💡 Help people get your complex chart

👋 Hi!

Yesterday this newsletter hit 15k subscribers. I never imagined sharing weekly actionable dataviz tips would reach so many people — thank you.

All past issues are here. If you found something useful, please forward this email to a friend!

Today I suggest we talk about using unconventional or complex charts, and how to get people to actually adopt them.

​

Unconventional graphs

​Data To Viz lists about 50 chart types. Yet the vast majority of charts you see are one of ~six common types.
​
That is fine! Bars and scatterplots work beautifully in many cases. But sometimes a less common chart will communicate your idea much faster.
​
For instance, here are a couple of ideas to visualise flows:

It would be very sad ignoring those options. Yet, people not necessarily know them, so how can we do?

​

​

A concrete example

Many years ago I saw a project by Nadieh Bremer. I loved it, and it motivated me so much to fall into dataviz.
​
It uses a chord diagram to show how phone customers switched brands. It is visually striking and, at first glance, complex. For this exact dataset, the chord diagram is the best way to show many-to-many flows.
​
​

The problem? Complex charts can scare people off. If the audience does not understand the visual, they stop reading the data. 😳

There is a solution though: introducing the chart step by step.

Start with a small, easy-to-understand piece. Add one connection or one layer at a time. Make each step useful on its own and invite the reader to continue.
​
Nadieh’s interactive chord walks you from outer circle to one simple flow to the full diagram. Each step teaches and keeps curiosity alive.
​
​

​
​How to apply it today

This stunning example uses JavaScript and d3.js to allow interactivity. It's the very best but is also very hard to learn.
​
But you can use this effect with any other tool!
​
In talks I love building complexity slide by slide. For instance, I first start with a single line that provides the context. Then add all groups. Then one annotation at a time. Then a companion chart for context.
​
It's easy to do. You just have to make several screenshots at each step of your chart creation. Or add some white squares to hide some specific parts.
​
Same pattern, different stack.

​
​The bottom line

When a chart is complex, reveal it gradually. You keep the audience, create suspense, and you can use more powerful but lesser known charts.

If you end up doing this at work, please let me know: it will make my day!

See you next week,

Yan

PS: My prtfol.com tool now has almost 463 users! A few ppl made very slick personal pages, like Pablo or Olivia. I'm working on a new feature that builds it from a LinkedIn profile. Hit reply to give it a go!
​
​PPS: I'm looking for sponsors for this newsletter
​