HubSpot connector
Overview
The HubSpot connector lets a Neblex workflow or agent treat a HubSpot portal as a source and a destination: read and write records across the CRM objects, walk the associations between them, manage lists, and start a flow when records are created or changed.
- Signs in with a private app access token, so the connection has exactly the scopes you grant that app
- Custom objects and properties work through the generic object, schema, and property operations
- Associations are read per related object type with Get association, then the related records are read in one batch call
- One connector powers deterministic workflows, agent tool calls, and marketing and sales sync jobs, all in the same run history
Why teams use the HubSpot connector
The three jobs this connector is picked for most often, and what each one gets you.
- Run a real two-way CRM sync. Contacts, companies, and deals stay in step with another CRM or ERP, with the mapping and the rules for who wins a conflict defined in the flow you can read.
- Turn a closed deal into an order. A won deal becomes a customer and an order in the back office, with large deals approved by a person first.
- Get marketing and sales data to the warehouse. New and changed records are loaded into your warehouse on a schedule, so reporting joins HubSpot with everything else.
Worked example
A real flow revenue teams build early, with the actual step-level detail.
Closed-won HubSpot deal to a NetSuite customer and sales order
When a deal reaches Closed Won, Neblex reads its company and contacts, creates or updates the customer in the ERP, and raises the order. Deals over a threshold wait for approval first.
- Trigger:
New or updated deal, polled on the schedule you set, with a Filter that continues only whendealstageisclosedwon. - Read associations:
Get associationfrom the deal to companies, and again to contacts, then a batch read for the records. - Gate: if
amountis 25,000 or more, a Wait for form step sends an approval task to the deal owner's manager with the company name and line items. On reject, log a note on the deal and stop. - Customer:
Upsert by external idon the NetSuite customer, keyed on the HubSpot company ID, mappingcompany.name → companyName,company.domain → website,company.country → billing country, and the primary contact'semail,firstname, andlastname. - Order: create the sales order from the deal's line items, mapping
line.hs_product_id → item,line.quantity → quantity,line.price → rate. - Close the loop: write the order number back to a deal property
erp_order_number, and post the order link to the deal owner in Slack.
Example flows
Common patterns other teams start from with this connector.
- Two-way CRM sync. Contacts, companies, and deals stay in step between HubSpot and Salesforce, with field and owner mapping and conflict rules you control. (RevOps)
- Lead enrichment and routing. New contacts are enriched, scored, checked against existing records, and assigned an owner before sales ever sees them. (Marketing)
- CRM to warehouse replication. New and changed contacts, companies, and deals land in Snowflake on each poll. (Data)
- Inbound triage agent. Given a form submission, an agent researches the company, checks fit against your ICP, and either routes to sales or nurtures, with a note explaining the call. (Agentic)
Supported objects
The connector covers the HubSpot CRM objects, plus lists, pipelines, owners, properties, schemas, imports, and exports. Custom objects are available on the HubSpot tiers that allow them.
Standard objects
Contact, Company, Deal, Ticket, Lead, Product, Line Item, Quote, Invoice, Order, Cart, Call, Email, Meeting, Note, Task, List, Owner, Pipeline
Custom objects and custom properties work through the generic object operations and the Schema and Property operations, including associations.
Security and access
The connection uses a HubSpot private app access token, so it can only touch the objects and actions that app is allowed to.
- Private app scopes. Grant the app only the scopes your flows need; a token with read-only deal scope cannot write deals.
- Encrypted credentials. The token is stored encrypted and never shown again after you save the connection.
- Webhook endpoints. If you point a HubSpot webhook subscription at a Neblex Webhook trigger, pick an authentication mode HubSpot can send, or keep the trigger's unguessable address private.
- Run history. Every run shows each step's status and errors; record data from successful runs is kept when the flow is set to keep run data for replay.
Limits, as HubSpot publishes them
The shape of HubSpot's limits, as published at the time of writing. Verify the exact figures for your plan; they change without notice.
- REST calls are limited to roughly 100 requests per 10 seconds per app per portal, with a daily cap that rises by subscription tier and the API add-on
- Batch endpoints process up to 100 records per call
- The Search API returns up to 200 results per page and 10,000 results per query, at about 4 requests per second
- Rate-limited reads are retried with backoff, honoring HubSpot's retry signal; creates and searches are POST calls, so turn on Retry on failure or an idempotency key for those steps
What a flow can do in HubSpot
Create and update contacts, companies, deals, and tickets, search records by any property, read and manage associations between records, add and remove list memberships, log calls, emails, notes, and meetings, batch create, update, and archive, and many more.
The connector has 336 operations, and 81.5% of them describe their output so later steps can map fields directly. Sign-in method: API token.
Every operation, with its inputs and outputs, is listed in the HubSpot connector reference.
How flows react to changes in HubSpot
The HubSpot connector has 46 ready-made triggers, checked by polling on the schedule you set, from every minute upward. A flow that starts from one of them runs with the changed record already in hand.
Flows can start on new or changed contacts, companies, deals, and tickets, new or changed leads, quotes, invoices, and line items, new or changed imports.
Frequently asked questions
Does the HubSpot connector support custom objects?
Yes, on the HubSpot tiers that allow custom objects. Custom objects and properties are reached through the generic object, schema, and property operations.
Can it react to changes as they happen?
The connector has 46 polled triggers, such as New or updated deal and New contact, checked on the schedule you set from every minute upward, so a flow starts with the changed record in hand. For an immediate start, point a HubSpot webhook subscription at a Neblex Webhook trigger.
How does it handle associations?
Get association lists the records of one related type, with their association labels, so a flow reads a deal's companies and contacts with one call per type and then reads the records in a batch.
How do I backfill history?
The Search API and the list operations page through existing records with a modified-since filter, so a first run can load history before the polled triggers take over for ongoing changes.
Does it work with sandbox portals?
Yes. Each connection uses the token of a private app in one portal, so a sandbox portal and the production portal each get their own connection.
Bring one real integration
We will build it with you against your own systems, with the run log open.