Turning a PDF report into a chart
The numbers you most want to chart tend to arrive in the least chartable format. Extraction has got good enough to skip the retyping, provided you keep one verification step.
Text PDFs are easy, scans are not
A PDF exported from software has real text underneath and extracts cleanly. A scanned or photographed document goes through OCR, where 8 and 3 and misplaced decimals are genuine risks.
Extract the table, not the narrative
Point the extraction at the specific statement or table you need. Feeding a whole eighty-page annual report produces a mixture of numbers from different sections and different periods.
Reconcile one subtotal
Pick a subtotal from the document and check it against the extracted data. If it matches, structure was preserved; if it does not, a column shifted somewhere and everything downstream is suspect.
Watch for footnote adjustments
Financial statements restate figures in footnotes constantly. Extraction takes the number in the table, not the correction below it, so a quick scan of the notes is worth the minute.
Import from a PDF or image →Done carefully this turns an afternoon of retyping into about five minutes, which is most of the reason quarterly charts stop getting skipped.
Frequently asked
Can I make a chart directly from a PDF?
Yes, by extracting the table first. Text-based PDFs extract reliably; scans go through OCR and need a subtotal check before you trust them.
What should I verify after extracting from a PDF?
One subtotal against the original document, the units and period of every column, and whether footnotes restate any of the figures.