Chatwoot Reporting Analytics

Chatwoot Reporting & Analytics

Guide for using Chatwoot's reporting tools to analyze support performance.

Report Types Overview

Chatwoot offers two generations of reporting APIs:

| API | Tools | Best for | | ------ | -------------------------------------------------------------------------------------------------------------- | ----------------------------------------------------- | | V1 | reports_account_overview, reports_account_summary, reports_agent_summary, reports_conversation_metrics | Real-time overview, quick summaries | | V2 | reports_v2_overview, reports_v2_agents, reports_v2_inboxes, reports_v2_teams, reports_v2_labels | Detailed breakdowns by dimension, date range analysis |

Recommendation: V2 endpoints are more comprehensive. Use V1 for real-time snapshots, V2 for historical analysis.

Real-Time Overview

Account overview (V1)

Get an instant snapshot of current conversation state:

reports_account_overview(account_id: 1)

Returns:

• Open conversations count
• Pending conversations count
• Unattended conversations count
• Agents online count

Conversation meta counts

For conversation-status breakdowns:

conversations_meta(account_id: 1)
→ { open: 45, resolved: 1230, pending: 12, snoozed: 3, all: 1290, ... }

V1 Reports

Account summary

Summarized metrics for a date range:

reports_account_summary(
  account_id: 1,
  since: "2026-01-01",
  until: "2026-01-31",
  type: "account"
)

Agent summary

Per-agent performance metrics:

reports_agent_summary(
  account_id: 1,
  since: "2026-01-01",
  until: "2026-01-31"
)

Returns per agent:

• Conversations handled
• Average first response time
• Average resolution time

Conversation metrics

reports_conversation_metrics(
  account_id: 1,
  type: "account"
)

V2 Reports (Recommended)

All V2 endpoints accept a consistent parameter set:

| Parameter | Type | Description | | ------------ | ------- | --------------------------------------------------- | | account_id | integer | Required | | since | string | Start date (ISO format or Unix timestamp) | | until | string | End date | | timezone | string | IANA timezone (e.g., "America/New_York", "UTC") |

V2 Overview

Aggregate metrics across the account:

reports_v2_overview(
  account_id: 1,
  since: "1704067200",
  until: "1706745600",
  timezone: "UTC"
)

V2 Agent Report

Per-agent breakdown:

reports_v2_agents(
  account_id: 1,
  since: "1704067200",
  until: "1706745600",
  timezone: "America/New_York"
)

Returns per agent:

• conversations_count — Total conversations
• avg_first_response_time — Average first response (seconds)
• avg_resolution_time — Average time to resolve (seconds)
• resolved_conversations_count — Number resolved

V2 Inbox Report

Per-inbox breakdown:

reports_v2_inboxes(
  account_id: 1,
  since: "1704067200",
  until: "1706745600",
  timezone: "UTC"
)

V2 Team Report

Per-team breakdown:

reports_v2_teams(
  account_id: 1,
  since: "1704067200",
  until: "1706745600",
  timezone: "UTC"
)

V2 Label Report

Per-label breakdown — useful for tracking issue categories:

reports_v2_labels(
  account_id: 1,
  since: "1704067200",
  until: "1706745600",
  timezone: "UTC"
)

Key Metrics Explained

| Metric | Description | Good benchmark | | ---------------------------- | ---------------------------------------------------- | -------------------- | | First Response Time | Time from conversation creation to first agent reply | 4.0 / 5.0 | | Unattended Conversations | Open conversations with no agent response | Should be 0 | | Agent Online | Agents currently available | Monitor for coverage |

Date Handling

V2 reports accept dates as Unix timestamps (seconds since epoch):

# January 1, 2026 00:00:00 UTC → 1767225600
# January 31, 2026 23:59:59 UTC → 1767311999

Set timezone to match your team's working timezone for accurate daily breakdowns.

Common Analysis Patterns

See METRICS_REFERENCE.md for detailed metric calculations.

See EXAMPLES.md for real-world reporting scenarios.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: fazer-ai
• Source: fazer-ai/chatwoot-skills
• License: MIT

Install and usage instructions live in the source repository linked above.