How to convert text-stored numbers to real numbers without excel's text to columns wizard
- Step 1Save your data as .xlsx — The inspector reads
.xlsxonly. If your data is a CSV you'd otherwise import column-by-column, open it in Excel and Save As an Excel Workbook first, then scan — the report will list which columns need converting. - Step 2Drop the workbook onto the tool — SheetJS parses it locally in your browser. The scan covers every sheet, so you get one consolidated list instead of checking sheets one at a time. No upload.
- Step 3Read which columns are affected — The report lines (
[Sheet] Cell: ...) reveal the affected columns at a glance — e.g. all the flagged cells are in column C and column F. That's your targeted Text to Columns plan; the count panel shows the total. - Step 4Run Text to Columns on just those columns — In Excel, select an affected column → Data → Text to Columns → choose Delimited or Fixed width (either works for single-column conversion) → Next → Next → set Column data format to General for numbers (or Date for text dates) → Finish. Excel re-parses every cell. Repeat only for the columns the report flagged.
- Step 5Or use the warning-diamond shortcut — For smaller selections, an even faster fix than the wizard: select the flagged cells, click the yellow warning diamond, choose 'Convert to Number'. The inspector's cell-level addresses make selecting exactly the right range easy.
- Step 6Re-scan to confirm everything converted — Drop the fixed workbook back in. 'No format mismatches found' means every column is done — no need to run the wizard on any more columns. That all-clear is your stopping point.
Inspector vs. running Text to Columns blind
Why scanning first beats wizard-on-every-column. The inspector is the diagnosis step, not a replacement for the conversion itself.
| Aspect | Text to Columns alone | Inspect first, then convert |
|---|---|---|
| Knowing which columns to fix | Guess; often run on all | Report names the exact columns |
| Multiple sheets | One selection at a time | All sheets scanned in one pass |
| Text numbers vs text dates | You decide per column | Report labels each ('number' / 'date') |
| Wasted passes on clean columns | Common | None — you only fix flagged columns |
| Who does the conversion | Excel wizard | Still Excel (inspector only diagnoses) |
| Confirming the job is done | Manual re-check | Re-scan returns the all-clear |
Which Text to Columns format to pick, by report label
Match the inspector's suggested type to the Step-3 column data format in the wizard.
| Report says | Example value | Text to Columns Step 3 format |
|---|---|---|
| looks like number | 1234, 1,234.56, -50 | General |
| looks like date | 2024/01/15, 01-02-2024 | Date (pick the matching D/M/Y order) |
(not flagged) $1,234 | currency text | Strip $ first, then General |
(not flagged) (1,234) | accounting negative | Replace parens with minus, then General |
(not flagged) 15 Jan 2024 | spelled month | Convert with DATEVALUE / manual, not the wizard |
Cookbook
Targeted-fix recipes: scan to find the affected columns, then apply the minimum number of conversions.
Three columns flagged across one sheet
Instead of running the wizard on all 12 columns, the scan shows only C, F, and J have text numbers. Three targeted runs and you're done.
Inspector report (excerpt, Sheet 'Import'):
[Import] C2: stored as text, looks like number ("500")
[Import] C3: stored as text, looks like number ("620")
[Import] F2: stored as text, looks like number ("12.5")
[Import] J2: stored as text, looks like number ("9000")
Plan: Text to Columns on columns C, F, J only
(General format). Columns A, B, D, E, G, H, I, K, L
are clean — skip them.Numbers in one column, dates in another
The report labels them differently, so you know to use General for the number column and the Date option for the date column.
Inspector report (Sheet 'Log'):
[Log] B2: stored as text, looks like number ("4500")
[Log] B3: stored as text, looks like number ("3300")
[Log] D2: stored as text, looks like date ("2024/01/15")
[Log] D3: stored as text, looks like date ("2024/01/16")
Plan:
- Column B -> Text to Columns -> General
- Column D -> Text to Columns -> Date (YMD)Warning-diamond shortcut for a small selection
Only four cells are text. Faster than the three-step wizard: select them and use Convert to Number. The cell addresses from the report make the selection precise.
Inspector report:
[Data] E10: stored as text, looks like number ("100")
[Data] E11: stored as text, looks like number ("250")
[Data] E12: stored as text, looks like number ("175")
[Data] E13: stored as text, looks like number ("90")
In Excel: select E10:E13 -> yellow warning diamond ->
'Convert to Number'. Done in two clicks, no wizard.Currency text the wizard alone won't catch
A clean first scan can be misleading if values carry a $. The detector skips $1,234 (the symbol breaks the pattern), so you strip the symbol first, then both scan and convert work.
Source column G has values like $1,234 (text).
First scan: 0 mismatches -> looks clean (misleading).
Fix order:
1. Find & Replace '$' and ',' -> nothing
2. Re-scan:
[Bill] G2: stored as text, looks like number ("1234")
3. Text to Columns (General) on column G -> real numbersMulti-sheet workbook turned into a fix plan
A workbook with five sheets is scanned once. The report shows only two sheets have problems, so you skip the other three entirely.
Inspector report (excerpt):
[Sales] C2: stored as text, looks like number ("500")
[Sales] C9: stored as text, looks like number ("480")
[Inventory] B5: stored as text, looks like number ("144")
Plan: only 'Sales' (col C) and 'Inventory' (col B) need
Text to Columns. Sheets 'Summary', 'Notes', 'Lookup'
returned no mismatches — leave them alone.Edge cases and what actually happens
CSV instead of XLSX
RejectedThe inspector reads .xlsx only. A CSV has no per-cell types to inspect (everything is text). Open it in Excel and Save As an Excel Workbook first; then scan to learn which columns need converting before you run Text to Columns.
Inspector doesn't do the conversion
By designThis tool diagnoses; it does not rewrite the workbook. Its output is a report of which cells/columns are text-stored numbers or dates. You apply the actual conversion in Excel (Text to Columns or Convert to Number). The value is precision — you only convert what's flagged.
Accounting parentheses negatives
Not detected(1,234) doesn't match the numeric pattern, so it's not flagged and Text to Columns (General) won't cleanly convert it either. Replace (/) with a minus sign first, then re-scan and convert. The detector and the wizard both need a clean numeric shape.
Currency or percent symbols
Not detected$1,234 and 50% carry a symbol that breaks the numeric pattern, so they aren't flagged. Strip the symbol with Find & Replace first; once the cell is bare digits-and-commas it's caught on the next scan and converts cleanly with the General format.
Spelled-out month dates
Not detected (and the wizard's Date option won't help)15 Jan 2024 isn't matched by the date detector, and Text to Columns' Date option expects numeric date parts. Convert these with DATEVALUE or by reformatting, not the wizard. The detector only flags numeric date shapes.
Leading-zero codes flagged
Expected00045 is pure digits and is flagged as 'looks like number'. If it's a zero-padded code you want to keep as text, do NOT run Text to Columns on it — General format would strip the leading zeros. The report flags it; you decide whether it's a value or an identifier.
Report capped at 100 inline lines
Truncated displayIf many cells are affected, the report shows the first 100 lines plus '... and N more.', and the count panel shows the total. That's fine for planning column-level fixes — you only need to see which columns appear, not every cell, to run Text to Columns.
Already-numeric cells
PreservedCells already stored as real numbers are skipped, so a clean column never appears in the report — which is exactly the point: you don't waste a Text to Columns pass on columns that are already fine.
Hidden sheets included in the plan
SupportedEvery sheet is scanned, hidden ones included. If a hidden sheet has text numbers, it shows in the report so your fix plan covers it — Text to Columns would otherwise require unhiding and selecting it manually.
File over the tier limit
Rejected (size limit)Free tier caps at 5 MB / 10,000 rows. A larger workbook is rejected before scanning. Upgrade to Pro (50 MB / 100,000 rows) or split by sheet. The conversion in Excel has no such cap, but the diagnosis step does.
Frequently asked questions
Is this a replacement for Excel's Text to Columns?
It's the diagnosis half. The inspector tells you exactly which cells and columns are text-stored numbers (and which are text dates) so you can run Text to Columns precisely instead of on every column. The conversion itself still happens in Excel — this tool doesn't rewrite your file; it makes the wizard faster by removing the guesswork.
Why is the inspector faster than just using the wizard?
Text to Columns works on one selection at a time and gives you no idea which columns are affected, so people run it on every column 'just in case'. The inspector scans all sheets at once and names the exact offending columns, turning a wizard marathon into a few targeted runs.
How do I do the actual conversion after scanning?
Two ways. For a whole column: Data → Text to Columns → Next → Next → set format to General (for numbers) or Date (for text dates) → Finish. For a small selection: select the flagged cells, click the yellow warning diamond → 'Convert to Number'. The inspector's cell addresses make the selection exact.
Which Text to Columns format should I pick?
If the report says 'looks like number', use the General column-data format in Step 3. If it says 'looks like date', use the Date format and pick the matching D/M/Y order. The report's label tells you which to use for each column.
Does it convert text dates too?
It detects them, labelling numeric-shaped text dates as 'looks like date'. You convert them in Excel with Text to Columns' Date option. Spelled-out months (15 Jan 2024) aren't detected and need DATEVALUE or reformatting instead.
Why didn't my $-prefixed values get flagged?
The $ (and %, and parentheses) break the numeric pattern, so those cells aren't flagged — and Text to Columns' General format won't cleanly convert them either. Strip the symbol with Find & Replace first, then both the scan and the conversion work.
What file types does it accept?
.xlsx only. It needs the per-cell type metadata XLSX stores to tell text from number. CSV has no type info. Save a CSV as an Excel Workbook first, then scan.
Will it convert leading-zero IDs and ruin them?
The inspector won't convert anything — it only flags. It will flag 00045 as 'looks like number', but that's a heads-up, not an action. If that column is a zero-padded ID, don't run Text to Columns (General) on it, because Excel would strip the zeros. Keep it as text.
Is my data uploaded?
No. The scan runs in your browser with SheetJS. Cell values and sheet names never leave your machine. Free tier needs no upload or account; only an anonymous usage counter may be recorded for signed-in users.
How big a file can I scan?
Free tier: 5 MB / 10,000 rows. Pro: 50 MB / 100,000 rows. Pro-media: 200 MB / 500,000 rows. Developer: 500 MB / unlimited. Over-limit files are rejected before scanning; split a large workbook by sheet if needed.
How do I know when I've converted everything?
Re-scan the fixed workbook. 'No format mismatches found' means every flagged column is converted — that's your signal to stop running the wizard. No more guessing whether you missed a column.
What related tools speed up the rest of the cleanup?
The date standardizer normalises mixed date formats in one pass, the formula-to-value tool flattens formula columns to static values, and the unpivot tool reshapes wide tables. Use this inspector to scope the type fixes first.
Privacy first
Every JAD Excel tool runs entirely in your browser using SheetJS and ExcelJS. Your spreadsheets, formulas, and data never leave your device — verified by zero outbound network requests during processing.