Integrate Hubspot with external system and save data in hubspot

Question

I know we can integrate Hubspot with an external system and display that data as cards. But can we store the same data in Hubspot? The idea behind this is to Save the data in Hubspot for reporting purposes. So that we can compare records based on other factors.

So is it possible to store external system data in Hubspot?

Answer 1

Yes, you can push the data using API endpoints provided by hubspot. If you want to update data in some custom field/property you can create that in your hubspot account.

Answer 2

Yes. Hubspot provides multiple ways to store custom data. Depending upon your implementation

1. You can create your own custom properties and add data to them from HubSpot CRM or by using API. https://knowledge.hubspot.com/crm-setup/manage-your-properties
2. On Enterprise accounts you can use Custom Objects and create your own object. Hubspot exposes API endpoints for your custom objects as well. https://developers.hubspot.com/docs/api/crm/crm-custom-objects

Answer 3

[ EDIT: Apologies. I misread your question and described a reverse scenario. You need to store data into Hubspot. I described pushing them out. ]

In my company we had a similar need. Besides reporting we also wanted to do some validation, enrichment and distribute smart notifications to wider business operations team when sales team was updating data (typically deals).

The architecture we went with:

1. We used Webhooks API to push updates/changes to an API endpoint running on AWS API GW.
2. We used AWS Lambda function to validate authenticity of incoming messages and pushed their payload to a SQS queue
3. We implemented business logic running containerized on Fargate which consumed from queue in small batches. Deduped, enriched, distributed to other systems. It also pulled/pushed additional data via Hubspot REST API

The solution worked without any major problems for 2-3 years and is probably still operational today.

Few complications that I can remember:

• At the time Hubspot did not publish information about IP ranges from which they send Webhook notifications. We could not IP whitelist our API and had to open it to the wide world. Maybe they do publish it now but I could not find it.

• There was no possibility of certificate-based authentication for Webhooks. The only line of defense was the lambda-based validator checking for X-Hubspot-Signature. See section Security in the docs

• Your receiving endpoint must return the response to Hubspot in 5 seconds. See section Retries in docs. With some communication overhead you'll get about 2 seconds for parsing, signature validation, response building and shooting the notification payload into a queue. So don't try to put any logic into the lambda validator. Do it later down the pipeline.

• Our lambda was Java based. On AWS a cold lambda can easily get 0.5 second startup time further cutting down time remaining for that 5 seconds response. Maybe other languages are better, or you can keep your lambda warm (or your sales team will be so effective the lambda never cools down! :) )

• Inevitably you are going to miss sending timely response here and there. Hubspot then makes a retry which leads to data duplication. Build some deduplication logic after that queue and you are good. Also possibly monitor the duplication rate. You don't want to pay for constant retries. There are only 10 retries, by 2nd or 3rd your lambda should be warm and catch the incoming notification on time. But if you miss all 10 retries you are going to lose that data. So monitor your system is responding in a timely matter.