Data storytelling: structuring a compelling narrative

C1
60 min
Premium
Video thumbnail
Scan to watchScan to watch
1

Think about these questions before watching. Share your ideas with a partner.

  1. In what ways can a narrative or story structure transform raw data, like sales figures or survey results, into something more persuasive or memorable?
  2. Recall a situation where you had to convince someone of a particular viewpoint using facts or evidence. What challenges did you face in making your argument not just logical, but also engaging?
  3. Beyond entertainment, where do you see the classic story elements—like a clear beginning, a central challenge, and a final outcome—being used effectively in a professional or academic context?
2

Watch the video carefully. Pay attention to the main ideas and key details.

Video script135 segments · click a timestamp to jump

SCOTT BERINATO: With data.

Everybody's doing it, right?

You're not doing it?

You should be doing it.

You've been told you should be doing it.

Storytelling with data is the big thing.

[MUSIC PLAYING]

Well, it turns out you should be doing storytelling with data,

but it's a lot easier than you probably think it is.

Once you know the three elements of a story, setup, conflict,

resolution, you can start telling stories with your data.

Setup conflict resolution.

Every story ever told from the beginning of time

follows this structure.

It's really that simple.

Setup.

Charlie Brown runs toward the ball.

Conflict.

Lucy takes the ball away at the last second.

Resolution?

Ah!

And a setup is just some reality.

And by reality, we just mean a situation.

It could be fictional, but it's a reality

you've created for the story.

And then you have conflict, which has changed that reality.

Without change, there is no story.

You just have a bored audience.

The resolution is just the new reality

that the change creates.

So with that in place, we can now take a chart,

and we're going to break it down, pull it apart, find

the stories in it, find the setups,

the conflicts, the resolutions, and we're

going to rebuild it as a storytelling device.

[MUSIC PLAYING]

OK.

Now we're back with a chart.

This is Global Real Home Price Index.

The index is 100 and you see a bunch of lines for countries.

There's a gray line in there.

That's the aggregate as well.

This chart is really showing home prices

in a lot of different places in the world,

and that's where I'm starting.

So, I want to find the stories in here.

And the first thing I start to notice, really,

are a couple of things.

I obviously noticed that big hump on the green line there.

That's Japan.

And I noticed that point where everything comes together.

To me, that immediately says those are probably

two places where there's conflict,

but I realized something as I'm looking at this,

and I'm glad I did, because otherwise,

I would have had the wrong story.

And that is that these home prices are indexed to 2005.

So, the fact they come together there at 2005

doesn't mean the prices came together.

That just means those are the dollars

that the people who made the chart

used to show the change in house prices over time.

So, there's really no conflict there.

And in fact, I think that's my setup.

That's where we start, because that's where they started.

I can still tell those two stories on the left

and on the right, but I start there

instead of at the beginning of the chart.

I can look backward and say, home prices rose steadily

in most places in the world for 30 years

except for in Japan, which experienced

a three decade long bubble, and that is a perfectly good story.

The setup is home prices rose steadily in most places,

except in Japan, which is the conflict.

And the resolution?

It experienced this 30 year bubble.

And then working from 2005 to the right,

I have another story, which is that there was a smaller house

bubble, housing price bubble in most places except Japan.

So now, the story has flipped, but something different

happened this time, and that is that the market's bifurcated.

And you ended up with three markets, Australia, Canada,

and New Zealand rising again, looking

much like a bubble, and the rest falling

and then rising back to about 2005 levels.

So, I've really started to see a couple of stories emerge here.

One, both starting in 2005, because that's

where our prices start, and one sort

of working backward in time, and one working forward in time.

[MUSIC PLAYING]

OK, so I'm going to spare you most of my sketching

because it's so messy and chaotic,

it probably would just give you a headache.

But I did a bunch of sketching, and I

arrived at these final charts.

And I've split the chart up into a couple of states.

It really helps your audience focus.

So, here's my setup.

You can see it.

Prices rose steadily in most places in the world.

I try to use my titles to actually reflect the story

and hear.

The steady increase reflects that setup.

Instead of just using a generic title about global real home

prices, which is boring and not helpful,

the title can really do some work for me here.

And then, I want to add the conflict and resolution

state, too.

There it is.

You can see we've added Japan, Except in Japan in the title,

and that conflict and resolution becomes clear

that Japan was this bubble that lasted 30 years that was

different than everywhere else.

And you can see as I present this, it almost looks seamless,

as if I'm just showing you one chart that changes state.

So then I'm just going to repeat this process for going forward.

And you see here, I've included my setup, conflict,

and resolution altogether in one state.

You see that little bubble and then

you see the bifurcation of the market.

But the most important thing is, I've really

highlighted the elements of the story

in both cases and nothing else.

I've left out any information that might distract

from telling that story.

I've not focused on anything that

doesn't matter to telling that simple story of setup,

conflict, and resolution.

Narrative is the most powerful, most human tool

we have to communicate.

If you can apply storytelling to your data,

it creates an emotional connection with the audience.

They're not only going to believe what you show them,

they're going to feel it.

[MUSIC PLAYING]

3

Answer these questions in your own words. Support your answers with evidence from the video.

01According to the speaker, what are the three fundamental components of any story, and what function does each one serve?
Sample answerThe speaker states that every story is built on three components: setup, conflict, and resolution. The 'setup' establishes the initial reality or situation. The 'conflict' is a change that disrupts that reality, which is essential for creating a story. Finally, the 'resolution' is the new reality that emerges as a result of that change.
02What did the speaker initially misinterpret about the point where all the lines on the chart converged, and how did correcting this misunderstanding alter his narrative approach?
Sample answerHe initially thought the point where the lines met in 2005 represented a conflict, a moment where all global house prices actually came together. However, he realised this was just the index year, meaning all price changes were measured relative to that point. This changed his entire approach; instead of being a conflict, the 2005 point became the 'setup' or the starting point from which he could tell two different stories, one looking backward and one looking forward.
03How does the speaker use the 2005 index point as a pivot to construct two distinct narratives from the same dataset?
Sample answerHe uses the 2005 point as a central anchor. For the first story, he looks backward in time from 2005, telling the narrative of steady price rises in most countries, contrasted with the major, long-lasting housing bubble in Japan. For the second story, he moves forward from 2005, describing the smaller global housing bubble and the subsequent 'bifurcation' where markets split into different recovery patterns.
04Beyond identifying the story elements, what specific techniques does the speaker apply to his final charts to make the narrative more compelling and easier for an audience to follow?
Sample answerHe modifies the charts in several ways to enhance the story. He splits the information into a sequence of 'states' to guide the audience's focus step-by-step. He also replaces generic chart titles with descriptive ones that explicitly state the story's point, like 'Except in Japan'. Most importantly, he intentionally removes any distracting information, highlighting only the data points essential to the setup, conflict, and resolution he wants to communicate.
4

Vocabulary

Vocabulary
These expressions will help you communicate more naturally about this topic.
Examples
To tease out a narrative — to carefully extract or discover a story or meaning that is not immediately obvious from a complex set of information.
Usage note: this is a sophisticated alternative to 'find a story'. It's often used when the information is complex or tangled, like in a dense dataset. Common collocations: tease out a narrative/story/meaning/implications.
To paint a picture — to describe a situation in a way that helps someone to imagine it clearly and vividly.
Usage note: this idiom is widely used in both formal and informal contexts. When discussing data, you might say, 'The sales figures from Q3 paint a rather bleak picture of consumer confidence.'
To take something at face value — to accept something as it appears to be, without questioning it or looking for a deeper meaning.
Usage note: this phrase is perfect for discussing the critical analysis of data. For example, 'You can't just take the headline numbers at face value; you need to investigate the underlying trends.'
To cherry-pick data — to selectively choose the most favorable data points to support an argument, while ignoring those that don't.
Usage note: this phrase has negative connotations, implying a biased or dishonest presentation. It's a crucial term for discussing the ethics of data storytelling. Example: 'The report was accused of cherry-picking data to make the results look more impressive.'
The key takeaway — the main point or most important piece of information that you want people to remember from a presentation or analysis.
Usage note: this is a standard phrase in professional and academic settings. It's a great way to signal the conclusion or 'resolution' of your data story. Example: 'The key takeaway from this chart is that our strategy is working.'
5

Decide if each statement is true or false. Correct the false ones.

01The speaker's creative process involved a significant amount of chaotic sketching before he arrived at the final, polished charts.
02The speaker argues that the fundamental story structure of setup, conflict, and resolution is a relatively modern concept in communication.
03After 2005, the global housing market became fragmented, with some countries experiencing continued growth while others saw a decline and subsequent recovery.
04The speaker suggests that the primary goal of data storytelling is to foster an emotional connection with the audience, leading to a deeper impact.
05The narrative concerning Japan's housing market shows it following the same general trend as most other countries, but with more extreme fluctuations.
6

Discuss these questions with a partner. Try to use vocabulary from the lesson.

  1. The video champions the setup-conflict-resolution model for data storytelling. To what extent do you agree this is universally effective? Consider a professional context where taking data at face value, without a strong narrative, might be more appropriate or even more ethical.
  2. Reflect on a major economic or social trend in your country over the last decade. If you had to create a data story about it, what 'conflict' would you highlight to paint a compelling picture for an international audience, and what would be the key takeaway?
  3. Where is the ethical line between effectively teasing out a narrative from complex data and deceptively cherry-picking data to support a biased conclusion? Discuss a real-world or hypothetical scenario where this distinction could be particularly blurry.