If the deal is in HubSpot, the order is in one sheet and the forecast is in another, and the three never quite match, I connect them so there's one set of numbers. HubSpot and your other systems agree, and if a sync ever fails, someone is told.
Data
Objects, properties and associations designed around how you actually sell, then the duplicates and gaps cleaned out.
Data and CRM architectureIntegrate
ERP, help desk and delivery tools talking to HubSpot, so nobody rekeys the same customer twice.
HubSpot integrationsAutomate
Approvals, renewals and handoffs that run on their own, without anyone chasing.
HubSpot implementationApps
Custom cards, app pages and portals, so your team works in one place instead of five tabs.
Custom apps and cardsAI
HubSpot's AI and agents working on data they can trust, plus Claude and ChatGPT set up for your team.
HubSpot AI readinessA HubSpot integration moves data between HubSpot and your other systems, such as your ERP, another CRM or your help desk, so nobody types the same thing twice. I'm an independent HubSpot Gold Solutions Partner. I use HubSpot's own connectors where they do the job, and build custom where they don't.
Whatever holds the other half of the customer picture. For most B2B firms that means some of these:
A simple check: if someone copies data between two systems by hand, or a report needs exports from both, it probably wants an integration.
Before I build anything I agree one system of record for every field that crosses between systems. Invoice status belongs to the ERP. Deal stage belongs to HubSpot. Each field syncs one way, from its owner, so the systems don't fight over the same data.
Most integration problems I see come back to this step. Matching on company name instead of a domain or account number creates a new duplicate every time someone spells it differently.
| Option | When it fits | Trade-off |
|---|---|---|
| Native app from the HubSpot Marketplace | It syncs the records and fields you need at your volumes | Limited control over mapping and matching rules |
| Zapier or Make | Simple triggers, low volume, one record at a time | Gets fragile and costly as it grows |
| n8n | Multi-step jobs, many records at once, document generation | Needs hosting and someone to own it |
| Workflow webhooks | HubSpot needs to tell another system something has happened | Needs Data Hub Professional |
| Private app and HubSpot APIs | Scheduled syncs, bulk loads and anything the others can't handle | Custom code, so it needs documenting and maintaining |
| Card or app page inside HubSpot | Users need to trigger the other system from a HubSpot record | A custom build, released properly |
I always check the native option first. A connector HubSpot maintains is cheaper to run than anything I build.
A UK manufacturer, one of the largest in its field worldwide, sends each customer site a test report after every visit, in the customer's own language. Those reports used to be put together by hand across several tools.
Now a button on the company record produces the report as a PDF in up to six languages, files it in HubSpot and records that it's done, in about 30 seconds. If a step fails, it retries, then shows the team a failure they can act on. Together with the rest of the project, that gave the team two weeks of work a month back. See the case study.
A London immersive art studio, with work touring internationally, needed revenue figures HubSpot can't calculate on its own. An n8n workflow reads each deal's line items and writes the financial-year splits and recurring totals back onto the deal. Leadership now sees revenue by product and when the money arrives. Read more about the revenue reporting work.
Sometimes it makes more sense to move the other system into HubSpot than to connect it. For an international technology firm, the support desk moved from Zendesk into HubSpot Service Hub: about 62,000 past tickets with their conversations and attachments, and no open tickets lost. That's covered on the Zendesk migration page.
If the integration needs a screen inside HubSpot, that's custom development. If you're not sure your data model is ready for it, start with data and CRM architecture.
The ERP, HubSpot and the help desk, each with its own copy of the customer. Someone carries the changes across by hand, slowly, and some of them go in wrong. Join the systems and the same records flow by themselves.
Use the native connector if it covers what you need, because it's cheaper to run and HubSpot maintains it. Go custom when it can't handle your records, your volumes or the fields that matter. I'll test the native option first.
n8n is a workflow automation tool, similar to Zapier but able to run longer, multi-step jobs that retry and report when something goes wrong. I use it when a job is too big for Zapier and doesn't need custom code.
It should tell someone. I build integrations to retry where it's safe, then show a failure on the HubSpot record or alert a named person. A lot of the clean-up work I see comes from integrations that stopped without anyone knowing.
Usually an API user or key with the narrowest permissions the job needs. Credentials are stored securely and never sit in code or files. Your data stays in your systems.
Book a short call. Tell me which systems you run and where the rekeying happens, and I'll tell you how I'd connect them.