Pardon the mess

You are viewing docs related to an older version of Sequin. We’re in the process of updating our provider-specific guides and will be done in a few weeks (February 2024). Please click here to view the latest version of the docs.

If you’re interested in HubSpot and want to learn more about what’s changing, send us a note.

HubSpot database schema

You can select to sync any of HubSpot’s standard CRM objects to your Sequin database:

  • Companies
  • Contacts
  • Deals
  • Line items
  • Owners
  • Products
  • Quotes
  • Tickets
  • Meetings
  • Calls
  • Tasks
  • Postal Mail
  • Communications
  • Notes

Additionally, Sequin can sync custom CRM objects and associations between any of these collections.

Need other objects in your sync? Just send us a note.

The syncing process

HubSpot OAuth Apps like Sequin have dedicated rate limits:

APISyncsRequest limitInterval
StandardOwners, Pipelines, Associations10010 seconds
SearchObjects (e.g. Company, Contact, custom objects)41 second

Because these limits are scoped to the OAuth application, Sequin won’t impact your existing API quota usage (whether from other OAuth apps or your own HubSpot Private app). Sequin selects the fastest API to use for each collection in each stage of your sync’s lifecycle and maximizes its consumption of the available rate limit.

Because your synced tables share these rate limits, we recommend syncing what you need. If you realize you aren’t using one or more synced HubSpot tables, removing it from your sync may speed up the other tables!

When you create a sync or update your selected columns, Sequin initializes your tables with the current state of your HubSpot data. After the initial backfill, it continuously polls those collections — either repeatedly sweeping a collection or listening for changed records — and updates your tables accordingly.

Because these rate limits are scoped to our OAuth app and your HubSpot workspace, we do not recommend syncing the same HubSpot workspace with more than one active Sequin sync.

Learn more about [our syncing process]/syncs.


When you associate two objects in HubSpot, the association is undirected. When Contact A is associated with Company X, Company X is reciprocally associated with Contact A.

You can use association labels to indicate the kind of association between two objects. For example, Contact A can be a Manager at Company X and also a Former Employee at Company Y; you would associate Contact A with both companies but label the associations differently.

For any pair of object types, Sequin syncs all associations between objects of those types to an associative table. For example, every association between a Contact and a Company syncs to the table associations_company_contact with the following structure:additional-fields

Column namePostgres typeExample valueDescription
company_idtext“9457785876”The associated company’s HubSpot ID.
contact_idtext“36701”The associated contact’s HubSpot ID.
labelstext[]ManagerThe association’s labels.

Because HubSpot associations are undirected, these associations will only appear in one table: if you have a table associations_company_contact, there won’t be a table with the opposite order (associations_contact_company). These table names are a fixed function of the type names.

You can JOIN using the associations tables:

  -- Contact data.,,
  -- Data from the corresponding company.,,
  -- Label data from the associations table.
FROM hubspot.associations_company_contact association
LEFT JOIN company ON = association.company_id
LEFT JOIN contact ON = association.contact_id;

This example uses Contact and Company, but the same patterns apply for any other HubSpot CRM objects you can sync with Sequin. For example, if you have custom Requests and custom Cars in HubSpot, Sequin would sync an associations_car_request table with the columns car_id and request_id.


To write back to HubSpot, you’ll connect via Sequin’s Postgres Proxy. The Proxy listens for changes. When you make a mutation, the Proxy applies the mutation to HubSpot’s API first. If the mutation succeeds, your database is updated as well. If it fails, your database mutation will be rolled back, and you’ll receive a Postgres error.

With this architecture, HubSpot is the source of truth. All your database changes are validated by HubSpot’s API first. There’s never a chance for the two to get out-of-sync.


To create a new record in the API, you’ll use the insert command. Here’s how to insert a HubSpot contact:

insert into hubspot.contacts (first_name, last_name, email)
values ('Paul', 'Atreides', '')
returning id;

(1 row)

An insert command must translate to a single API query. Because HubSpot does not allow for transactional batch inserts, inserts on HubSpot are limited to a single record at a time.

Without transactional batch inserts, bulk inserting records could get your data into an inconsistent state. For example, imagine you batch inserted 20 HubSpot contacts. 2 of them failed. Is your code going to pull those out and try them again? What about the other 18 you created successfully – should they remain created or do you want to roll them back?

To avoid these tricky situations, we limit HubSpot to one insert at a time. For updates and deletes, those are safe to retry, and so we don’t impose the same limitation.


To update a record in the API, you’ll use the update command. Here’s how to update a HubSpot contact:

update hubspot.contacts
set email = ``, note = 'Moved to new planet called Dune'
where id = '0038b00002gszToAAI';

(1 row)

An update command must translate to a single API query. That means you can update a maximum of 100 records at a time.


To delete a record in the API, you’ll use the delete command. Here’s how to delete a HubSpot contact:

delete from hubspot.contacts
where email = '';

(1 row)

A delete command must translate to a single API query. That means you can delete a maximum of 100 records at a time.

  1. There may be other fields in your table, like _sync_hash, _sync_inserted_at, and _sync_updated_at. Don’t be alarmed: Sequin uses these columns to keep data flowing into Postgres smoothly. Just don’t tamper with this data↩

Was this page helpful?