Auto-tagging
Automatically label conversations by intent, channel, sentiment, and custom keywords.
Tags make it easier to filter your inbox, run reports, and trigger downstream automations. Auto-tagging applies tags to conversations the moment they arrive, without any manual effort from agents.
AI-based auto-tagging
Liya detects intent and other signals on every inbound message and can automatically apply tags based on those signals. No rule-writing required for the basics:
- Enable Auto-tag by intent in Settings → AI → Tags and Liya will apply your intent taxonomy labels (e.g.
billing,shipping,technical) automatically. - Enable Auto-tag by sentiment to tag conversations with
frustrated,positive, orneutral. - Enable Auto-tag by urgency to apply
urgentorhigh-prioritytags.
Rule-based auto-tagging
For custom tagging logic, create automation rules with a Add tag action:
| Rule example | Condition | Action |
|---|---|---|
| VIP customer | Contact tag has 'enterprise' OR contact attribute plan = 'Enterprise' | Add tag: vip |
| First-time customer | Contact conversation count = 1 | Add tag: new-customer |
| Churn risk | Message contains 'cancel' OR 'cancellation' OR 'leaving' | Add tag: churn-risk, set priority: high |
| Legal language | Message contains 'lawsuit' OR 'attorney' OR 'legal action' | Add tag: legal-flag, assign to Legal group |
| Weekend ticket | Day of week is Saturday or Sunday | Add tag: weekend |
| Long-form complaint | Message word count > 300 AND sentiment is frustrated | Add tag: complex-complaint |
Managing your tag taxonomy
Maintain a consistent set of tags to keep your inbox and reports clean. Go to Settings → Tags to:
- Create new tags with names and optional colours
- Merge duplicate tags
- Archive tags no longer in use (existing tagged conversations retain the tag)
- View the conversation count per tag
Using tags in analytics
Tags are first-class filters in the Analytics and CSAT dashboards. Filter reports by tag to understand performance on specific ticket types — e.g. "What is our average resolution time for billing tickets?" or "What is the CSAT score on churn-risk conversations?"