How to build Drag inbox dashboards in Metabase

Drag turns Gmail and Google Workspace into a shared inbox and help desk with kanban boards, cards, and tags. Metabase is where you turn that activity into shared, trustworthy dashboards. This guide covers two complementary paths: a lightweight MCP route for fast, AI-assisted questions, and a durable pipeline route that syncs Drag into a database so you can build dashboards anyone can read.

How do you connect Drag to Metabase?

Most teams combine these: use the MCP route to explore and triage, and the pipeline route for the dashboards people depend on.

What can you analyze from Drag data in Metabase?

• Card volume — created vs. closed by day and board
• Time to first reply — how long customers wait for a response
• Board and column health — backlog by shared inbox and status
• Backlog and aging — open cards and how long they've been waiting
• Team workload — assignments and emails sent per member
• Tag drivers — volume and resolution speed by tag
• Contacts — new vs. returning senders

Which Drag dashboards should you build in Metabase?

How do you use the Drag and Metabase MCP servers together?

Pair the Drag MCP server with the Metabase MCP server for live, conversational analysis. The Drag MCP reads current boards and cards; the Metabase MCP queries the models and dashboards you've already built.

Example workflows

• List unassigned cards in the support board and summarize by tag.
• Pull today's open cards from Drag and compare response trends against a Metabase model.
• Triage: "show cards with no reply in 24 hours."

Be honest about the limits

• MCP is great for live lookups — not for scheduled or audited reporting.
• It does not create history; trend analysis still needs synced data.
• The MCP acts on your Gmail/Drag data with your permissions.
• The Metabase MCP server is built in; an admin enables it under Admin → AI → MCP.

How do you set up the Drag and Metabase MCP servers?

{
  "mcpServers": {
    "drag": {
      "command": "npx",
      "args": ["-y", "@dragapp/mcp-server"],
      "env": {
        "DRAG_API_KEY": "your-drag-api-key"
      }
    },
    "metabase": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://your-metabase.example.com/api/metabase-mcp"]
    }
  }
}

Generate a Drag API key in your account settings. Confirm the exact MCP package name and install command in Drag's developer docs before standardizing.

Can you generate a Drag dashboard with AI?

Yes — and this is the fastest way to a strong first draft. Use the prompt below with the Metabase MCP server and any assistant that can inspect your warehouse schema and create Metabase questions. It assumes Drag data is already synced into a database Metabase can read, treats MCP as exploratory, and tells the agent to skip metrics the schema can't support instead of faking them.

Create a polished Metabase dashboard for Drag shared-inbox analytics using the
available Drag tables in this database.

Goal: Help team leads understand inbox volume, response time, workload, and tag
drivers from Drag (Gmail shared inbox) data.

First, inspect the schema and identify the available Drag tables. Do not assume
exact table names. Map the available raw tables into these analytical concepts
where possible: Boards (shared inboxes), Cards (conversations), Columns (statuses),
Threads/Emails, Tags, Users, and Contacts.

Important:
- Treat MCP data access as exploratory only.
- Build the dashboard from durable database/warehouse tables.
- Use medians (p50) and p90 for reply times, never averages.
- Define "first reply" as the first outbound email from a team member, excluding
  internal notes and automated messages.
- If card movement history is missing, do not calculate time-in-column or reopen
  rate. Use a caveat instead.
- Do not claim Metabase connects natively to Drag unless that is explicitly true
  in this environment.

Dashboard title: Drag Inbox Overview

Sections:
1. Executive summary (KPI cards): Cards created last 7 days; Closed last 7 days;
   Open backlog; Median time to first reply; Volume by board.
2. Volume & backlog: Created vs closed by day; Open by column; Backlog aging;
   Volume by board.
3. Response time: First reply p50/p90 by week; Cards by tag; Reopened by week
   (only if history exists).
4. Team workload: Closed by assignee; Open assigned by assignee; Emails sent by
   member; Median handle time by assignee.
5. Tags & drivers: Volume by tag; Slowest tags; New vs returning contacts;
   Volume by hour.

Filters: Board, Assignee, Tag, Column/status, Date range.

Before finalizing, create or recommend reusable Metabase models:
modeled_drag_cards, modeled_drag_boards, modeled_drag_emails,
modeled_drag_users, and modeled_drag_contacts.

Output: Build the dashboard if you have permission; otherwise provide the exact
questions, SQL, model definitions, and layout. Include caveats for any metric
that cannot be calculated from the available schema. Keep it practical, dense,
and executive-readable. Avoid vanity metrics.

How do you sync Drag data into a database or warehouse?

For dashboards that need history and reliability, land Drag data in a database first, then connect Metabase to that database.

Connector options

• Drag REST API (v2)(raw) — the source of truth; read boards, cards, tags, and contacts. The base URL isv2/ with API-key auth and a rate limit around 100 requests/minute.
• dlt(code) — wrap the v2 API in a Python pipeline for incremental loads and schema control.
• Webhooks — Drag includes webhooks; subscribe to card events to keep your warehouse fresh in near real time.

Notes

• Land raw tables first, then build clean models on top.
• Respect the rate limit — page and back off rather than hammering the API for large historical loads.
• Sync email-level data so you can compute reply times, not just card counts.

How should you model Drag data in Metabase?

Core tables

Modeling advice

• Treat a card as the conversation/ticket grain for most dashboards.
• Define first reply from the first outbound, non-note email on a card.
• Map columns to a small status set (open/in-progress/done) so charts stay stable.
• Treat tags as a bridge table so a card can carry many tags.
• Define "closed" once and reuse it everywhere.

Which Drag metrics should you track in Metabase?

What SQL powers Drag dashboards in Metabase?

These assume the modeled tables above (PostgreSQL dialect). Adjust identifiers to match your warehouse.

SELECT
  date_trunc('day', c.created_at) AS day,
  COUNT(*)                                          AS created,
  COUNT(*) FILTER (WHERE c.is_closed = true)        AS closed
FROM cards c
WHERE c.created_at >= CURRENT_DATE - INTERVAL '30 days'
GROUP BY 1
ORDER BY 1;

WITH first_outbound AS (
  SELECT
    e.card_id,
    MIN(e.sent_at) AS first_reply_at
  FROM emails e
  WHERE e.direction = 'outbound'
    AND e.is_note = false
  GROUP BY e.card_id
)
SELECT
  date_trunc('week', c.created_at) AS week,
  percentile_cont(0.5) WITHIN GROUP (
    ORDER BY EXTRACT(EPOCH FROM (f.first_reply_at - c.created_at)) / 60.0
  ) AS median_first_reply_min
FROM cards c
JOIN first_outbound f ON f.card_id = c.id
GROUP BY 1
ORDER BY 1;

SELECT
  b.name             AS board,
  COUNT(*)           AS open_cards
FROM cards c
JOIN boards b ON b.id = c.board_id
WHERE c.is_closed = false
GROUP BY b.name
ORDER BY open_cards DESC;

SELECT
  u.name             AS assignee,
  COUNT(*) FILTER (WHERE c.is_closed = false) AS open_cards,
  COUNT(*) FILTER (WHERE c.is_closed = true)  AS closed_cards
FROM cards c
JOIN users u ON u.id = c.assignee_id
WHERE c.created_at >= CURRENT_DATE - INTERVAL '30 days'
GROUP BY u.name
ORDER BY open_cards DESC;

What are common mistakes when analyzing Drag in Metabase?

Related

Related analytics

Related integrations

FAQ