How to build HubSpot sales dashboards in Metabase

# HubSpot (remote, OAuth 2.1 + PKCE in browser)
claude mcp add --transport http hubspot https://mcp.hubspot.com

{
  "mcpServers": {
    "hubspot": {
      "url": "https://mcp.hubspot.com"
    }
  }
}

# Install the Metabase CLI
npm install -g @metabase/cli

# Log in (opens your browser; requires Metabase v62+)
mb auth login --url https://your-metabase.example.com

# Load a HubSpot CSV export — creates a table AND a model
mb upload csv --file hubspot-deals.csv --collection root

# Refresh that same table later from a new export
mb upload replace <table-id> --file hubspot-deals.csv

Create a polished Metabase dashboard for HubSpot sales analytics.
Work end to end: get the data into Metabase if it isn't there yet, then build.

Goal: Help sales leaders understand pipeline, forecast, win rate, sales cycle,
and rep activity from HubSpot CRM data.

Step 1 — Find or load the data:
- First, check what already exists in Metabase (search for HubSpot tables and
  models). If durable HubSpot data is already present — synced from a warehouse
  or uploaded earlier — use it and skip to Step 2.
- If nothing is there, pull it with the HubSpot MCP server (read-only scopes):
  deals, deal pipelines/stages, contacts, companies, and engagements. Write each
  result to a CSV, then load it with the Metabase CLI — run
  "mb upload csv --file <export>.csv" so each upload creates a table and a
  ready-to-query model. Use "mb upload replace <table-id> --file <export>.csv" to
  refresh an existing table instead of creating duplicates.

Step 2 — Inspect before querying:
HubSpot CSV exports are usually flat and pre-aggregated (one row per deal, contact,
or company, with columns like stage, amount, close date, and owner). Warehouse
tables are raw and include deal property/stage history for velocity metrics.
Inspect the actual tables and column names first; do not assume exact names or
that stage-history exists.

Important:
- Build on whatever data is present; don't claim Metabase connects natively to
  HubSpot — it reads a database or CLI-uploaded tables.
- Only recompute rates over the correct base (e.g. win rate) when raw counts are
  available; if the data already provides a rate, chart it directly.
- Define "won" and "closed" once (closed-won vs. any closed stage) and reuse the
  definitions everywhere.
- For win rate, state the denominator explicitly (won / closed vs. won / created)
  and hold the cohort fixed.
- Report sales cycle length and deal size as medians (p50) and p90, never plain
  averages — both are right-skewed.
- If deal stage history is missing, do not calculate sales cycle length,
  time-in-stage, or stage conversion. Use a caveat instead.
- Convert all amounts to a single reporting currency; caveat any mix.
- Only build a card if its underlying column/metric exists in the data.
- A single CSV is a point-in-time snapshot: only build trend cards if there is a
  usable date column or multiple periods have been uploaded.

Dashboard title: HubSpot Sales Overview

Sections:
1. Executive summary (KPI cards): Open pipeline; Pipeline coverage vs. target;
   Win rate (last 90 days); Median sales cycle length; Average and median deal
   size; Deals closing this period.
2. Pipeline: Open pipeline by stage; by owner and segment; aging and past-due
   close dates.
3. Forecast: Weighted vs. unweighted pipeline; commit vs. best-case vs. closed;
   deals expected to close by week.
4. Win rate & conversion: Win rate by month and segment; stage-to-stage
   conversion; loss reasons.
5. Activity: Calls/emails/meetings by rep; activities per open deal; new-lead
   response time.

Filters: Pipeline, Stage, Owner, Segment/Industry, Source, Date range.

Reuse the models Metabase auto-created from uploaded CSVs, or (for a warehouse)
create reusable models: modeled_hubspot_deals, modeled_hubspot_contacts,
modeled_hubspot_companies, modeled_hubspot_stage_history, and
modeled_hubspot_activities.

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 data. Reconcile totals against
HubSpot reports. Keep it practical, dense, and executive-readable. Avoid vanity
metrics.

Won as a share of closed deals over the last 12 months.

SELECT
  date_trunc('month', d.closed_at) AS month,
  COUNT(*) FILTER (WHERE d.is_won)                 AS won,
  COUNT(*) FILTER (WHERE d.is_closed)              AS closed,
  ROUND(
    100.0 * COUNT(*) FILTER (WHERE d.is_won)
      / NULLIF(COUNT(*) FILTER (WHERE d.is_closed), 0),
    1
  ) AS win_rate_pct
FROM modeled_hubspot_deals d
WHERE d.is_closed
  AND d.closed_at >= CURRENT_DATE - INTERVAL '12 months'
GROUP BY 1
ORDER BY 1;

The current funnel — deals and value by stage, in order.

SELECT
  s.stage_label,
  s.display_order,
  COUNT(*)               AS open_deals,
  ROUND(SUM(d.amount), 2) AS open_amount
FROM modeled_hubspot_deals d
JOIN modeled_hubspot_stages s ON s.stage_id = d.stage_id
WHERE NOT d.is_closed
GROUP BY s.stage_label, s.display_order
ORDER BY s.display_order;

Days from create to close-won; medians beat averages here.

-- Median days from create to close-won, by month closed
SELECT
  date_trunc('month', d.closed_at) AS month,
  percentile_cont(0.5) WITHIN GROUP (
    ORDER BY EXTRACT(EPOCH FROM (d.closed_at - d.created_at)) / 86400.0
  ) AS median_cycle_days,
  percentile_cont(0.9) WITHIN GROUP (
    ORDER BY EXTRACT(EPOCH FROM (d.closed_at - d.created_at)) / 86400.0
  ) AS p90_cycle_days
FROM modeled_hubspot_deals d
WHERE d.is_won
GROUP BY 1
ORDER BY 1;