I’ve been asked a number of times how I went about writing Practical Charts because it’s so, well, practical. This is a wonderful compliment to receive and so, as a cheap thank-you, here’s the story behind the book.

It started with someone else’s workshop…

In 2013, I attended a training workshop delivered by legendary dataviz educator Stephen Few. Prior to that, I’d been making charts as best I could, but that workshop was my first exposure to underlying concepts and best practices, and it blew me away. Steve and I stayed in touch after the workshop and, a few months later, I pitched a partnership to him under which he’d train me to teach his courses in exchange for a share of the revenue from workshops that I’d teach. Incredibly, he accepted my offer, and I ended up training thousands of professionals in workshops in over a dozen countries from 2014 to 2019.

While delivering those workshops, I noticed that the dataviz mistakes that chart creators made were remarkably consistent from one client organization to the next, and the specific chart-making challenges that arose were also remarkably similar. I also noticed that many of those mistakes and challenges weren’t covered in any of the dataviz courses or books that I was aware of, so I started noting them, as well as clever solutions that I came across (often, from uncredited geniuses who attended my workshops) or solutions that I thought of on my own.

In 2019, Steve retired from teaching and encouraged me to develop my own courses (though I continued to offer his courses, as well). The first step was to go back and review my notes and… yikes. Hundreds of completely random thoughts and ideas, with no discernable sequence or structure. Not even a central theme. Just a bunch of “If I ever develop my own course, make sure to mention…”

At the time, the only thing of which I felt confident was that those notes contained at least 90% of the common mistakes and 90% of the common challenges that arise in real-world chart-making. How could I know that? Because, after teaching workshops for three or four years, almost all of the dataviz mistakes and challenges that I came across among new clients were already mentioned somewhere in my notes, and it was quite rare for me to get a new question that I hadn’t been asked already in a past workshop.

If I could develop a course that taught participants how to avoid all of those mistakes and handle all of those challenges, that course would, almost by definition, turn participants into expert chart makers, able to make clear, compelling, and problem-free charts, every time.

Developing the Practical Charts course

The hardest part of developing the Practical Charts course was organizing all of those hundreds of random ideas into a structure and sequence that made sense and that would be easy for participants to follow. Oh, and cramming all those ideas into a course that could be taught in two days (or four half-days). I was worried that the course would feel like a quasi-random collection of tips and guidelines that would be hard for participants to absorb.

The organizing and sequencing process took many weeks and fried many brain cells. Figuring out which concepts or guidelines I needed to talk about before I could talk about which other concepts or guidelines was difficult. One of the techniques that I used to group and sequence ideas in my own mind was to sketch out decision trees for making various chart design choices, which turned out to be very helpful. I’m embarrassed to admit that it didn’t occur to me to actually include those decision trees in the course itself until quite late in the course development process, since many people now consider those decision trees to be the most valuable parts of the course.

Over time, a structure emerged, and I was able to organize all the ideas into a sequence that made sense. I was amazed that all my hundreds of random notes somehow “fit together,” like a puzzle. There were many small refinements and iterations after I launched the course in 2020, but I was lucky enough to get lots of positive feedback from participants from day one. As I’d hoped, the most common adjective that people used to describe the course was “practical,” that is, that it provided specific guidance on how to tackle almost all of the real-world challenges that they faced when creating charts.

I should mention that early drafts of the course contained a certain amount of dataviz theory (pre-attentive attributes of visual perception, visual hierarchy, color theory, subitizing, etc.). Like many other experienced chart creators, I assumed that people would need to know at least some of those theoretical concepts in order to create effective charts. I did want to minimize the amount of theory in the course, though, so I started cutting everything that wasn’t necessary in order to fully understand a particular practical guideline in the book. When I was done cutting, there was no theory left. I realized, then, that, if you’re just creating “everyday” charts for reports and presentations (and not highly customized, scientific, or artistic charts), I don’t think you need to know any dataviz theory.

BTW, I still regularly deliver the Practical Charts course and continue to update and improve it. If you’re interested in attending an upcoming online public workshop or scheduling a private online or in-person workshop for your team, contact us.

I provided (and still do provide) a PDF of the course slides to workshop participants, but I was often asked for a book version that would be easier to reference after the workshop, since the course slide PDF contains over 500 slides (!).

Writing the book

Once the course had undergone multiple iterations and was in a pretty fine-tuned state, I thought that it would be relatively easy to write the book, since the book would basically just be a transcript of me delivering the course.

Hah.

I’d been cobbling together passages for a book for quite some time, but progress was slow, and I figured that the only way that I’d ever finish it was to block off weeks of uninterrupted writing time in my calendar. During an eight-month period in 2022, I only delivered a handful of workshops so that I could focus mostly on writing, and that turned out to be the right call since I finished the manuscript in early 2023, with monumental amounts of help from my colleague, Bryan Pierce. Even though I thought that the course was very fine-tuned already, I ended up tweaking many of the examples and guidelines from the course while writing the book, which I then worked back into the course after the book was written. As they say, if you want to understand something fully, write about it…

The first draft of the manuscript ended up being almost 400 pages long, which was far more than the 250 pages that I was aiming for (most people don’t want to read a brick). I noticed, though, that about 20 of the 50 chart types that I covered in the book were ones that roughly half of my workshop participants told me they’d never use because their audiences “weren’t data-savvy enough” to grasp them, such as scatter plots, histograms, and cycle plots. To address all of these issues, I decided to split the book in two:

Practical Charts, which covers 30 common chart types for the many chart creators who only create charts for “non-data-savvy” audiences and clocks in at a manageable 302 pages
More Practical Charts, which covers 20 additional chart types that should only be used with data-savvy audiences and is a very manageable 98 pages

Since the book was published…

When Practical Charts was released in November 2023, it was the top-selling new release in two (admittedly niche-y) Amazon categories and, thankfully, it’s been selling well since. It’s also been picked up as a textbook at Yale and is currently being evaluated by several other universities. Even more gratifying is the fact that I see social media posts and get messages virtually every day from readers who’ve found the book to be valuable, and from workshop participants who’ve found it to be an essential post-workshop reference. One review on LinkedIn read, “I have many great dataviz books on my shelf, but Practical Charts is the one that stays on my desk,” and there have been many similar comments which have been particularly gratifying, since that that was one of my main goals when writing the book.

How is the book holding up against my other goal of making it complete enough so that, if I came across a chart that I considered to be poorly designed for any reason, I could point to a part of the book that described the problem and what to do about it? Thankfully, pretty well. Over 90% of the chart design problems that I see “in the wild” are covered somewhere in the two books. As the field of data visualization continues to evolve, that percentage will undoubtedly start to drop, but that’s what second editions are for 😊.

So what?

What can you get from this story, especially if you’re thinking of developing your own course or writing your own book? To be honest, I’m not sure, but here are a couple of random thoughts:

As you can see, Practical Charts was written “from the bottom up,” not “top-down.” I didn’t start out with a central idea or theme that I developed into increasingly detailed sections, as is so often taught in writing courses. It started out as a collection of details, which were then cobbled together into a structure and sequence that only emerged relatively late in the process. A central idea (i.e., that charts are “graphics for doing a job”) only emerged very late in the process. Basically, “bottom up” can be a valid way to develop a course or book, even though that’s not how it’s normally taught.
The book also isn’t “the important things that I think people should know about data visualization.” Instead, it’s “what I saw that people needed to know in order to create expert-level charts, regardless of how important or unimportant that I felt that it was.” Each workshop that I taught provided invaluable “market research” that allowed me to identify the common gaps chart creators’ knowledge, what challenges they faced most often in real-world practice, and which solutions they found to be most useful. I tried to focus only on that and forget about what I, personally, thought was important or unimportant for them to know.
I waited several years before starting to work on the book in earnest and, in hindsight, I think that that was the right decision. If I’d started writing it as soon as I’d decided that I wanted to write a book, the ideas wouldn’t have been fully baked and the book would have been considerably less useful. I mention this because it goes against the common advice that the best time to start any big project is today. Sometimes it is, but sometimes you’re better off waiting until you feel like things are more fully formed in your mind.

I hope you enjoyed this look at the messy process of writing Practical Charts. Did anything jump out at you as surprising or useful? Have you had similar experiences working on a large project? Let me know on LinkedIn or Twitter, or in the comments below.

BTW…

If you enjoyed this article, consider subscribing to my email list (one to three emails per month to level up your data visualization and dashboard design chops).