Step away from the charts with your hands up

What is this and who is it for?

If you’ve ever been given some data and asked you to turn it into a chart, this series is for you. You don’t need to work in data visualisation, or to have known that’s what it’s called. If part of your job is explaining something to other people, and you use data to help you do it, then sooner or later you’ll reach for a chart.

This series is about the bit that happens before that — the process that decides between a chart working or getting in the way.

It’s also for people who already do this every day and build charts, maps and dashboards without a second thought. The risk there isn’t that you don’t know how to make a visualisation, it’s that you’re good enough at it to skip straight to making one. Something I do constantly.

A caveat before we move on. I’m not claiming this process is the right and only way to do things, far from it. Rather this is an attempt to distill how I’ve ended up approaching things.

It’s a method I’ve assembled over years of getting things wrong and occasionally getting it right, and it’s a method that’s still changing. I’ll revisit these posts as my thinking moves on, so treat them as something in progress rather than set in stone.

Take what’s useful, ignore what isn’t, and if you’ve found something that works better then I’d genuinely like to hear it.

I should also clarify my background. I work in data visualisation at an energy supplier with a focus on renewables and low carbon technology, so most of my examples come from electricity and energy. None of what follows is specific to that, but it does explain why I focus on it throughout the series.

A chart without a story has no purpose

For most people their experience of data visualisation is through charts and maps, and they likely had no idea (nor cared) that it’s even called data visualisation.

I’ve lost count of the number of times I’ve heard, “We’d like to make a chart using this data” or something similar. It’s one of the surest ways to tell me that the purpose of a data visualisation, the story and the reaction to that story, probably hasn’t been thought about.

Now don’t get me wrong, I’m as guilty of this as anyone else. So what’s the problem?

The short version is that without a story and a desired reaction, a data visualisation has no meaning and purpose. It’s just a pretty chart or map, or any of the many other types of visualisation (more on that later…). And without a story it is nearly impossible to invoke an emotion and, importantly in some cases, a reaction.

At best this will be something nice to look at, and at worst it can end up misinforming and even causing damage to the reader’s understanding and reaction to a topic.

A well chosen and executed story can make the difference between someone learning (and doing) absolutely nothing, and someone truly understanding something and changing their behaviour.

Now of course not every chart or visualisation needs to be so profound, however it’s critical to consider when the results of that understanding (and subsequent actions) can make the difference between the public at large supporting and reaching a low-carbon future or not.

To put it another way, a simple story can be the difference between driving straight past someone, or bringing them along for the ride.

Bad Excel

So why do we reach for charts without thinking?

I think part of this is down to our exposure to data visualisation (in the loosest possible sense) being mostly via the media or through spreadsheet software like Excel.

What do you do when you want to make sense of a table in Excel? You click the “Charts” button and pick whatever looks nicest. Job done. Very few people stop to think about the story they want to tell with that table, or what the desired reaction is once someone sees that story. Why would you?

In the media we often see data represented as line or bar charts in the news, or perhaps the occasional map showing where something has happened in the world. There are absolutely some very valid examples of this done well, though more often than not the charts are there purely to make the data more visually interesting (tables are boring). There’s often no context, story or a desired reaction.

The result is that when given data our immediate reaction is often to turn it into a chart in an attempt to make sense of it. And that can work just fine, though it’s missing the one critical piece of the puzzle that truly makes sense of any dataset — a story.

And frankly how can you blame anyone? These are things you wouldn’t even consider unless you have particular experience or training on the nuances of data visualisation.

Simple stories, not novels

What do I mean by stories?

I mean less of the verbose stories that you’d find in books and instead more of the simplistic, impactful narrative that gets straight to the point.

Think about the kind of story you’d tell when bumping into a friend and you don’t have much time. Simple and concise. You wouldn’t open with the background and the caveats, or ask them to remember a pile of backstory before you get to the point. You’d cut to the chase and get to the point quickly, because you know that if you don’t then you won’t have time to finish the story before you need to be elsewhere.

The temptation is always to include everything you know and explain every little detail in the data, especially when you know a subject so well that it all feels relevant. But the more a story carries and the longer it gets, less of it lands and you’re more likely to lose your audience.

Keeping it focused on the things that actually matter is what holds someone’s attention long enough to get the reaction you’re after.

The five questions

None of this needs a tool, a dataset or any particular skill. It’s five questions, and you can answer all of them before you open anything.

1. What is the story?
2. Who are you telling the story to?
3. What is the reaction you want to the story?
4. What data do you need to tell the story?
5. What visualisation techniques (if any) best tell that story?

The first three are the thinking, and they’re the ones that decide whether any of this works. The last two are the doing.

The order matters more than you’d expect, particularly the fact that data isn’t the first thing to think about. That one may seem counterintuitive, and it’s worth its own post later in this series.

You don’t have to answer all five every time. Asking yourself just one of them and properly thinking about it will still leave you better off than randomly picking a chart type and moving on.

Where this goes next

None of this is about charts and visualisations yet, and that’s deliberate. The five questions above are all things you can answer with a notepad and no data in front of you.

In the next posts I’ll take the first three — the story, the audience, the reaction — and work through them properly. Those three are the ones that do the heavy lifting, and they’re the ones that will have the biggest impact on how you approach data visualisation in the future.

Later in the series we’ll look at what data you actually need to tell the story (this might be later in the process than you’d expect), and how to pick a visualisation technique that works with your story rather than against it.

For now, keep those five questions in mind next time someone asks you for a chart. And if you want to see results immediately, start by asking what you want the reader to do when they see that chart.

I think you’ll be surprised how much that helps…