An original resource by Waseet Mahyudin
PDF to Excel conversion: a checklist for usable data
Turn suitable PDFs, scanned tables and images into structured spreadsheets. Define the columns, preserve source references and review uncertain text before delivery.
Free to reuse for your own work. No signup or file upload needed.
Who this helps
For businesses hiring help with PDF conversion, document data entry or image-to-spreadsheet preparation.
A conversion brief and acceptance checklist for an Excel, CSV or Google Sheets deliverable that can be checked against the source.
1. Inspect the document before agreeing the task
A PDF is a container, not a guarantee of extractable tables. Selectable text, scans and mixed layouts need different amounts of checking. Review representative pages before agreeing scope.
- Record the file count, page range, language and whether pages contain selectable text, scans or handwriting.
- Check a sample that includes repeated headers, multi-line cells, page breaks and the least legible page.
- Confirm permission to process the documents and agree a secure transfer method. Start an enquiry with a description, not confidential files.
- Agree which tables or fields are required. Exclude decorative text, footnotes or images only when their meaning is not needed.
2. Define the spreadsheet you need
A visually similar sheet may still be hard to sort or import. Specify what one row represents and how the output will be used.
- Choose a row unit, such as one product or one document line item, and supply the required column names.
- Keep identifiers and codes as text when leading zeros matter. Agree decimal separators, date conventions and units.
- Include source filename and page references so a reviewer can locate the original information.
- Agree Excel, CSV or Google Sheets delivery. CSV cannot preserve multiple worksheets, formatting or review comments without separate files.
3. Extract, then check the difficult cases
OCR and AI-assisted extraction can help with suitable documents. Their output needs human review; a plausible-looking value is not evidence that the source said it.
- Check common substitutions such as O/0 and I/1, dropped minus signs, decimal points and wrapped cells.
- Flag illegible or ambiguous values instead of guessing. Keep missing, unreadable and deliberately blank values distinguishable.
- Compare row counts by source section and check totals only where the document provides a meaningful reconciliation.
- Review agreed critical fields and representative records across every layout; expand checking when a repeated error is found.
4. Deliver the table with its review notes
The handover should explain what was converted and what still needs a decision. File conversion is not verification that the original facts are correct.
- Provide the structured file, field definitions, source-page mapping and an exception list.
- Reopen the export in the intended tool and inspect text encoding, long identifiers, dates and formula-like strings.
- State the pages processed, exclusions, review method and unresolved fields; keep original documents unchanged.
- Ask the named reviewer to approve the agreed use. A successful file open alone does not establish data accuracy.
Try a small example
Conversion is only the beginning.
A tidy table can still contain a wrong value. See how source checks and unresolved fields affect a handover.
Make the columns understandable
Field definitions
Adapt these definitions to the brief. All examples below are illustrative.
- Source reference
File, page and original table or row location.
Example: catalogue.pdf · page 4 · table 2 · row 7
- Row unit
The item represented by one output row.
Example: One product variant, with size and colour in separate columns.
- Extraction status
Whether a value is readable, missing or requires review.
Example: Review needed: printed code could contain a zero or the letter O.
- Target format
Output tool, column types and conventions.
Example: CSV, UTF-8; product codes as text; units in a separate column.
Illustrative decision / not a client result
A decision in practice
Illustrative situation: a scanned product code reads 001O8, but the fourth character is unclear.
Action
Preserve the source reference and flag the code for confirmation.
Reason
Replacing O with 0 without evidence could match the wrong product, and numeric formatting could discard leading zeros.
Follow-up
Request a clearer source or the client's confirmation; record the decision before approving import.
Acceptance checklist
Use these as review questions. A checkbox only means something after the work behind it has been checked.
- The agreed pages and tables are accounted for.
- Output columns and row units match the brief.
- Identifiers, units and dates retain their intended meaning.
- Uncertain values and source references remain visible.
- The exported file opens correctly and the review scope is documented.
What this does not establish
Feasibility depends on source quality, layout and authorized access. Handwriting, damaged scans and complex layouts may require additional review or a narrower scope. No automatic extraction method guarantees error-free results.
Try a related demonstration
Inspect the approach in an original interactive project. It uses synthetic data and is not a record of paid client work.
Data cleaning studioPut the brief to work
Need help with your next project?
Tell me the outcome, the source material and the constraints. Start with a description; keep sensitive records out of the enquiry form.
Discuss your project