Skip to main content

Using Fin Attributes in workflows, reports, and the inbox

Leverage Fin attributes in Workflows, Reports, and the Inbox to automate routing, generate insights, and enable teammate review.

Written by Dawn
Updated over a week ago

Fin Attributes in workflows

Fin Attributes allow you to use AI-detected conversation attributes—such as conversation category or intent—to automate your workflows. When configured correctly, Fin automatically detects these attributes during the conversation and makes them available for use in subsequent workflow steps, including branching and escalation logic.

Note: To ensure Fin Attributes are effectively applied, always include a "Let Fin handle" step in your workflow, as Fin can only detect and assign attributes when engaged in an active conversation. Attributes will not be triggered in workflows missing this critical step.

How to use Fin Attributes in workflows

  1. Set up Fin Attributes

    Configure the Fin Attributes you want to detect (for example, Urgency or Topic). These determine what Fin will analyze during customer conversations.

  2. Create Escalation Rules (optional)

    • Create an Escalation Rule and use one or more Fin Attributes as the criteria for automatic handover to your team.

    • You can also include additional attributes such as Tag, User, or Company attributes.

  3. Add a Fin block to your workflow

    Insert a Let Fin handle step in your workflow. Fin automatically detects the configured attributes during this block, no separate “Detect attribute” step is required.

    Important: Fin attributes can only be detected after a Fin block runs, not before or without one. This is what enables Fin to be involved in the conversation.

  4. Triage based on Fin Attributes

    After the Fin block, add a Branch step.

    • Set the branching condition to one of your Fin Attributes by searching for it by name.

    • Define what should happen next. For example:

      • If the topic Fin detected is Projects, assign to Projects team.

      • Else, assign the conversation to Tier 2 Support.

  5. Branch based on Escalation Rules
    After the Fin block, add an Escalation triggered step.

    • Select the Escalation Rule you created.

    • Define what should happen next—for example:

      • If Escalation = Plan cancellation, assign to account Owner.

      • Otherwise, assign the conversation to Tier 2 Support.

Tip: When searching for Fin Attributes within branch conditions, use the attribute name directly (for example, “Issue”) don’t search for “Detect” or “AI.”

Additionally, test configurations using Preview tools in the Fin Setup to validate attribute detection and workflow conditions.


Fin Attributes in reports

Automatically filter and segment your Intercom reports using Fin Attribute data to analyze conversation trends. By using AI-detected data like Topic, Sentiment, or Urgency as filters, you can build powerful, targeted reports to analyze trends, pinpoint issues and find actionable insights.

Using reports to track billing issues

Imagine you need to know not just if customers are talking about billing, and if they're upset about it.

By combining two Fin Attributes, you can build a report to track exactly this:

  1. Filter for the Issue: Create a report filtered for Issue is Billing. This isolates all conversations related to that topic.

  2. Layer on Sentiment: Next, add a second filter for Sentiment is Negative.

Your report now instantly filters out the noise and shows you only the high-priority conversations where customers are frustrated, angry, or at risk of churning.

Using reports to track feature requests from high-value customers

Imagine you want to find the most urgent feature requests from your most important customers to help prioritize your product roadmap. You can build a report by combining Fin Attributes with your own company data:

  1. Filter for the Topic: First, create a report filtered for Topic is Feature Request. This gathers all conversations where customers are suggesting new functionality.

  2. Layer on Urgency: Next, add a Fin Attribute filter for Urgency is High. This narrows the list to only those requests that customers feel are critical.

  3. Add a Customer Filter: Finally, add a filter using your own company data, such as Company size is Enterprise.

From here, you can Drill-in to view the specific conversations in your list or Export aggregated data for further analysis.


Fin Attributes in the inbox

Teammates working in the inbox can see when a attribute has been detected by Fin in two places:

  1. Conversation events within the conversation thread: Fin’s reasoning for classifying is visible when viewing conversations in the Inbox, enabling you to see when and why Fin applied attribute values.

  2. The inbox sidebar (when conversation attributes are pinned).

Teammates can:

  • Edit the attribute values in the Inbox sidebar

  • Apply macros to quickly update attributes

  • Manually complete required attributes before resolving a conversation

Note: Ensure that you have ‘Show conversation events’ enabled to see the conversation events and reasoning logs.


FAQ

Can I detect Fin Attributes without Fin answering the customer?

Yes, it is possible to categorize conversations without an AI reply by using a "Let Fin answer" block followed immediately by an Escalation Rule. This allows Fin to detect the attributes and then immediately exit the conversation to your team before sending any message to the customer.

  1. Add a "Let Fin answer" block: This is required to trigger the AI's detection engine.

  2. Configure Escalation Rules: Set up rules in the Fin AI Agent settings to trigger immediately based on the attributes you want to detect (e.g., if "Category" is "Billing" or "Channel" is "Email").

  3. Automatic Exit: When the workflow hits the Fin block, Fin will identify the attributes and, finding a matching Escalation Rule, will immediately exit the conversation before sending a reply.

  4. Route the Conversation: Once Fin exits, the conversation continues to your next workflow step (like "Route to Team") using the now-applied attributes.

Is "AI category detection" still a standalone feature in Intercom?

No, AI category detection is no longer available as a standalone feature and has been replaced by Fin Attributes. While legacy workflows containing these blocks will still function, you cannot add new "AI Category Detection" blocks to any workflow.

Do I need a specific workflow block to detect Fin Attributes?

Yes, you need a "Let Fin answer" block for attribute detection to occur; Fin Attributes cannot be detected independently of Fin being invoked in the conversation.


💡Tip

Need more help? Get support from our Community Forum
Find answers and get help from Intercom Support and Community Experts


Did this answer your question?