Not the job you came for? Tell me what you need
The chart is the easy part. Everything that makes a dashboard wrong happens before you get to it — and it's all in the data you started with.
Search for a dashboard template and you'll find thousands, all free, most
of them fine. Almost none of them are the reason people give up. You
download one, paste your export into it, and the totals are wrong, or the
chart has one bar per row, or a column that should add up says
0. The template was never the problem.
Here's what's actually in the way, in the order it bites.
This is the commonest one by a distance. An export writes
05/04/2026, 5 Apr 2026 and
2026-04-05 in the same column — often because the rows came
from different systems or different people — and Excel reads some as dates
and some as words.
A month breakdown then covers only the rows Excel understood, and it says so nowhere. The total at the top looks right, the chart underneath is missing a third of the year, and nothing on the sheet contradicts anything else.
What to do: before anything else, select the date column and check the alignment. Excel right-aligns real dates and left-aligns text. If any of them are hugging the left edge, they aren't dates, and no amount of formatting will change that — formatting changes how a value is displayed, not what it is.
Most exports end with a TOTAL line, and plenty have subtotals
part-way down. Paste that into a dashboard and every figure is
exactly double what it should be, because the total is
being added to the rows it already contains.
Double is the lucky version. With a few subtotals scattered through, you get a number that is wrong by an amount nobody can work out by eye, and it looks entirely plausible.
What to do: sort by any text column and the total rows group together at one end, usually with blanks beside them. Delete them before you build anything on top.
Some numbers are meaningless added up, and a spreadsheet will never say so. Three kinds catch people out:
What to do: for each numeric column ask "if I had twice as many rows, should this figure be roughly twice as big?" If not, it isn't something to total.
You group by a column and get a bar for every single row, which is a list wearing a chart's clothes. It happens when the column you grouped on barely repeats — an invoice number, a reference, a timestamp.
Timestamps are the trap here, because they look like a category and behave
like a fingerprint: 07:41 and 08:15 are never
going to appear twice.
What to do: a column is only worth grouping on if its values repeat. Count the distinct values — if that's close to the number of rows, it's an identifier, not a category.
The one nobody plans for. You build it, it's right, you add next month's rows — and the totals don't move, because they were typed in, or the formula range stops at the last row that existed on the day you wrote it.
What to do: reference whole columns
(=SUM(Data!C:C)) rather than a fixed range like
C2:C400, or put the data in a proper Excel Table so the range
grows with it. And never type a figure into a dashboard cell. If it can't
recalculate, it's a screenshot.
Every one of the five is fixable by hand in a spreadsheet you already own, and if you've got a clean export and an afternoon, do that — a dashboard you built yourself is one you can change.
It stops being reasonable when the file is long, or when it arrives again every month, or when it came out of a system that writes dates three ways and puts a total row at the bottom of every section. That's the point at which people quietly stop updating the dashboard, which is worse than not having built it.
Two of them, and they answer different questions.
A dashboard for your trade. Pick what you do — gyms, cafes, trades, hair and beauty — and you'll see a finished dashboard before you hand anything over, on numbers that are obviously not yours. Drop your own export in and the same boards redraw with your data: what you took, what sells, when you're busy, and the questions that belong to that trade rather than to spreadsheets in general.
A workbook you keep. Inspect reads a spreadsheet and hands back an Excel workbook with your cleaned data on one tab and a dashboard beside it — totals, a breakdown, and month by month, with the charts already drawn.
It works out which columns are what, so a balance isn't totalled, a rate isn't added up, a postcode isn't mistaken for a quantity and a total row isn't counted as data. Every figure in it is an Excel formula, not a number we typed — so you can click any cell and read the arithmetic, and it keeps working when you paste next month's rows in.
Both are free, nothing is stored, and the file never leaves Australia.