How to Turn Ugly Charts That Make No Sense Into Simple, Clean Visual Stories
“Above all else show the data.” ― Edward R. Tufte, The Visual Display of Quantitative Information
This pioneer in the field of Data Visualization loves data. He strongly believes, “To clarify, *add* data.”
Some unfortunately believe in the exact opposite- To confuse the audience even more, add data (read ugly charts and graphs). Sad part is some charts are not just plain ugly, they don’t make any sense too. What is this column chart trying to say, can anyone tell?
Feeling dizzy? No? Let’s increase the dose (cue evil laughter):
Now you know how I feel! (The creator of the above graph must be saying right now- Haters will hate)
Data Visualization exists to serve one major function- simplify the data and show it visually. Ideally speaking, we are supposed to find the story hidden in that pile of data and use the most apt graphic to tell that story. Back to the real world- presentation slides with data tell stories that will make anybody’s blood run cold.
All right, I went a bit extreme. Not all real world graphs and charts are that bad! Some are intentionally misleading; that will be the discussion of another post. Some are using the wrong data visualization tool for their data; sorry that will be discussion of another post too. Some are just ugly and unintentionally making their data more complicated. That is the discussion of this post. And since a column chart or bar diagram is the most commonly needed and used graphic to show comparison across categories, we have picked the same to demonstrate how to clean them up step by step and visually narrate your message in the shortest time.
Check out the before 3-D column chart and the after version of the same:
Let’s see what we did and why:
Step by Step Polishing of Graphs
We have to admit that the example in the before is not an epic disaster. At least we can discern that it is a 3-D Column Chart depicting average maximum temperature over the months in 5 cities. But then let’s try to understand the data- can you tell what was the average temperature in Christchurch in the month of say August? Or that of Geneva in September? I had a neck exercise, pretty much like the one below, while trying to decipher the values (the guy in center was smart enough not to give a damn!):
Locating the cities and mapping it to their respective bars is a left-right, right-left, top-bottom and bottom-up… a sprain-inducing exercise. When the city is located, guessing the maximum temperature is an additional headache- the 3D design of the chart greatly hinders in easy comprehension. Overall, a disaster and failure of data visualization.
So, let’s jump in and save the day:
Step 1- Change Chart Type
Clearly, we have to drop this 3D idea no matter how lovely the effect looks in theatres. We can convert the 3D chart to a 2D chart to make it less horrible but that would mean lots of columns that would still take good 15 minutes to get the message. Here’s how the 2D version looks:
Personally I like horizontal bars over the columns. It looks more in tune with times. That’s probably because audience is saturated looking at those vertical bars. So, my first step will be to change the chart type to a 2D Clustered Bar diagram:
Right click on the chart area and click Change Chart Type (see the screenshot below):
The Insert Chart window will open giving you options to choose the chart type. Go to the Bar category and select the first option- Clustered Bar. Click OK to make the changes (see the screenshot below).
The new chart will open on your screen.
The result is not so good but have patience- beauty takes time!
Note: If your chart was not created in PowerPoint but in Excel and you had saved that as an image and thereafter inserted in PowerPoint, we advise that you create chart in PowerPoint (most ugly charts are created in Excel). Go to the Insert tab in the PowerPoint ribbon, locate the Charts button and choose the Clustered Bar option. You can copy-paste your Excel data into the default sheet that will open on your PowerPoint slide.
Step 2- Edit the Data (Keep One Series Only)
The best way to analyse the mind-boggling data is to split it into 5 graphs- one for each city. The presentation advice- Keep one message per slide- also applies to charts and graphs. For cleaning up this bar chart, right click on the chart > Edit Data > Edit Data in Excel (see the screenshot below).
The Excel sheet will open on your screen. Keep the data of one city, select the remaining 4 cities and press Delete.
Step 3- Remove the Gridlines, Legend and Chart Title
The bar chart below looks much cleaner now. Let’s clean it up further. We do not need the Legend any more since the data is of just one city. We do not need the Chart Title too. Even the Gridlines aren’t adding any value. Select the chart, locate the + button to the right and deselect the Chart Elements- Chart Title, Gridlines and Legend (see the screenshot below).
Step 4- Use a Professional Color
Since charts and graphs talk about serious data, let us use professional colors that have been trusted by all business organizations- greys and blues. Select the horizontal bars, right click, open the Fill drop- down menu and choose a professional color you like. Alternatively, click the Format Data Series… option. The Format Data Series window will open to the right, go to the Fill tab and change the color.
Grey can never go wrong; even if it’s 50 shades of grey (in PowerPoint!)
Step 5- Sort the Data
The graph may be looking clean now but what is the data trying to say? Right now, it shows only the fluctuations in temperature. Which month recorded the average maximum temperature? Which the lowest? It takes at least 2 minutes to figure that out. And coincidentally, 2 months (January and November) recorded the maximum temperatures. As many as nine bars separate the two. Let’s make this analysis insanely easy for the audience. Let’s arrange it in ascending order- smallest to largest!
Accomplish this with the click of a button.
- Right click on the chart > Edit Data > Edit Data in Excel.
- The Excel sheet opens on your screen. Locate the Sort & Filter menu on the right.
- Click Custom Sort
The Sort window will open on your screen. Select the Sort by column which is only one in our case. In the Order menu, select Smallest to Largest. Click OK.
Close the Excel sheet and see your bar graph- it is neatly arranged in that order. Audience is already blessing you for that!
Step 6- Highlight the Most Important Data
Let us make reading bar graphs even more easy for the audience. We have made the food for them, let us serve it too on their table. We have arranged the data for them, let us highlight the main points and guide their eyes straight to that!
Since the story in our data is noticing which months recorded the maximum and minimum average temperature, let us give the two a different color. Right click on the chart area, go to Fill and give it a color of your choice. We gave the maximum temperature bar a dark blue color and the lowest bar a mustard color. Here’s how the data for the Wellington now looks:
Now repeat the steps for all the remaining cities- Insert Clustered Bar, copy your data and paste it into the Excel sheet, sort the data, and highlight the most important values.
Minimize the chart size of all the cities so that we can fit them into one slide. Because of neatly arranged charts, it is very easy to compare the data across all the cities. Here’s how your final graph now looks:
Could we have done anything else? Yes, a line chart showing the fluctuations. We tried that but horizontal bar chart looked the best.
BONUS TIP 1: Summarize the Findings of Data in Slide Heading
Following the same steps we had used above, we transformed another hard-to-understand 3D chart into a clean and crisp bar diagram. This time, we also added a slide title and sub-title which summarize the main findings of the data. Here’s the before and after:
BONUS TIP 2: Add Icons If Possible
We found another poor specimen of a chart gone horribly wrong. This time since it was a survey result of social media usage in percentages that totalled to 100, a pie chart looked best. So, we went to Chart styles > Pie > Doughnut chart. We split the data for easy comprehension, added the maps of the respective continents as icons. Again, the title and sub-title was used to highlight the finding of the data and quickly tell the audience the key takeaway from the data visualization. Here’s the transformation from ugly chart to a stunning data visualization:
How did you like the presentation tutorial? Was it beneficial? Do you have any good and horrible data stories to share? We’ll love to hear your feedback in the comments below.
Do you need to use tables in your presentation? If yes, learn how to turn shabby PowerPoint tables into neat and professional ones by clicking the link below: