I am using the API via a synatic feed to update or add companies and contacts into hubspot.
I am using a sync that first looks in Hubspot for a custom identifier and if it exists it then takes the record down one route to update the record. If the ID dosent exists it then goes down a different route to add the record into the system.
I am loading around 9,000 company records with around 200 unique fields each and around 8,000 contact records with 5 unique fields each.
The first full load for the 9,000 company records took 2 hrs 30 minutes hours, a second load on a subsequent day took around 2 hrs 20 minutes.
The first load for the contact records took 12 hrs and a subsequent load is at 7 hrs and counting, and looking at the contacts processed, its likely to be around 12 hours again.
For contacts after the load, I am then running a workflow in Hubspot to link the Company and the contact using a unique id.
Do these times seem reasonable for the amount of records that I am loading into Hubspot. The information has a level of volatility and will be changing on a regular basis so we need to carry out regular loads, but if contacts take 12 hours to import, we will just end up continuously updating them.
I can potentially look some processes to only send the records that have changed to Hubspot, so once an initial feed is done it shoudl be quicker, but I just wanted to check if other users have a similar timescale for importing records.
Hi @BMI
Thank you for reaching out to the Community!
I’d like to invite some community members who are subject matter experts to join this conversation. @sylvain_tirreau@SteveHTM@MichaelMa - Would you be able to share any insights on this? Your expertise would be greatly appreciated.
Best,
Victor
@BMI - although this is now marked as solved, I am interested to dive deepr on this. Are you using the upsert batch functionality or loading contacts one by one after your filtering. I’ve previously run tests on import times (although specifically for HubDB records) and can give some calibration approaches to understand how to optimize the workflows if helpful.
It sounds like you’re doing an API request to search for the record and then another request to update the record. So 2 requests per record. Are you doing this sequentially (eg, you proccess each record one at the time)?
If so, then yes, I can see that the times you have being reasonable. The back and forth of searching for the record, waiting for response, sending the update and getting the update response before going to the next record will be slowing you down quite a bit.
You can probably build in efficiencies, like the one you mentioned of only processing modified records after the initial load, to ensure a speedier update.
Have you looked into using Upsert instead of searching for the record? It seems like this would reduce the loading time by quite a bit since you might be able to batch update/create