Skip to content
Utah Community Learning

Cleaning up a chart so the point is obvious

About 20 minutes

Cleaning up a chart so the point is obvious

Last time we picked the right chart type for the job. Bar, line, or pie, matched to the question you are truly asking.

This lesson is about the part most everybody skips. You make the chart, it technically shows your data, and somehow it is still confusing to look at. That is a cleanup problem, not a chart-type problem, and it is fixable in about five minutes once you know what to look for.

A chart is meant to make one point obvious at a glance. If a person has to squint at it and ask you what they are looking at, it is not doing its job yet, even when the underlying data is perfect.

Start by naming the one point

Before you touch any formatting, say out loud (or type in a cell somewhere) what the chart is supposed to prove. Something like "we spend far more in December than any other month." One sentence.

If you cannot get it down to one sentence, that is not a formatting problem, that is a too-many-ideas problem, and no amount of cleanup will fix it. Split it into two charts instead.

The usual mess, and how to clear it

Do away with the legend if you do not need it. If you have only one series (one line, one set of bars), a legend is just wasted space telling you "this is the only thing here." Click it and press delete. Excel and Sheets both let you label the data directly instead, which is cleaner.

Fix the title. "Chart 1" or "Sheet1!A1:B12" is not a title, that is a filing error. Double-click the title box and type the actual point. Not "Monthly Spending," but "December spending is far higher than every other month." Make the title earn its keep.

Turn off gridlines you do not need. A few are fine, not necessarily all of them, but a chart covered in faint horizontal lines everywhere gets busy in a hurry. Try turning off the vertical ones first and see if it reads easier.

Cut the decimal places. Nobody needs to see $342.17 on a bar label. Round it. Right-click the axis or the data labels, format as currency or number, and set the decimals to zero.

Reconsider your colors. If every bar is a different color for no reason, that is just noise. One color for most of the bars, and a second color ONLY on the bar you want to draw attention to, is a trick that works nearly every time. It leads the eye right where you want it.

Look at your axis scale. This one catches folks. If your spending runs from $200 to $600 but your axis starts at zero and climbs to $5,000, all your bars stand about the same height and the real difference vanishes. Adjust the axis minimum and maximum so the actual range of your data fills the chart.

A quick honesty check

I will own up to this: I once built a whole chart, a really fine one, showing our grocery spending by month, and felt very proud of myself. I showed Casey. He looked at it for about two seconds and said, "so we spend a lot in December." That was it. That was the whole insight.

And that is the point. That is the chart working. I wanted him to be impressed by the chart itself, and instead he understood the point at once and went on with his day. A good chart is not made to be admired. It is made to disappear and leave you with the fact.

A word about trusting what is in front of you

Here is where I bring up the two weeks. Early on, in my very first real spreadsheet job, I had a formula pointed at the wrong column. Just one column off. It sat there wrong for two solid weeks and nobody caught it, myself included, because the chart built from it still looked reasonable. Nothing looked broken.

When I finally found it I went cold for a moment, the way you do when you think you have made a bad mistake. Then I actually checked what it changed, and it turned out the wrong numbers had been wrong in a way that did not change the conclusion. I laughed out loud, alone at my desk.

The lesson is not "it always works out." The lesson is that a chart can look clean and confident and still be pointed at the wrong data underneath. Before you clean up the formatting, double-check that the chart is truly pulling from the range you think it is. Click on it, look at the highlighted source range, and make sure it matches what you meant. It takes ten seconds. Do it before you spend twenty minutes making a wrong chart pretty.

Try this at home

Pick a chart you have already made, from the budget or the tracking sheet. Say the one-sentence point out loud. Then work through it: legend, title, decimals, colors, axis scale. See how much clearer it gets once you are only fighting for one idea instead of five.

Before next time: bring the chart you cleaned up, even a rough version. Next lesson we are going to talk about when a chart is the wrong tool entirely and a plain number serves you better.

  • C