Can someone help me understand the architecture? | Community
Skip to main content

I’m returning to Intercom after a long break, and I could use some advice from folks who’ve set up recent workflows.

It looks like every customer interaction starts as a Conversation. And if it takes longer to resolve or needs extra structure, it can be turned into a Ticket. That all makes sense.

But I’m trying to set things up in a way that keeps reporting clean and consistent.

My goals:

  • I want to capture the same core metadata for every interaction (things like feature area, source, etc.)
  • I don’t want to double-count anything when pulling metrics
  • I want to avoid having agents fill in fields twice, or inconsistently across conversations and tickets

My questions:

  • Do you only use tickets for escalations, or convert everything?
  • Can you safely report on Conversations only, and still get reliable counts and resolution metrics?
  • Are you adding custom fields to Conversations for universal stuff, and using ticket-only fields for more complex cases?

Trying to figure out how to avoid having my metrics split across two objects and how to train the team without overcomplicating.

Thanks in advance if anyone has gone through this setup and has advice.

 

Hey William! 👋
great questions — I totally get where you’re coming from. most teams I’ve seen use tickets mainly for escalations or multi-step issues, while keeping everyday interactions as conversations to keep reporting simple.

You can definitely get reliable metrics from conversations alone if your custom fields are set up right. Just make sure shared fields (like feature area or source) exist on both objects so data stays consistent.

It takes a bit of tweaking at first, but once the structure’s clear, it works really smoothly. 👍

Terms & ConditionsAccessibility statement