checklist
How do you find and fix numbers stored as text after PDF to Excel?
Run one probe before you total anything: a SUM over the column returning zero while values sit visibly in their cells means the conversion delivered text wearing number costumes, and text adds up to nothing without complaining.
Why conversions produce text numbers
Converters inherit whatever characters the PDF contains. Thousands separators, currency symbols, trailing spaces, and non-breaking hyphens all ride along, and once attached, the safest reading of "1,234.56 " is a string, not a number. Excel displays it plausibly, which is the danger: a column of text numbers looks finished and totals to zero, and the zero propagates into every report built on top of it.
One status note first: you can open these pages and sign in to the workspace today, but that does not mean external platform accounts are connected, and it does not mean the product has formally launched. Everything below runs on the files you supply.
What you need before you fix
The converted workbook and the source PDF with its printed totals. The totals are the finish line: fixing text numbers is proven done when the recomputed SUM matches the figure on the page.
A scratch column for conversion formulas. Converting in place destroys the evidence of what the text looked like, which is exactly what you need when one pattern refuses to convert.
Step 1: Probe each money column with one formula
Put ISNUMBER on a sample row and SUM across the whole column. Left alignment is a hint but not proof; the formula is proof. A SUM of zero against visibly populated cells confirms the diagnosis in one step.
Step 2: Count the offenders precisely
Compare COUNTA against COUNT on the column: the difference is exactly how many text values hide among the numbers. Filtering to those rows gives you the full work list instead of a scrolling hunt.
Step 3: Diagnose the pattern before converting
Look at five offenders side by side. Common culprits are trailing spaces, thousands separators, currency symbols, and stray characters invisible in the cell. One pattern usually explains most of the column, and knowing which lets you fix causes rather than individual cells.
Step 4: Convert in a scratch column, reversibly
Build VALUE-based formulas, or use text-to-columns, in the scratch column beside the originals. Keep the original text until the new column verifies; multiplying-in-place tricks convert fast and destroy the audit trail just as fast. When the scratch column reads clean, replace the original and delete the scaffold.
Step 5: Re-total and tie back to the source
Re-run the SUM and compare against the printed total in the PDF. Then repeat the COUNT versus COUNTA probe to confirm nothing was skipped. The column is fixed when the arithmetic closes and the offender count is zero.
Verification
The fix holds when the column's SUM equals the source's printed total, ISNUMBER returns true on a fresh sample, and no cell needed a manually typed replacement. Typed-over cells are unconverted failures wearing a fixed costume, and they belong on the work list, not in the sheet.
Limits worth stating plainly
Tables are not cell-perfect. Check every number against the source page before you rely on it. A whole multi-page report converted in one pass is not reliable today. Rows from different periods can land in one cell and a column can be dropped. Work one table at a time. Audit-style statements with dollar signs, em-dash blanks, and parenthesised negatives are not handled. Data rows can be lost. Scanned PDFs with no text layer are refused rather than guessed at. There is no OCR path you can rely on here today. A document with no printed table is not something to convert. If you are handed a sheet built out of running prose, discard it and tell us.
Tablefold flags suspicious cells including values that arrived as text-shaped material, but flagging is where its job ends: the conversion decisions and the tie-back to the printed totals are yours, and the source PDF outranks the sheet wherever they disagree.
What Tablefold does in this workflow
Tablefold surfaces cells whose content looks off — including number-shaped strings carrying separators or symbols — so the probe in Step 1 starts from a flagged shortlist instead of a blind column scan.
It declines to silently coerce values during conversion, because quiet coercion is how a misread character becomes a plausible wrong number. You decide what each flagged value should be, with the source page open.
FAQ
Questions this guide is for
Why did my SUM return zero when the numbers are right there?
They are text, not numbers. Text participates in no arithmetic, and SUM treats the cells as empty. Probe with ISNUMBER, count offenders with COUNTA minus COUNT, then convert in a scratch column.
Is multiplying the cells by 1 an acceptable fix?
It converts, but it overwrites the original text and hides what pattern failed to parse. A VALUE formula in a scratch column converts the same cells while keeping the evidence until the totals verify.
The column passed SUM but one cell still looks odd. Does it matter?
Yes. A single text cell that escaped conversion means the pattern list was incomplete, and the same pattern likely appears in next month's file. Diagnose it now rather than meeting it again at reconciliation.
Start in the workspace
Convert with the suspects flagged
Sign in or create an account and you return to the Tablefold conversation. Upload the PDF and work from a grid where number-shaped text is flagged before it reaches your SUM.