How do I configure intent tracking?
Required Feature Flags
The following feature flags and permissions are required to use this feature:
Feature Flag | Description |
Reason for Contact | Enables Reason for Contact and Intent tracking |
Reason for Contact Visibility | Enables viewing of Intent data in the UI |
Required Permissions:
Manage reasons for contact (
analytics.manage_reasons_for_contact) — required to create, action, archive, and restore intentsView reasons for contact (
analytics.view_reasons_for_contact) — required to view intents and their usage in conversations
Intent tracking captures what customers and agents are trying to do during a conversation. It works alongside Reason for Contact (which captures why they got in touch) to give you a complete picture of what's happening on each call or chat.
This guide covers how to set up your intent taxonomy, review AI-suggested intents, and view intent insights across your conversations.
Step 1: Open the intent management screen
Go to Analytics > RFC & intent and open the Intent tab.
You'll see the list of intents with their status, hierarchy, and usage.
Step 2: Create your intent taxonomy
Intent uses a two-level hierarchy:
Level 1 (L1) — Broad action categories like Request Information, Make a Purchase, Escalate Issue
Level 2 (L2) — Specific actions under an L1, like Product Specifications, Pricing Details, Availability Check (under Request Information)
Add an intent
Click Create new
Enter a Name
To make it an L2, tick Belongs to a parent Intent and pick the parent from the dropdown
Click Save and close (or Save and add another to keep going)
New intents are created as Active straight away.
Tips for naming intents
Frame intents as actions, not topics — use verbs (Request Callback, Cancel Service, Dispute Charge)
Don't duplicate your Reason for Contact categories
Keep L1 broad enough to be a useful grouping
Use L2 for variations of the same action
Names must be unique across your taxonomy
Step 3: Control AI suggestions
The AI continuously analyses conversations and proposes new intents when it spots a pattern that doesn't match your active taxonomy. This is how you catch emerging intents you hadn't thought of.
To control whether the AI generates suggestions, use the Generate suggestions toggle on the intent management screen:
Yes — AI proposes new intents when it sees recurring patterns
No — AI stops generating new suggestions (existing suggestions stay until you action them)
Step 4: Review and action suggestions
To find pending suggestions:
On the intent management screen, tick the Suggested status filter
Click Review next to a suggested intent
You'll see the suggested name and how many conversations it's been identified in.
Activate a suggestion
Click Activate. The intent's status changes to Active, and it'll start appearing in evaluations and reports for new conversations.
Decline a suggestion
Click Decline. The intent is archived and the AI won't suggest the same pattern again. You can review it later from the Archived list and restore it if you change your mind.
Step 5: View intents on a conversation
Open any analysed conversation from Conversations > Imported contacts.
In the Insights sidebar, open the Reason and Intent section. The intent panel shows:
Each detected intent (L1 and L2)
Evidence snippets from the conversation showing why each was identified
Which part of the conversation triggered the intent
Multiple intents can be detected on a single conversation — that's expected, since customers and agents often have several goals during one interaction.
Step 6: Filter and report on intents
Filter conversations
In the Imported contacts filter panel, find the Intent filter under the evaluagent Insight Fields section. Filter to a specific intent (L1 or L2) to find every conversation tagged with it.
Discovery tab
Go to Conversations > Imported Contacts > Discovery tab.
Use the Group by option to combine intent with other metrics — for example, group by intent and look at xCES per intent. That tells you which actions are causing the most customer effort.
For more on Discovery, see A user guide: Discovery tab.
Manage existing intents
Archive an intent
Find the intent in the list, open its action menu, and select Archive. Archived intents:
Aren't applied to new conversations
Stay visible in historical data
Can be restored anytime
Restore an archived intent
Tick the Archived status filter, find the intent, open its action menu, and select Restore. The status changes back to Active.
Important: Intent is not retrospective
When you enable Intent or add new intents:
Only new conversations ingested afterwards get tagged
Existing conversations stay untagged
Previously analysed conversations can't be reprocessed
If you enable Intent today, only conversations imported from today onwards will have intents identified.
Troubleshooting
Suggestions aren't appearing
The AI only suggests after 75+ instances of a pattern — give it time
Check the Generate suggestions toggle is set to Yes
Confirm the feature is enabled for your organisation
Intents aren't showing on conversations
The conversation hasn't been analysed yet
The conversation predates Intent being enabled
The AI didn't match any active intents (consider whether your taxonomy is too narrow)
Can't manage intents
You need the Manage reasons for contact permission. View access alone is read-only. Talk to your administrator.