Import Paige documents into HubSpot from CSV
Export a batch of documents from Paige and load the one-row-per-document CSV with HubSpot's import tool — contacts, companies, deals, or custom objects created or updated from Paige's fields, mapped column by column, no subscription.
Before you begin
- A HubSpot user with import permissions
- A Paige scan job whose document type's fields match the HubSpot properties you want filled
At the end of this guide, a batch of documents that Paige read becomes a batch of HubSpot records — contacts from intake forms, deals from signed orders, whatever your document type holds — with properties filled from Paige's CSV, in one import.
HubSpot's importer takes a CSV with a header row and maps each column to a property; that is exactly Paige's one-row-per-document CSV. Files aren't part of the import — for the PDFs on the records, use the Zapier path.
Step 1 — Get the CSV out of Paige
In Paige, open Documents, click Export, and in the Export Documents dialog choose the Scope, the Filters (Reviewed only, Field Data with fields, Status not yet exported), tick Download under Destinations, and confirm. Paige emails a link; the zip contains the PDFs and the data files, including the one-row-per-document CSV.
Open it in a spreadsheet. What you should see: a header row with the PDF file name, the document type, and one column per Paige field, then one row per document. Delete the columns HubSpot shouldn't see (the file name and IDs) or leave them and skip them during mapping.
Name Paige's fields the way HubSpot names its properties (Email, First Name, Company Name, Phone Number) and HubSpot maps them automatically.
Step 2 — Import
In HubSpot, open Data Management > Imports (from Settings, or More in the navigation) and click Import data.
Choose Advanced imports (all objects), then the object (for example Contacts), and click Next.
Choose Upload a file, select Paige's CSV, and continue.
On the mapping step, check each column's Import As property. Set anything you don't want to Don't import column. Create a new property on the spot if the document has a field HubSpot lacks.
Give the import a name, review the options (for example whether to update existing records that match on email), and click Finish import.
What you should see: the import completes with the number of records created and updated, and the records appear in the object's list with the properties filled.
When it doesn't work
| Symptom | Where to look |
|---|---|
| Rows skipped | A required property is empty, or a value doesn't fit (a text date into a date property). Check the import's error file. |
| Duplicates created | HubSpot de-duplicates contacts by email; for other objects, include a unique ID property and choose update-on-match. |
| Wrong columns imported | Set the file-name and ID columns to Don't import column. |
| The same batch imported twice | Use Paige's Status: Not Exported filter when exporting and export each batch once. |
If a Paige field is consistently misread, improve its extraction guidance: Write extraction guidance the AI can follow.