Map extracted fields to your destination

Configure how extracted document fields map to columns and properties in Notion, Google Sheets, or Airtable, including row layouts.

3 min read

On this page

Field mapping controls exactly how extracted document data flows into your connected sheet, database, or table. On the extractor feed's Mapping tab you connect each extracted field to the matching column or property, optionally clean values with enrichments, and choose how line items become rows.

Custom Mapping

Map any extracted field to any column or property

Enrichments

Clean or categorize values with rules before they are written

Itemization

One row per document, or one row per line item

Before you start

You need a Document Extractor feed with a destination chosen on the Destination tab. The fields available for mapping come from the Document tab: that is where you pick the document preset (Receipts, Invoices, and so on) and which fields the AI extracts.

Available Receipt Fields#

With the Receipts preset, these fields are extracted and available for mapping:

merchant

Store or vendor name

date

Transaction date

total

Total amount paid

subtotal

Amount before tax

tax

Tax amount

currency

Currency code (USD, EUR, etc.)

category

Expense category

items

Line item details

Other presets (Invoices, Transactions, Balances, and so on) extract their own field sets, and you can add or remove fields on the Document tab before mapping them here.

Setting Up Mappings#

The Mapping tab has subtabs: Fields, Enrichments, and (once a destination table is chosen) Advanced.

Configuration Steps

  1. Open the feed's Mapping tab

    From the Feeds tab, open your Document Extractor feed and select Mapping. Make sure you have already picked the sheet, database, or table on the Destination tab.

  2. Map each field on the Fields subtab

    For each extracted field, select the column or property it should write to. Unmapped fields are skipped, so map at least one field or no rows will be written.

  3. Clean values with enrichments (optional)

    On the Enrichments subtab, attach rules that fix or categorize values before they are written, for example normalizing merchant names.

  4. Pick a row layout on the Advanced subtab

    Set Itemization to Row Per Document (one row per receipt or invoice) or Row Per Item (one row per line item).

  5. Test with a real document

    Upload or forward a test document, then check that the new row lands with values in the right columns.

Choose a row layout (Itemization)#

LayoutWhat you getBest for
Row Per DocumentOne row per receipt or invoice, with totalsExpense summaries, tax prep, bookkeeping
Row Per ItemOne row per line item on the documentItemized budgets, inventory, per-product analysis

Destination-Specific Tips#

Notion#

Notion Property Types

Use Number properties for amounts (total, tax, subtotal), Date properties for the receipt date, and Select properties for categories. Title property works best for merchant names.

Google Sheets#

Sheets Formatting

Format amount columns as Currency and date columns as Date in Google Sheets for proper sorting and calculations. Add headers in row 1 for each mapped field.

Airtable#

Airtable Field Types

Use Currency fields for amounts, Date fields for dates, and Single Select for categories. Link fields can connect receipts to other tables like Projects or Clients.

Common Mapping Patterns#

Expense Tracking

Map merchant, date, total, and category for personal expense management

Business Expenses

Map every amount field (subtotal, tax, total) and use enrichments to tag departments or projects

Tax Preparation

Map tax amounts and categories for easy tax reporting

Itemized Records

Switch Itemization to Row Per Item to capture every line item as its own row

Missing Fields

If a document doesn't contain certain information (like tax on a tax-free purchase), those cells are left empty in the row. Unmapped fields are always skipped.

Next Steps#

Set up the Document Extractor

Use this page with your AI assistant

Every BankSync doc is available as plain Markdown for agents and LLMs.