Question 1

What Clerk data can Polytomic sync?

Accepted Answer

Users, organizations, and organization memberships. Those three cover the questions most teams have: who signed up, which accounts exist, and who belongs to what with which role. Invitations, sessions, JWT templates, domains, and allowlists are not exposed as objects. Organizations do carry a pending invitations count, and SAML domain information appears nested inside user records, so some of that data reaches you indirectly even without its own table.

Question 2

How is Clerk metadata handled?

Accepted Answer

Public, private, and unsafe metadata are preserved as JSON objects, and Polytomic additionally infers their child keys during schema discovery by sampling records, so common metadata fields become individually selectable rather than requiring you to parse JSON. Two caveats. The inference is not recursive, so deeply nested structures stay as objects. And it works from a bounded sample, so a key that only appears on a handful of users may not surface. The complete object is always there regardless.

Question 3

How do email addresses and phone numbers come through?

Accepted Answer

As typed arrays of objects, preserving Clerk's model where a user can have several of each with different verification states. They are not exploded into separate rows, and there is no automatically flattened primary email column, though users do expose the identifier of their primary email address and phone number. So getting a single primary email means joining that identifier back to the array in your data warehouse, which is a small transformation worth building once.

Question 4

Does Clerk sync incrementally?

Accepted Answer

Users do. They are requested in descending updated time order with the checkpoint set to the newest record seen, and later runs read until they reach records older than that checkpoint, deliberately allowing a little overlap where timestamps are equal so nothing at the boundary is skipped. Organizations and memberships are full reads on each ETL execution. One consequence of the user cursor is that deletions are not discoverable, so a periodic full resync reconciles removed users.

Question 5

Why sync auth data into a data warehouse?

Accepted Answer

Because for a product led business the signup record is where the customer journey begins, and it usually lives nowhere else. Joining Clerk users and organizations to product usage, billing, and CRM data lets you measure activation, see which accounts are adding seats, and spot organizations whose membership is shrinking before they churn. Reverse ETL on the same platform then pushes new signups into your CRM or lifecycle tool automatically, so sales and marketing act on them without an engineer wiring up a webhook.

Clerk

Perfect pairings

Users and org structure

Metadata discovered

Incremental user reads

Example workflows

What you can sync

From Clerk

To your destinations

How it works

How to set up

Frequently asked questions

Why teams choose Polytomic

No engineering required

Flexible data modeling

Handles scale automatically

Start syncing your data with Clerk in minutes

More integrations beyond Clerk