# Extract Data from PDFs to Excel, CSV & JSON | DigiParser

Source: https://www.digiparser.com/solutions/extract-data-from-pdf

[Solutions / Extract Data from PDF](/solutions)

PDF DATA EXTRACTION - A PRACTICAL GUIDE

# Extract data  
from PDFs.

Skip the copy and paste.  
Get the fields and rows your team needs.

Turn customer POs and supplier invoices into structured data for Excel, CSV or a configured system. Start with a useful output, then build a repeatable process.

[Try your PDFs](https://app.digiparser.com/auth/join) [Follow the setup guide](#plan)

YOUR PDF BATCHExample

invoice-a.pdf**3 item rows**

invoice-b.pdf**2 item rows**

invoice-c.pdf**4 item rows**

Extract - review - export

**One useful spreadsheet.**References + item data

**9**item rows  
across 3 documents

Illustrative row count: 3 + 2 + 4 = 9.  
Check each document, not only the batch total.

Digital and scanned PDFs Fields and repeating tables No code needed to start

THE SHORT ANSWER

## Choose the output before you upload.

For recurring order or invoice PDFs, use DigiParser to define fields, extract repeating rows, review the result and export the data. The right test is whether your team can use the output in its next task.

[See a document-to-data example](/solutions/pdf-parser) [Find your export option](#export)

YOUR SETUP GUIDE

[01Choose your fields](#plan)[02Prepare the batch](#batch)[03Check every record](#review)[04Choose the export](#export)[05Make it repeatable](#repeat)

**Keep the record intact.**

A three-page invoice is one document. A PDF containing three invoices needs three separate records.

01

START WITH THE TASK

## Choose the fields you actually use.

Open the spreadsheet or system where the information belongs. Use its required columns to decide what to extract. Keep customer orders and supplier bills separate.

SELLER / CUSTOMER ORDER

### Customer PO -> order data

*   Customer name and PO reference
*   Ship-to and requested delivery date
*   Item code, quantity and unit
*   Price and repeating item rows

[Customer PO workflow](/solutions/purchase-order-parser)

BUYER / ACCOUNTS PAYABLE

### Supplier invoice -> AP data

*   Supplier and invoice reference
*   Invoice date, due date and currency
*   Item rows, amounts and tax
*   PO reference when present

[Supplier invoice workflow](/solutions/invoice-parser)

**Use types that preserve meaning.**

Keep IDs such as `00124` as text. Use numbers for quantities and amounts, dates for document dates, and a Table field for repeating rows. A missing delivery date should be checked, not silently replaced with the invoice date.

**Take a field plan to your first test.**

Editable CSV starter worksheets. Adapt the fields and checks to your documents.

[Download PO field plan](/downloads/pdf-extraction/customer-po-field-plan.csv)[Download invoice field plan](/downloads/pdf-extraction/supplier-invoice-field-plan.csv)

[Field and table setup instructions](https://www.digiparser.com/docs/guides/working-with-parsers)

02

SORT THE RECORDS, NOT EVERY LAYOUT

## Give each document the right route.

Different layouts of the same document type can use the same field definitions. Different kinds of documents need a routing decision before extraction.

Normal batch

### Same type, separate files

Upload the files to the appropriate parser. Check representative supplier or customer layouts against its fields and table columns.

Separate or classify

### Mixed document types

Create separate parsers. Sort the batch yourself, or configure a Document Classifier to route each type and define what happens to unmatched files.

Split the pack

### One file, several records

Use PDF split options for a combined pack. Check that every resulting document starts and ends on the correct page.

Preserve the record

### One record, several pages

Keep the pages together. Check item rows across page breaks, including long descriptions and repeated headers.

Document Classifiers require Pro or a higher plan, or an active full-feature trial. Classification and AI splitting add credit usage; check the current guides before scaling.

[Upload a batch](https://www.digiparser.com/docs/guides/document-processing/uploading-documents)[Configure routing](https://www.digiparser.com/docs/guides/document-classifiers)[Split combined PDFs](https://www.digiparser.com/docs/guides/how-to-split-documents)

03

CHECK THE COMPLETE RESULT

## Nine rows can still hide a missing item.

A batch total does not tell you which document contributed each row. Use a per-document check so an extra row in one file cannot hide a missing row in another.

Scroll the example table to see every column.

Illustrative batch check - not measured product output

Source file

Expected rows

Extracted rows

Result

invoice-a.pdf

3

4

Extra row: inspect

invoice-b.pdf

2

1

Missing row: inspect

invoice-c.pdf

4

4

Count matches; check values

Batch total

9

9

Same total, two records need review

Method: fictional expected counts 3 + 2 + 4 and extracted counts 4 + 1 + 4 both sum to 9. Equal counts do not prove correct values or complete records.

*   **References and dates:** make sure each record belongs to the correct document.
*   **Rows and units:** compare item codes, quantities, prices and units with the source.
*   **Exceptions:** name the person who fixes missing values or unclear scans.

Configure assigned review stages when automatic exports must wait for approval. All required stages must be approved before the configured export runs.

[Set up Reviews & Approvals](https://www.digiparser.com/docs/guides/reviews-approvals/overview)

04

MAKE THE OUTPUT USEFUL

## Choose what a row should represent.

A file format alone does not define the export. Choose whether you need document fields, item rows or separate table sheets.

[![DigiParser Export Data dialog with XLSX, CSV and JSON format choices and options for fields, separate tables or flattened fields and tables.](/_next/image?url=%2Fimages%2Fsolutions%2Fpdf-extraction%2Fexport-dialog.png&w=3840&q=75&dpl=dpl_CPQnEhggmYn43eUERpeVtf8KHiV2)](/images/solutions/pdf-extraction/export-dialog.png)

Real product screen from the public documentation. Open the full-size image to inspect the choices.

01

### A list of orders or invoices

**Choose All Fields**

Use document references, dates and totals without item-table rows.

02

### A working sheet with item rows

**Choose All Fields and Tables (Flatten)**

Use CSV or XLSX. Select the needed columns and inspect the resulting row structure.

03

### Document fields and separate item sheets

**Choose XLSX -> All Fields and Tables**

Use separate sheets for tables when one combined sheet is not the right structure.

04

### Structured data for another system

**Choose JSON or XML**

Keep the structure your receiving software needs; test its import or mapped record.

**Export the right set of documents.**

Apply a folder, status, date or search filter, or select documents and use Export Selected. Use Customize to include only the columns your receiving sheet needs.

[Follow the export instructions](https://www.digiparser.com/docs/guides/managing-data/exporting-data)

05

FROM A TEST TO A WORKFLOW

## Automate the route that already works.

Start with a representative batch and a file your team can use. Then connect the document source and destination that remove recurring manual work.

### In a small office

The person entering orders or bills can own the sample check and export. Keep a short exception list and confirm who approves the pilot.

### Across a larger team

Assign extraction review and destination setup to named owners. Agree how to handle revised documents, failed handoffs and changes to field definitions.

**Extraction gives you data. A destination turns it into a record.**

Customer POs may feed order tracking or a sales-order record. Supplier invoices may feed an AP sheet or accounting bill. A connected system needs supported record types, field mapping and a successful test write; matching, accounting decisions and payments remain downstream.

[Explore configured integrations](/integrations)

### Before expanding, answer three questions.

1.  Are the fields and item rows complete enough for the next task?
2.  How much correction and review work remains?
3.  Does the actual file or destination record work for its user?

[Build a recurring document workflow](/solutions/automated-data-extraction)[Manufacturing workflows](/solutions/for-manufacturing)

PDF EXTRACTION QUESTIONS

## Clear answers.  
Before your first batch.

Files, fields, tables and costs.

[Read the credit guide](https://www.digiparser.com/docs/guides/troubleshooting/credits-and-billing)

How do I extract data from a PDF?

Create a parser in DigiParser, choose a document type and define the fields and table columns you need. Upload a representative PDF, compare the extracted results with the source, then export CSV, XLSX, JSON or XML. Test a batch before connecting a recurring workflow.

Can I extract data from scanned PDFs?

Yes. DigiParser accepts scanned PDFs. Use complete, readable files and compare fields and item rows with the source. Check faint characters, cropped edges, quantities and units; a scan can need correction or a better source.

How do I extract a folder of PDFs into Excel?

For files of the same document type, add several files to the appropriate parser and check the processed results. For a normal spreadsheet, use XLSX with All Fields and Tables (Flatten). Mixed document types need separate parsers or a configured Document Classifier. Classifiers are available on Pro and higher plans, and during an active full-feature trial.

Should I split a multi-page PDF before extraction?

Keep the pages together if they belong to one order or invoice. If a combined file contains separate records, use the upload split options and check each new document's page boundaries. Splitting one multi-page invoice into individual pages can separate its reference from its item rows.

What is the difference between OCR and PDF data extraction?

OCR turns text in a scan into text that software can read. Structured extraction assigns information to named fields and table columns. For order or AP work, you usually need references, dates, quantities and repeating rows rather than a block of raw text.

Can DigiParser send the data to our existing system?

DigiParser offers configured integrations and an API in addition to file downloads. The destination needs field mapping, supported record types and a tested result. Customer and item IDs, accounting choices, related-document matching and payment stay in the relevant downstream workflow.

How much does PDF data extraction cost?

DigiParser uses page credits. Document length and enabled processing options affect credit usage, and reprocessing uses credits again. Use the current pricing and credit guide to estimate your actual files instead of assuming that one PDF always costs one credit.

START WITH YOUR NEXT REAL TASK

## Stop moving values by hand.  
Start with one useful batch.

Bring your PDFs and the columns your team needs.  
DigiParser handles the extraction; you define what useful means.

[Start with DigiParser](https://app.digiparser.com/auth/join) [Discuss your workflow](/schedule-demo)

[PDF to Excel](/solutions/pdf-to-excel)[PDF to CSV](/solutions/pdf-to-csv)[Scanned invoice tables](/solutions/extract-tables-from-scanned-invoices)[Pricing](/pricing)