Data queries

• Data queries only run against external databases (PostgreSQL, MySQL or Amazon Redshift). They don’t work with integrations, APIs or organization databases.
• A data query has no parameters. It runs exactly as saved. If you need the same query with different filters, save one query per filter, or let the coworker or app query the database directly.
• Data queries are visible to your whole organization, and each name must be unique in your organization.
• You manage data queries under Settings → Data → Data queries. They don’t appear on the Connections page.

​

Create a data query

select date_trunc('month', started_at) as month, sum(amount) as mrr
from subscriptions
where status = 'active'
group by 1
order by 1 desc

​

Run a data query

​

With a coworker

Run the “Monthly MRR” data query and chart the last 12 months.

What does “Active users last 30 days” return right now?

​

In apps

import { runDataQuery } from '@/client-lib/data-query-client';

const response = await runDataQuery('Monthly MRR');
const rows = response.data;

​

Edit a data query

​

Delete a data query

​

Tips

• Use descriptive names. “Open support tickets by priority” is easier to find and reuse than “tickets”.
• Try it in chat first. Ask a coworker to explore the data and write the SQL with you. Once the results look right, save the SQL as a data query.
• Keep each query focused. One query should answer one question. Split queries that try to do several things.