You have a spreadsheet and no Excel. Upload it to the Excel viewer and it opens in this tab: XLS, XLSX, CSV, ODS and the macro and binary variants, up to 20 MB, deleted once you close it.
Then one thing worth understanding, because it accounts for most of what looks wrong in a spreadsheet you did not write: a viewer does not recalculate. Everything below follows from that.
The numbers on screen are the numbers Excel saved
Every formula cell holds two separate things: the formula, and the result Excel worked out the last time it calculated. Anything reading the workbook without a calculation engine reads the second one. Usually you never notice, because the saved result is correct. Three situations make it visible.
Calculation was set to manual. Under Formulas, Calculation Options, someone chose Manual — normally to stop a heavy model recalculating on every keystroke — and then saved. The stored values are whatever was on screen at that moment, which can be hours or years out of date, and they are wrong in a way that looks entirely normal.
A machine wrote the file. Workbooks generated by a reporting tool or a script often contain formulas with no cached result at all, because the library that wrote them has no way to evaluate a formula. Those cells come up blank or zero. Nothing is corrupt; nothing has ever run them.
The function was volatile. =TODAY(), =NOW(), =RAND() and =OFFSET()
recompute every time Excel opens the file, so what is stored is the value at the
instant of the last save. A sheet headed "as of today" is telling you when somebody
last pressed Ctrl+S.
A link to another workbook is the same problem, one file further away
A reference like ='C:\Reports\[Q3-Actuals.xlsx]Summary'!B4 points at a file on
somebody else's disk. You uploaded one workbook, not the folder it lived in, so the
cached figure from the last successful refresh is what appears. When the link was
already broken before you received it, #REF! is stored in the cell and #REF! is
what renders — an error saved as data, not an error happening now.
Where the grid actually ends
The two Excel formats have very different ceilings, and the difference bites when somebody saves down. XLSX holds 1,048,576 rows by 16,384 columns, the last of them column XFD. The older XLS stops at 65,536 rows and 256 columns, ending at column IV. Save a modern workbook as .xls and everything past those limits is discarded at the moment of saving. A viewer cannot show you rows that are no longer in the file.
Dates are numbers wearing a costume
A date in a spreadsheet is a count of days that a number format displays as a date. Clear the format and the date becomes an integer again. Two consequences follow.
Workbooks authored in older Mac versions of Excel may use the 1904 date system rather than the 1900 one. Open such a file under the other convention and every date moves by 1,462 days — four years and a day — while every value stays technically valid.
And Excel counts 29 February 1900, a day that did not exist, because Lotus 1-2-3 did and compatibility mattered more than the calendar in 1985. Any date arithmetic that crosses that point in a workbook carries the extra day.
Things that look like damage and are not
- A cell showing
#####is a column too narrow for its formatted number. Widen it in the source file and the number is there; the value was never affected. - A pivot table renders from the pivot cache stored alongside it. It is a snapshot of the source data at the last refresh, which is why a pivot can disagree with the sheet it was built from.
- An .xlsm opens and its macros do not run. Rendering a workbook is not executing it, and a viewer has no reason to run anyone's VBA.
- Sheet protection is not encryption. A protected sheet is a flag saying "do not edit" and it displays normally. A workbook with a password required to open it is genuinely encrypted, and nothing will render it without that password.
If spreadsheets are the format that keeps stopping you, the Excel viewer is the page worth a bookmark.