Data Storytelling: Examples that are Good and Bad

When I am trying to learn a new skill or develop thinking around a subject, such as data storytelling, I find examples from others helpful. The trick to reviewing their work is thinking about what is effective, not what is good or bad.

Sometimes I learn more from the examples that are ineffective then the effective ones. The effective examples are just good and you know they are good. The ineffective ones help you understand why certain approaches don’t work and, of course, avoid the issue yourself.

Example 1: Data Storytelling Must Make Impact

You may think a data story has to be a video or have lots of numbers. Watch this data story that Ben Wellington presented during his TEDx Talk. Wellington points out that he wants his data storytelling to have Impact!  This is the same as being persuasive. Why analyze data if you don’t want to change minds? Effective data storytelling offers a takeaway.

One point that he makes earlier in the talk is that it is easier to tell a data story when you talk about what you know. He understands urban life and often uses the NYC data portal to look for stories. Here he shows a simple chart of traffic tickets around fire hydrants, but this one chart has a significant impact. So don’t think that a data story has to more than a single fact.

I’m starting at the 11:21 mark – the whole talk is worth watching.

Example 2: Create an Emotional Impact with Data Storytelling

When you can bypass the logical mind and talk directly to the viewer’s heart – you’ll have an emotional impact. It’s how you make them care about your message. Data storytelling must have an emotional impact.

This data story has an emotional impact. Do you really understand what a million lives might look like – what about when millions and millions of lives are lost. This data story illustrates not only what the United States sacrificed, but certainly what Russia also sacrificed. It gives a new dimension to what impact Hitler had on Europe as a whole. It had a lasting emotional impact on me.

Example 3: Lead the Audience to Your Point

This data story was a great example of effective data storytelling – and since it is in Chinese I had to read it. But I watched the entire video and was waiting to discover the answer! [More tips for persuading audiences … ]

The storyteller wants to persuade you that she can use an algorithm to determine the best subject for an art award. Then she walks you through the data – which is artwork so you understand what the machine saw. She leads you right to the same conclusion the machine had. Tell me if you are convinced and thinking of entering your artwork.

Example 4: Data Stories Should Drive a Message

Some data storytellers are really just presenting facts and not persuading you toward a conclusion. In a recent blog post, I talked about the real point of data stories. I think there are two kinds: informative and persuasive. Informative data stories are really just that – presenting facts. Persuasive data stories are about bringing people to the same conclusion you have about the data. That is effective data storytelling.

This data story is a good example of being closer to the informative than persuasive. If I had to give it a theme it would be – “You can’t predict hurricanes” which doesn’t seem like a new fact. However a data storytelling example, it is well presented and well visualized.

Example 5: People are Interesting, Not Numbers

Hans Rosling made the statement “Numbers are Boring, People are Interesting” in one of his Ted Talks and it’s really true. What made someone do something or say something interests me more than their age or number of degrees.

As much as I love data, even I had difficulty following the message and taking in all the numbers. And the ending wasn’t very satisfying or even new information, “If you are single, try online dating.”

Ok … is this something no one has figured out in our online obsessed society? (Hint: has 10 million daily active users.)

I think that is where this data story goes off the track – its over-focused on numbers. In fact, it is so overfocused that it doesn’t seem to have a message. [Here are 7 things a data storyteller must get right to be successful.]

The storyteller was very interested in visualizing the research while missing a golden opportunity to draw in the audience by presenting information about the human aspect of dating.

Example 6: Find the Compelling Data

Many times data storytellers fail to understand that if you don’t have a compelling point in your story you might also fail to have any impact. The Freakonomics team does a great job of finding the compelling threads in the data or at least asking the question in a compelling way. Effective data storytelling finds a compelling data theme.

The present question many of us have and provide some supporting data that actually points to an unexpected conclusion. In this example they ask – should you hire a real estate agent to help you sell your home. Answer: Maybe.

Example 7: Think of Your Data in Different Ways

It would have been easy to just provide a bar chart that showed how tall each mountain was – but why not let the mountain do it?

When being visual you have to think about making the connection and keeping the numbers simple. It was difficult to follow the online dating story numbers above, but this story helps you think about the size of these mountains. It’s great data storytelling!

In this beautiful data story, the storyteller wants to remind you to visit the national parks. And the data is offered in a visually stunning way.

Tip! The author of this data story recently wrote a data visualization book.

Is There a Right Way to Do Data Storytelling?

I don’t like to use the word “right” when talking about presenting data. It’s restrictive and implies that there is a set formula. There’s not.

It’s about reaching your audience and presenting data in a way that influences them. Let’s call it an effective way.  If you look at any of these examples above, none were really wrong, they were just ineffective.

Here is the Zencos four-step method to creating a data story.