Skip to content
AskVault Docs
Try Free →

Tagging conversations in AskVault

Written by , Founder, AskVault · Reviewed by Aashiq
Last updated: · 4 min read

What tags are for

Tags solve three problems for an inbox of growing size:

  1. Filter the inbox. "Show me all billing conversations from the last 7 days." Tags make this a one-click filter.
  2. Route SLA policies. Billing conversations get a 4-hour SLA; technical conversations get 24 hours. Tags determine which SLA applies.
  3. Segment analytics. "What % of conversations are about refunds vs technical issues?" Tags enable this segmentation.

A useful tag taxonomy has 10 to 30 tags. Beyond that, tagging effort exceeds the analytical value.

Tag structure

Each tag has:

  • Name. Short string, lowercase-with-hyphens by convention. Example: billing, technical, refund-request.
  • Color. Visual marker in the inbox UI. Useful for quick scanning.
  • Description. What this tag means. Helps team members tag consistently.
  • Active flag. Tags can be archived without deletion. Archived tags don't appear in tagging dropdowns but historical conversations keep them.

Configure under Dashboard > Tags.

Manual tagging

Agents tag conversations during normal handling:

  • From the conversation header. Click the tag dropdown, pick one or more tags.
  • Bulk-tag. Multi-select conversations in the inbox view, apply tags to all selected.
  • Keyboard shortcut. T opens the tag picker on the focused conversation.

Tags apply immediately. No save button. Removing a tag works the same way: click the X on the tag chip.

Auto-tagging by intent

The AI agent can tag conversations automatically based on detected intent. Common auto-tags:

  • billing when the conversation mentions plans, prices, invoices, refunds.
  • technical when API errors, integration failures, or bug reports surface.
  • returns for e-commerce return-related conversations.
  • sales when buying signals trigger.
  • at_risk when sentiment_router flags frustration.

Auto-tagging runs continuously during the conversation. Tags get added as new signals appear. An angry customer asking about billing might end up with billing, at_risk, urgent tags by the time they escalate.

Configure auto-tag rules under Dashboard > Tags > Auto-Tagging Rules.

Tag-based SLA routing

SLA policies can apply differently based on tags:

  • Default SLA. 24-hour first response.
  • billing SLA. 4 hours (faster because billing issues affect customer's ability to pay you).
  • urgent SLA. 1 hour.
  • at_risk SLA. 30 minutes (frustration mitigation).

Configure under Automation Rules > SLA Routing. Each rule maps a tag (or tag combination) to an SLA tier.

Tag-based inbox views

Saved inbox views filter by tag combinations:

  • "Billing queue." Tag = billing, status = open or escalated.
  • "At-risk customers." Tag = at_risk OR urgent, status = any.
  • "Today's resolved." Status = resolved, resolved-date = today.

Tag filters are AND/OR composable. Useful for routing different team members to different specialized queues.

Tags vs audience tags vs custom fields

Three similar concepts that solve different problems:

  • Tags. Per-conversation labels for filtering, routing, analytics. Multi-valued.
  • Audience tags. Per-document labels for access control. Restrict what content the bot can retrieve based on the verified user's audience set.
  • Custom fields. Per-customer or per-conversation key-value pairs. Used for structured data like plan tier, account ID, etc.

Don't conflate them. Tags are for ops workflows. Audience tags are for content access. Custom fields are for structured customer state.

Tag taxonomy best practices

Three patterns we see work well in real teams:

  • Topic tags. billing, technical, returns, sales, feedback. Routes to the right team.
  • Severity tags. urgent, high, normal, low. Drives SLA and prioritization.
  • Resolution tags. bug-fixed, policy-explained, escalated-to-engineering. Powers post-resolution analytics.

Avoid tag-spam: tags for every possible nuance. Stick to 10 to 30 tags total. Reduce as your team's process matures.

Bulk-tagging from the API

For programmatic operations (CRM sync, bulk operations from external tooling):

Terminal window
curl -X PATCH https://api.askvault.co/v1/conversations/conv_xxx \
  -H "Authorization: Bearer ak_xxx" \
  -H "Content-Type: application/json" \
  -d '{"tags": ["billing", "urgent"]}'

Sets tags to exactly the provided list (replacing any existing tags). For additive operations, fetch first then submit the merged list.

Limits

  • Per-workspace cap. 100 active tags on Free and Starter, 500 on Growth, unlimited on Business.
  • Per-conversation cap. Up to 20 tags per conversation.
  • Auto-tag rule count. Up to 50 rules on Growth, 200 on Business.

Common pitfalls

Tag proliferation. Team members create new tags freely; taxonomy bloats. Periodically audit and consolidate under Dashboard > Tags > Audit.

Inconsistent tagging. Different agents tag the same conversation differently. Document your tag taxonomy and review onboarding consistency.

Auto-tag misfires. Rule fires on the wrong conversations. Refine the rule's keyword matchers or move from keyword to intent-based matching.

Tag-based SLA routes wrong. Tag was applied late in the conversation, after the SLA timer was already running against the default. Set auto-tag rules to fire as early as possible.

FAQ

Can I rename a tag without breaking historical tagging?

Yes. Renaming the tag updates the label on every conversation currently tagged with it. The tag's underlying ID stays the same.

Can the bot tag conversations on its own?

Yes, via auto-tagging rules. The bot evaluates the conversation context and applies tags that match. Configure under Dashboard > Tags > Auto-Tagging.

Are tags visible to customers?

No. Tags are agent-side only. Customers never see them. For visible labels on customer-facing UI, use custom message templates instead.

Can I import tags from another support tool?

Yes via CSV. Dashboard > Tags > Import CSV. Columns: name, color, description, archived. Up to 500 tags per import.

How do tags interact with team permissions?

Members can view and apply existing tags. Only admins can create or delete tags. Configure under Settings > Team > Permissions.

Was this page helpful?