If you spend too much time on the internet learning about data visualization, you might get fooled into thinking it’s hard. All that stuff about “ggplot2” and Python or R--you don’t need it. Honestly, your trusty Excel is enough.
The secret to making made-in-Excel graphs beautiful and professional? Don't let them look like you made them in Excel. Here are two ways to get that look, along with examples:
Don’t default to the defaults
The people who created Excel were many things, but they were not designers at heart. Another reason to move away from the defaults is that you--and only you--know what your data needs. This all means that the default graph design is rarely the best one.
Play around with different chart layouts. Excel has a collection of predefined layouts that you can test and then change to fit what you want. Bonus points if you decide to play around with the font or change the colors to match your company colors.
Here’s an example. Let’s say you want to show unremitted foreign earnings over the last few years using data from idaciti, to show if a company took advantage of the “tax holiday” the government gave large corporations last year.
This is what you might get when you select the data and ask Excel for a bar chart in all its default glory:
That’s all well and fine, but let’s make a few small changes to improve it, like:
- Applying a preset design option from Excel
- Changing the color of the bars to something brighter
- Making the title, x-axis labels, and y-axis labels larger
- Simplifying the y-axis
And now we have a much more readable, visually appealing graph:
Declutter your graph
Less is more when it comes to visuals. In the spirit of minimalism, here are a few things you might want to opt out of if your data will allow it:
- Gridlines (can be distracting and make the graph feels crowded)
- Legend (especially if you only have one dataset graphed)
- Bar outlines (if you’re using a bar graph)
- Decimals (if they’re unnecessary)
- Error bars (does anyone other than a scientist really need these?)
- Labels (especially for data points that you don’t need to highlight)
You can also deemphasize the elements that you don’t want someone focusing on. For example, if you’re showing a graph of revenue across five products but you only want to talk about product #4, you could leave #4 in color and put all the others in shades of gray. Or, if you don’t like gray, by changing the opacity (making the other four product divisions lighter).
Using the same example as above, let’s take what we ended with and declutter a bit by:
- Deleting the legend since we’re only showing one datapoint
- Getting rid of gridlines
- Increase the text size of the data labels above the bars
Here’s the finished graph that’s even cleaner than what we had before:
Highlight the really important parts
Now we have a good-looking graph that's easy to read. But if you're presenting the graph or showing it to others, you also want to make sure that they see exactly what you're trying to show them. One way to do this is by adding a rich data label.
Here is our graph with a short callout that highlights what we're really interested in and even offers an explanation:
Graphs aren’t hard. And a good graph doesn’t need to be complicated or built in a programming language.
If you go a few extra steps past the defaults, you can make graphs with your favorite spreadsheet app. And they'll still be easy to understand, even with a minimal amount of effort from you.
Data in this post is found in this idaciti card and the highlight in this card. Download the data by “flipping” the card by clicking on the 🔁 icon in the top right hand corner. In the graph view, you can see data for other top 20 companies by clicking on the filter. In the data download view, you can also see more details by clicking on the data points.