn how to capture and track Sunshine Conversations in Zendesk when customers don't interact with bot buttons. Step-by-step solutions for Zendesk admins to prevent lost revenue and conversations.
If you're using Zendesk's Sunshine Conversations with a custom bot, you've probably encountered a frustrating problem: customers start a conversation, your bot presents them with button options, and then... nothing. They close the chat without clicking anything. These conversations essentially disappear into the void, invisible to your support agents and impossible to follow up on.
This isn't just an operational headache—it's a potential revenue leak. Every lost conversation represents a customer who reached out for help but never got it. Let's fix that.
The Core Problem
When a customer sends their first message in Sunshine Conversations, your custom Zendesk bot might reply with options like "Get a Price Quote" or "Technical Support." If customers don't click a button—maybe they got distracted, closed the chat, or just didn't want to use the buttons—these conversations become completely invisible to agents.
As one Zendesk user put it: "These conversations at in Sunshine Conversations only, completely invisible to agents and lost revenue."
The root cause? Unlike standard Zendesk tickets, Sunshine Conversations doesn't automatically generate "Ticket" type conversations that agents can see in their usual workflow. The data exists, but accessing it requires jumping through API hoops that aren't practical for daily operations.
Solution 1: Enable AI Agent Tickets (Official Zendesk Feature)
The most straightforward solution is to enable Zendesk's AI Agent Tickets feature. This is an official Zendesk capability that creates visible, trackable tickets for conversations handled entirely by your AI agent.
How AI Agent Tickets Work
AI agent tickets are created when a customer begins a messaging conversation with an AI agent or in the Web Widget. These tickets remain read-only as long as the conversation isn't escalated to a human agent. When escalated, the ticket becomes a regular, editable ticket that agents can work with normally.
Benefits of AI Agent Tickets
- Full visibility: All bot conversations appear as tickets in Agent Workspace
- Conversation audit trail: Read full conversations between your AI agent and customers
- Automated resolution tracking: Examine solved tickets to confirm interactions are being resolved appropriately
- Customer context: AI agent tickets appear in customer interaction history, giving agents more context for future engagements
How to Enable AI Agent Tickets
To turn on AI agent tickets, navigate to Admin Center, click Channels in the sidebar, then select Messaging and social > Messaging. Click Manage settings, expand the AI agent conversations as tickets in Agent Workspace section, select Turn on AI agent conversations as tickets, and click Save settings.
Important Considerations
Keep in mind that AI agent tickets don't appear in existing lists, views, or searches by default to prevent disruption. You'll need to update or create new views to include them.
To search for AI agent tickets in Agent Workspace:
- Perform an advanced search
- Select the Tickets filter
- Select Support type
- Select AI agent
Source: Zendesk Help Center - Understanding and viewing AI agent tickets
Solution 2: Create a Trigger to Auto-Generate Tickets
If you need more immediate control over ticket creation, you can set up a trigger that automatically creates a ticket when a conversation starts.
The Timeout Approach
One effective method is to create a custom integration with a timeout fallback. Here's how it works:
The Process:
- When a customer starts a conversation, wait 30-60 seconds
- If the customer doesn't click a button within that timeframe, automatically create the ticket
- This ensures no conversation falls through the cracks
According to community feedback, this approach requires custom middleware code. You'll need to set up a custom webhook that catches the Sunshine webhook and creates the ticket programmatically. While this can be done with Python or similar languages, it cannot be configured entirely within Zendesk's platform interface.
The Challenge: As one user noted, "there is no way to do inside of platform" for this middleware approach.
Creating a Basic Trigger
Ticket triggers are business rules that run immediately after a ticket is created or updated. They automatically perform actions if specified conditions are met. For example, a ticket trigger can notify customers when a ticket has been opened.
To create a trigger that fires when conversations start:
In Admin Center, click Objects and rules in the sidebar, then select Business rules > Messaging triggers. Click Create trigger. Use the Run trigger dropdown to determine the event that initiates the trigger. Select "When a customer requests a conversation" to initiate the trigger when the conversation is passed to an agent.
Source: Zendesk Help Center - Creating ticket triggers and Creating messaging triggers
Pro Tip: When building triggers, make your conditions as specific as possible. Use the "Ticket | Is | Created" condition to ensure the trigger only runs when a ticket is created, not every time it's updated.
Solution 3: Webhook Integration for Conversation Logging
For teams with development resources, implementing a webhook integration provides the most flexibility and control.
How Webhook Integration Works
The integration links your business with Sunshine Conversations and the channels you want to connect to. This includes creating a webhook and selecting the conversation events that trigger it.
A webhook can capture every conversation engaged and plug it into any logging system your team uses. This gives you:
- Complete conversation history
- Custom data processing and storage
- Integration with external analytics tools
- Full control over how conversations are tracked
Setting Up a Conversations Integration
In Admin Center, click Apps and integrations in the sidebar, then select Integrations > Conversations integrations. Click Create integration and enter a unique name to identify the integration. In the Webhook section, enter the webhook endpoint URL. When a trigger event occurs, the webhook sees the event and sends the data to the target URL.
Source: Zendesk Help Center - Creating conversations integrations
While this solution requires development work, it offers the most flexibility for teams that need custom conversation tracking or want to integrate with data warehouses or other systems.
Solution 4: Consider Third-Party AI Tools
If the native Zendesk bot flows continue to cause issues, you might want to explore third-party AI agent solutions that offer more direct control over conversation workflows.
Third-party tools can provide:
- Immediate ticket creation: Automatically create tickets when conversations start
- Better conversation control: More granular control over when and how tickets are generated
- Enhanced engagement: AI that can engage visitors and send their first message automatically, reducing the chance of drop-off
While this adds another tool to your stack, it can solve the "leaky bucket" problem where conversations die without generating tickets.
Best Practices for Implementation
Regardless of which solution you choose, follow these best practices:
1. Start with AI Agent Tickets
Enable the official AI Agent Tickets feature first. It's the easiest to implement and provides immediate visibility without custom development.
2. Set Clear Escalation Rules
Decide whether you want:
- All conversations as tickets (for complete analytics)
- Only escalated conversations as tickets (to reduce ticket volume)
With Zendesk Chat, there was an option to enable automatic ticket creation when a chat ended with the transcript attached. In Sunshine Conversations, a Create Ticket Action needs to be added to achieve this result.
Source: Zendesk Help Center - Migrating advanced AI agents
3. Create Custom Views
Set up dedicated views for bot-only conversations so agents can easily identify and follow up on them when needed.
4. Monitor and Iterate
Track metrics like:
- Percentage of conversations that never receive agent response
- Common drop-off points in bot conversations
- Customer satisfaction for bot-handled vs. agent-handled tickets
5. Consider the Customer Experience
Remember that creating immediate tickets for every bot interaction might create notification overload for customers. Balance visibility with customer experience by:
- Using appropriate trigger conditions
- Setting reasonable timeout periods before creating tickets
- Managing notification preferences carefully
The Bottom Line
Lost conversations in Sunshine Conversations represent more than just a technical problem—they're missed opportunities to help customers and grow your business. Whether you choose the official AI Agent Tickets feature, implement custom triggers, build webhook integrations, or explore third-party tools, the key is ensuring every customer who reaches out gets the help they need.
Start by enabling AI Agent Tickets for immediate visibility, then layer on additional solutions based on your specific needs and technical capabilities. Your future self (and your customers) will thank you for closing that leaky bucket.
About Macha: Macha helps teams get more value from their Zendesk investment with tools and insights designed for modern support operations. Learn more at getmacha.com.
