TableXL / Practical guide

Convert PDF table to Excel for calculations you can check

Get a text-PDF table into Excel, then convert checked numbers yourself while preserving identifiers, units and the source table structure.

Start with an editable source if one exists

To convert a PDF table to Excel, first look for the original workbook or a CSV supplied with the report. Those formats avoid reconstructing cell boundaries from page positions. If PDF is the only source, test whether you can select individual words in the table. Where your Excel installation includes the PDF connector, choose Data, Get Data, From File, From PDF and inspect the detected tables. Select Transform Data to review the structure before loading it. This can avoid purchasing a separate converter when the source is straightforward.

Recover the logical columns

Compare the extraction with the printed header before changing data types. Reports often use a group heading above several year or region columns. Preserve that meaning in explicit column names such as Revenue 2025 and Revenue 2026 rather than leaving identical Revenue headings. Remove a title or footnote only after confirming it is outside the data. Check whether a line break is part of one description or a new record. A table can look aligned in the worksheet while values have moved into the adjacent field.

Turn verified text into usable numbers

Keep a raw extraction sheet and perform calculations on a working copy. Set product codes and account identifiers to Text before loading. Convert numeric columns with a locale that matches the printed decimal convention. A value such as 1,250 might mean a whole number or a decimal depending on the report. Keep units in the headers: a value printed in thousands should not silently become an ordinary currency amount. Do not replace a dash with zero unless the report defines that dash as zero rather than missing or inapplicable.

Combine pages only after matching schemas

Compare the last row before a page break with the first row after it. Continuation tables may repeat headings, add a subtotal or reorder columns. Append only tables representing the same fields and units. Keep unrelated tables on separate sheets, even if they have the same number of columns. Add a page reference to a combined analysis table so an unusual value can be checked later. If a row count drops after removing repeated headings, verify the removed rows individually rather than trusting a broad text filter.

Validate the analysis table

Count source records and output records, check a few ordinary rows and every unusual boundary, and recompute printed totals from the extracted detail. A total agreeing does not prove that identifiers and descriptions stayed with the correct amounts. Check negative values, percentages, footnote markers and blank cells separately. If Excel's connector is unavailable or the detection fails, Tabula offers a free local way to select text-PDF regions and export CSV. Import that CSV deliberately, retaining the original PDF and the raw extraction before making further edits or calculations.

Separate reported values from calculation columns

Consider a report with columns headed Region, Sales (£000) and Margin (%). Keep the printed sales and margin text in a raw sheet. In the analysis sheet, document that 125.4 in the sales column represents 125,400 pounds, while 12.5 in the margin column represents 12.5 percent rather than 12.5 whole units. Convert each column under its own rule. Test one row by hand before applying the rule to the table. A grand total might reveal a scale error, but a percentage column can be wrong without affecting that total, so verify units independently.

Use a shape check before a pivot

Make sure every intended data record has the same fields after extraction. A report may show a merged group label once above several detail rows; decide explicitly whether to repeat that label in a new analysis column. Keep original detail labels so repeated names in different groups remain distinguishable. Exclude reviewed subtotal rows from the pivot input, otherwise detail and subtotal may both be counted. Filter each required column for blanks and conversion errors, then compare group counts with the printed table. Keep this rectangular analysis sheet separate from the visual source layout and from unrelated tables elsewhere in the report.

Where TableXL fits

TableXL extracts cell text for review rather than guessing number or date types. You can edit cells and headers, merge adjacent columns or split a column at a literal separator; row joining remains manual. XLSX gives each detected table its own sheet. One PDF per conversion may contain up to 15 MiB, 100 pages, 3,000 rows and 40 columns. Encrypted PDFs, handwriting and complex layouts may fail. The trial converts the first two pages with data found within the first six pages; remaining successful pages require credits. Download your checked work before sources and results expire within 24 hours.

Check two pages with data free See page packs

Written by Zijian Xiao. Examples are synthetic. Review all extracted results against your original PDF.