How to tell stories with data

 How to tell stories with data

Image by Tumisu from Pixabay

by David Ainsworth, freelance journalist

Data is becoming ever more present, and ever more important, in the way we work. Increasingly, we’re gathering more information on operations, funding and services.

But it’s not enough to gather and share data. We have to do it in a way that makes it easy to understand, which makes it interesting, and which engages potential audiences. We have to make the data tell a story.

What makes a good story?

Make the data tell a story.

Easily said. Not so easily done.

The first thing we have to do is to understand what a good story looks like.

A good story involves something new

For a story to be strong, something new has to be happening.

The crucial thing is change. A good story tells how things change. Ideally, it involves a single succinct statement about a clear, significant change.

You can see this any time you pick up a newspaper. Most articles will have an introduction that tries to summarise the story. It is usually just one sentence, usually less than 25 words, and usually contains only a single idea.

If your data contains more than one strong, significant change, it may be best to think about how you can tell more than one story. How easy this is depends on the medium and the audience. More on this later.

A good story is relevant to the reader

Most changes don’t matter to most people. If the weather in Brazil changes, and that changes the prospects for the coffee harvest, that’s a strong significant change, but it’s no issue to you and me unless it means we can’t have a cup of coffee in the morning.

If there’s a rise in the number of charity insolvencies, on the other hand, Brazilian coffee investors won’t care, but I’ll take a long look.

So the story has to mean something to the people reading it. That means that to create a good story, you have to know who you’re talking to. If you don’t know who your audience is, and you don’t know what’s important to them, you can’t tell a story.

A good story creates a reaction

Lastly, the change you show must be interesting. The reader or listener must care.

So it’s good to know the type of reaction you’re trying to generate. Perhaps we’re expecting an emotional reaction – a feeling of surprise, or awe. Perhaps we’re shocked by challenging opinions or preconceptions. Perhaps we’re expecting to get the “Aha!” moment when we create new knowledge.

With luck, reaction will lead to action. Our story will get someone to do something.

How do you tell the story?

For a story to be good, something has to have changed. It has to have changed in a way that’s relevant and accessible to the audience. And that change has to provoke feeling – it has to surprise people, or intrigue them, or otherwise make them care. If you’ve got all those things, you’ve got a story.

Then we need to think about how to tell it. Here are some tips.


Good stories can be summed up in a single sentence. They don’t need a blizzard of numbers.

So cut it down, make it simple. A good data story uses as few numbers as possible – the most compelling and relevant numbers.

Let’s take an example. It’s from a book about numbers – a book which contains thousands of statistics. It’s called Capital in the 21st Century, by the French economist Thomas Piketty.

Despite writing tens of thousands of words and studying thousands of numbers, Piketty is able to reduce his story to three symbols – not even numbers.

R > G

R stands for return on capital. G is economic growth. Piketty’s assertion is that every year, the stock market grows faster than the economy as a whole. So, wealth increases faster than income, which means that more and more money will inevitably accumulate in the hands of people who have wealth. A powerful story is told in just three symbols.


Another key thing about finding a story in the data is to dig beneath the surface. Don’t just present the big and obvious numbers. Hunt around. Manipulate the information. Push and pull at it. Test it. Try and work out what conclusions you can draw.

The key to a good story is to not just present but to analyse. Explain now just how it is, but why it is that way.

Here’s a dataset I was looking at the other day.

Graph of Google searches for the term 'mental health' from 2011-2021 showing the amount of searches steadily increasing

It’s searches for the term ‘mental health’ in the past ten years.

What happens when you compare that to data on charitable grant giving in the last ten years? If we map grants given to mental health causes, does total funding follow the same line as that map? The answer, broadly, is yes. That correlation makes both data sets more interesting.


It’s often better to tell a story in pictures. Again, let’s look at an example.

This is the Gapminder Foundation image of population, income and life expectancy in 1919.

Visualisation of life expectancy in 1919 with each country represented by a coloured circle whose size represents the country's population

And here it is in 2019:

Visualisation of life expectancy in 2019 with each country represented by a coloured circle whose size represents the country's population

The Gapminder Foundation was founded by Hans Rosling, a Swedish physician, inspirational speaker and part-time sword-swallower, to tell another compelling story: things are getting better, almost everywhere, for almost everyone, all the time. We’re living longer, we’re better educated, we’re happier, and we’ve got more cash.

These graphs are one of his main tools to tell that story.

Having said all that, remember that pictures are a powerful tool to tell a story. But so are numbers. And the most powerful tool is still (usually) words.

Focus on loss

Rosling tells us, correctly, that human history is the continual story of things almost always getting better for almost everyone.

This is obviously true. So why do we find this idea of progress so hard to believe?

It’s in our genes, basically. Studies by behavioural economists have repeatedly found that human beings tend to be powerfully loss averse. We view £1 lost as similar in value to £2 gained. And so stories about loss gain much more traction than stories about gain.

This isn’t to say that we can’t cut through with good news. We’ve all read stories about people winning the lottery. We’ve all read about Captain Tom. But these stories must be pound-for-pound more interesting to get the same traction. If your graph shows sales going down, you can bet it will get more attention than a graph showing sales going up.

Tell stories about people

People like stories about people. That’s the nature of things. So if you want to make your data more compelling, make it about people. If your story is about small charities, make it about one small charity chief executive. If your story is about services, make it about one service user.

Again, this goes back to behavioural economics and evolutionary psychology. We evolved to live in tribes of less than 100 people. Our brains are constructed for an existence where you will only ever encounter a few hundred other people. No amount of knowing that that’s not true will have any impact on how we feel.

And finally…

Let me sum up. There isn’t that much.

Tell a simple story. Explain why. Use pictures. Focus on stuff going wrong. Make it about people. And make sure it is new, relevant, and interesting to your audience.

If you do that, your data will tell a good story.

To join the Data4Good Festival mailing list and get updates about other interesting blogs and future events, sign up here.