Simplicity v Complexity
This week I created two vizzes, and I am not quite sure I quite nailed either one. With the benefit of hindsight, one was too simple, and one was too complex.
The first was my redesign for Makeover Monday to show Andy Kriebel’s Amex spend in 2016. I went for a simple, stripped-back approach, showing cumulative spend over the year as an area graph.
So why do I think it is too simple? Because there are a number of questions which could be answered from the data which aren’t addressed, e.g.
- how many transactions did Andy make?
- what type of spend was it?
- what days of the week did Andy spend the most?
If I get the time I will look to expand the viz to incorporate answers to these questions: to keep the design simple, but not the message.
The other viz I created this week was this Sankey diagram, using guidance from Chris Love.
So why do I think it is too complex?
Firstly the graph type is not a standard one in Tableau, and the method Chris has used is very complex – I would never have worked out how to do it on my own.
Secondly for this type of Sankey, where there are distinct categories on the left and the right, there are better, simpler ways of showing the breakdown in volume, as Chris himself explains in his excellent talk on embracing simplicity.
The other lesson for me is I lost sight of my original intention, which was to redesign this diagram:
I do still think a Sankey could still work for this, with Q3 on the left and Q4 on the right, but because the instructions I found didn’t cater for change in volume over time, I just built a Sankey using dummy data.
So I proved that I could follow and adapt instructions, but in terms of the ultimate objective of data visualisation – to communicate – it was a hollow exercise.