The timing of week 15 was pretty special for me personally. 10 years ago I downloaded Tableau and began sharing what I was learning with anyone that would listen. It’s been amazing for me to watch the Community grow, see people develop, and that’s where projects like Makeover Monday can really help you. Use it as a way to practice because without practice, you simply won’t get better.
Given the timing of this week, I decided to let everyone makeover one of my very first Tableau vizzes. I thought this was important because it shows not only how far Tableau to product has come, but it also shows everyone that is just now starting to use Tableau how much they can develop if they put in the effort. By showing where I used to be, I hope it inspires people to improve, progress and learn.
The data set was pretty simple: monthly price data of oil, gold, and the consumer price index. Originally I created a scatter plot which seems only natural when you want to compare two metrics. This worked really well when I first created the chart, but given how much the world has changed in the last eight years, a scatter plot didn’t tell the same story this time. More about that in a minute. Having looked at this data myself, I was very curious to see how people would approach it.
As we’re doing each week now, I’ll review some of my observations before highlighting my favorite vizzes of the week.
USING METRICS THAT ARE EASIER TO UNDERSTAND
Gold and oil prices (or any two measures for that matter), which have very different price scales, are difficult to compare. If you place them on the same axis, oil prices look totally flat. If you place them on separate scales, then you run the risk of confusing and/or misleading your audience. Many people overcame this challenge this week by create a gold-to-oil price ratio, giving the reader much more context as a result.
USING QUESTIONS IN TITLES
I must admit that the idea of putting a question in the title didn’t really “stick” with me until I read Stephanie Evergreen’s latest book Effective Data Visualization. By putting a question in the title, you are telling your reader EXACTLY what they can expect. Plus this can help your frame your story. Using a question helps you determine what to include and, possibly more importantly, what to exclude. If you don’t use this technique, try it next week. See if it helps you.
SEEKING FEEDBACK AND ITERATING
First, it’s important that everyone understands that a data visualization is never “done”. You merely get to a point where it’s “good enough” and you publish. The last several weeks, Eva and I have noticed more and more people asking for feedback. And better yet, they are acting on the feedback and evolving their work. As an example, check out how Louise Heath iterated on her heatmap and created a much simpler visualisation:
SIMPLIFY WHEREVER POSSIBLE
“Perfection is achieved not when there is nothing more to add, but when there is nothing left to take away.” – Antoine de Saint-Exupery
Like Louise demonstrated above, continually removed elements from your visualisation that don’t add to the story. This week, I saw too many examples where it appears people are trying to make their charts look fancy or different. That won’t win you any favor with me and if you’re creating busy-looking content like that at work, I bet you’re not helping your audience at work either.
If I can’t understand your viz quickly, as a static image, then I’m not very likely to keep it in mind for the weekly recap. I prefer simplicity, cleanliness, minimal noise. Next week when you’re working on Makeover Monday, consider what you can take out of your visualisation yet still retain the message. Are you using color for a specific reason? If not, then get rid of it. Use color sparingly. The same goes for legends, shapes, sizes, the whole gamut. Give that a try next week.
USE COLOR FOR IMPACT
No one I know teaches about the effective use of color better than Cole Nussbaumer. In her classes and in her book, she puts major emphasis on using color only where it’s needed, using color to highlight what’s important, and using color for a very specific purpose. In addition, I would recommend that you use colors that match the topic of the visualisation. This week that should have been a no-brainer. Gold for gold, black for oil, yet many people chose to use other colors.
My tip for you: be intentional about using color. Is color necessary? Does color aid understanding? Does the color match the theme?
SCATTERPLOTS AND CONNECTED SCATTERPLOTS
It’s only natural for people to gravitate towards scatterplots when comparing two dimensions. Scatterplots allow you to quickly see if there is a relationship between the data. Throw a trend line on and you can see how strong the correlation is. However, I don’t think scatterplots work for this particular data set because the R-squared isn’t significant enough.
Mina Ozgen demonstrates this well in this series of scatterplots. As she points out in her notes, you can basically follow the pattern with your eyes well enough.
Some people tried connect scatterplots. A connected scatterplot is used to connect the dots through time to see if any patterns emerge. The best example I saw this week was by Matt Francis.
While Matt’s viz looks really good, it’s very difficult to follow the line and no overall pattern really emerges. Matt does a nice job of including annotations to provide the needed context.
My lesson for you: try scatterplots and connected scatterplots when you have two measures to compare over time. You might see a nice pattern emerge, but then again you might not but at least you know because you tried.
WEEK 15 FAVORITES
This week was so hard to not pick tons and tons of favorites. I saved about 20 and reduced my favorites down to these eight. Congrats to everyone on another great week of vizzing and learning!
What I like:
- Mirrored design
- Nice, intentional use of color
- Annotating the low and high prices
- Including the year labels between the charts
- Lines between the years help guide the eye
- Great hover action to update the comparison viz at the bottom
- Crediting icon sources
- Clean, easy to read design
What I like:
- Using a 6-month moving average for the line and keeping each month as a dot for context
- Putting a question in the title
- Tooltips include the answer to the question
- Nice summary text
- Background color makes the dots and line stand out
- Including very light grid lines
- Clean design in which Mark has removed so much that doesn’t add to the viz
- Every pixel has a purpose
What I like:
- Nice overall layout with one viz across the top and two supporting vizzes below
- Good use of area charts
- Including annotations for context
- Lots of text for additional information and explanation
- Shading the background of the outliers in the gold/oil ratio chart
- Clean design
- Appropriate color choices
What I like:
- Very similar to Michael Mixon’s viz, but turned on it’s side so that time goes from top to bottom
- Incredible design!
- Perfect use of color
- So engaging!
- Including a “more info” button
- Fantastic explanations throughout
- Great use of annotations for adding context
- Including to overall range of prices
- Wow, just wow!
What I like:
- Pablo has been crushing Makeover Monday and he keeps getting better and better
- Great overall design that’s very engaging
- Love the “five key moments” callouts and the line that links them together
- Like Sarah’s one chart at the top supported by two at the bottom
- Really nice color choices; they work well together
- Using a question in the title of the first chart
- Including icons that add to the story
- Good chart choices