MM Week 47: My First Custom Color Palette
And how color influenced my decision to use a black background
For American teens, Snapchat is the most popular social media app; but, that wasn’t the case in the spring of 2015. In fact, since the spring of 2015, Snapchat has grown from 11% of American teens reporting it as their favorite to 47% in the fall of 2017.
For my Week 47 Makeover Monday –Demonstrating Snapchat’s growth relative to the competition.
Click for interactive version
- Creating my first custom color palette
- Design choices
- Why a black background
Creating my first custom color palette: Eva makes it easy
Not much to say here other than – Bookmark this blog post by Eva Murray, Creating Custom Color Palettes in Tableau Desktop. I first read Eva’s post last week, so I came into week 47 thinking ‘custom color palette’.
Not only does Eva explain the steps exceptionally well, she also provides a spreadsheet formatted to build the required script; just add the color codes and name the new custom colors. All I had to do was research the codes for the various apps’ brand colors, fill in Eva’s spreadsheet, and update the Tableau preferences file with the output.
Design choices: My simple mind couldn’t see past lines
I’ve been experimenting with line charts of late and saw an opportunity to further explore some ideas. Specifically, I’ve been toying with dual axes and fixed ranges on continuous date axes. I implemented both in this viz.
I’m working on a separate blog post to dive into the steps I took here. It will come complete with credits and links to resources I found helpful. Look for a complementary post soon!
For now, here is the 30,000 ft view:
- Dual axis. The dual axis allows for different labels on each line. It’s really two charts, one directly on top of the other. In this case, I was after different labels on each end of my lines; the app name on the left end and the most recent poll results on the right.
- Fixed range. The default axis setting for a continuous date is an automatic range. Left unchanged, there will usually be insufficient room to place labels directly beside the ends of each line. The tip here is to set the axis to ‘fixed’ and select a wider date range. Doing so ‘slims’ down the chart relative to the length of the x-axis leaving room for labels.
Why a black background: Doesn’t the yellow pop?
Once I had my lines and labels in place, I added the custom colors and set the labels to match the mark color. Immediately, I came to two realizations – I couldn’t read the yellow Snapchat label on the white background and the use of four distinct colors was well outside my comfort zone. I suddenly wasn’t so sure about my design choices.
Typically, I would have used a pairing such as blue and grey to distinguish Snapchat from the field and kept the background white. In this case, however, I wanted to incorporate the branding. Plus, my goal coming into the week was to use a custom color palette. So, I had to figure out how to make my yellow label work.
As I tested a few light grey backgrounds, it occurred to me that a black background would incorporate Snapchat’s secondary brand color. Hummmm, I thought, should I give a black background a shot? I’ve read time and again that I should be cautious with changing my background from white; I had better have a good reason.
A case for the black background – It incorporates Snapchat’s secondary color while effectively highlighting their primary color. Consistent with the story, Snapchat’s brand takes over the viz.
The yellow pops so well. But not just in front of the black background; it also stands out well against the three darker, dimmer colors of the other lines. The contrast pulls the viewers’ attention directly to the story – Snapchat’s growth relative to the competition. It fit the story too well to ignore.
Sensing the finish line around the corner, I had to ask myself the question – Does the viz work?
I landed on yes. Do you agree? Please share your opinion on my thought process as well as the viz. Let’s connect on Twitter or via the comments section.
Thank you for reading. If you found this post informative or interesting, please Follow me on Twitter so we can keep the conversation going.
Relatable Data 