Data

Data storytelling: How to use data to tell great stories

With great data comes great responsibility... to tell an engrossing story. It's no longer enough to have just the numbers, you need an articulate and engaging narrative to go with it. That's where data storytelling comes in.

What is data storytelling?

Imagine you’re asked by your manager to create a presentation for your team on the last years’ performance, and your company’s projected performance for the next three years. You’re given a big spreadsheet of rows and columns full of statistics, and you’re asked to extract some sort of conclusion or hypothesis from it – with no context. As you present the statistics the next day, you watch your audience. They look confused, or not engaged at all. Being presented with huge amounts of data is a daunting prospect, and enough to make anyone zone out, especially if you’re not a data expert.

The solution? You need to tell a ‘story’ with this data. Huge datasets like the one mentioned above can tell you ‘what’ is going on, but it is the ‘why’ that matters most. It is the overall narrative that you can create from the data that will make your audience listen closely and motivate them to instigate change.

Data storytelling is – in simple terms – telling a story using data to back up your narrative. A data story turns insights gained from statistics, facts and figures into a clear, digestible, and informative ‘story’ or explanation to be acted upon.

Storytelling is an ancient art that dates back to the earliest human records as a way to communicate. Our brains prefer being presented with a story over a lot of dense data, simply because it processes so much daily information that it needs to determine what is worth remembering. Storytelling also engages the language-comprehending, emotional, and empathetic side of our brains. If you can engage more parts of your audiences’ brains, they are more likely to remember the story and store it as a long-term memory!

But what makes a good story? The philosopher Aristotle recommended that a story must have a strong and defined beginning, middle, and end. He therefore suggested the three-act structure of a play. The first act establishes the plot, the second act introduces conflict, and the final act is the resolution. This is widely known as the setup, conflict, and resolution model, which is very useful to remember and apply when creating your own data story to present to your stakeholders. Present the data, then the problem, and finally the solution or action for your audience.

What makes a good data story?

A data story is made up of three core parts: data insights, data visualisation, and a narrative.

1. Data insights are the basis of the story. Your data insights are the detailed understandings gained after data analysis. Effective data insights allow an organisation to improve business processes based on reliable, solid data.

For example, maybe you want to know the overall online sales versus in-store sales of a particular product. This could be gained from descriptive analytics (see Descriptive data in Types of Data section below). Or, say if you wanted to predict sales of the same product in the next year, you could use tools such as AI or machine learning to gain your insights (see Predictive data in Types of Data section).

2. Data visualisations are any kind of visual aid, such as graphs, charts, pictures, diagrams, or videos (read on to discover what visualisations are in more detail).

3. A narrative is a written or verbal communication or storyline which supports your visualisations. For an effective narrative, using engaging storytelling techniques such as establishing the plot, presenting a problem, and resolving it, are key. Here, you must clearly recommend what specific actions your audience should take with the narrative you present.

A data story has the potential to connect with the emotional side of the brain, increasing audience engagement and allowing them to visualise, remember, and act upon the data story. An effective data story can improve communication, aid with making informed decisions, and connect different areas of the business.

The benefits of telling stories through data

Aside from simply being a more engaging way to present data, data storytelling has some real measurable benefits. It can:

  • increase understanding of sophisticated, complex data sets, metrics, and forecasts.
  • lessen ambiguity – it is important to make business decisions based on data, not just assumptions.
  • get senior leader buy-in.
  • empower stakeholders to make more informed decisions.
  • present and offer value to your stakeholders.
  • highlight key points to your audience for maximum impact.
  • enhance credibility as a business.
  • improve communication at all levels of an organisation.

It's a soft skill that's becoming increasingly important in the workplace. With many businesses looking to foster a data culture within their organisation, having team members that can articulate the numbers at their finger tips is becoming a much sought after attribute.

How to improve your data storytelling

Data storytelling is the combination of the ‘hard skills’ of in-depth data analysis with the ‘soft skills’ of constructing an effective, engaging story.

Data Apprenticeships and Data Training Courses have always included these hard software and analysis skills, but it's becoming increasingly common for this training to focus on the soft skills as well.

Like writing any good story, you need to have a plan. You can create this plan by asking yourself a few key questions:

  • What do you want to prove or disprove?
  • What data should you collect?
  • What is the purpose of your story?
  • What do you want to say and convey to your audience?
  • How will you make your introduction and conclusion strong?
  • What questions will you ask?
  • Was your hypothesis correct?
  • What is the goal you’ve created for your audience? What actions should they take?

You should not only consider data that supports your theory, but also the data that doesn’t. Rest assured you don’t have to add every single piece of data available, but you should use data sets that support and add to your narrative in an unbiased manner. A great data story will always lay out a clear roadmap for an action or change that can help your organisation.

Data storytelling skills and tips

How do you become a great data storyteller? It involves mastering a range of abilities. These skills include:

  • analysing and visualising data.
  • communicating with business stakeholders.
  • focusing a story on relevant insights.
  • the ability to generate accurate visualisations that can be adapted to the audience in question.
  • data literacy skills – reading, understanding, and working with data, as well as deriving actionable insights and communicating them effectively.

Considering your audience

When delivering your data story, you need to consider whether your data story is suitable for your audience.

Firstly, identify your stakeholders. This is anybody who needs to be influenced to act from your data story. This could be anyone from your CEO to a customer who wants to buy your product.

It is important to consider exactly what your audience is looking for from your data story – you’ll need to adapt the story depending on your stakeholders and their needs and expectations. You may even want to present multiple versions of a data story depending on who you’re presenting to and what action you want to persuade them towards.

Your story needs to have value and provide new knowledge. If your data story is lacking, this can damage trust and reduce engagement. Basing your data story around a clearly stated business need or question is essential to keep your data story on track.

Ask: is this narrative suitable for my audience? Check in with relevant stakeholders, or even enlist the help of a stakeholder to help guide your data story towards the best angle by fully comprehending their needs. Knowing what their background is and their level of understanding of the topic is important when deciding which parts of the data story to emphasise or omit. Make sure you’re teaching your audience something new.

Finally, make sure you’ve received and taken on board any feedback from your stakeholders. It’s rare that anyone would write anything and hand in the first draft – great storytelling needs to be edited and revised. Your data story should be informed by your stakeholders’ response. This in turn increases trust and shows them that you are taking their needs into account.

Types of data

There are many ways to optimise your data storytelling, and understanding how best to read and interpret your data is central. There are four main groups of data.

Descriptive – answers the question ‘what happened?’

- For example, calculation of the mean, median, mode or range of a particular set of data.

Diagnostic – answers the question, ‘why did this happen?’

- For example, running tests to determine the cause of a technology issue, or a study that aims to identify why a certain product is not selling well.

Predictive – answers the question ‘what might happen in the future?’

- Nowadays, organisations may make use of AI or machine learning to help ‘predict’ outcomes of decisions, such as whether a machine may malfunction in the future, or your company’s projected profits for the next year.

Prescriptive – answers the question ‘what should we do next?’

- This type of data combines information about possible situations or scenarios, the resources available to the team, past performance and current performance, to suggest strategic action.

What is data visualisation?

Data visualisation is the process of transforming large, inaccessible amounts of data into easy-to-digest charts, graphs, or visual aids.

Data visualisations support your overall story. They can help reveal patterns, provide context, and keep your audience engaged.

Human eyes are drawn to colours and patterns. Data visualisation is a form of visual art which can grab your audience’s attention and capture their interest. Most people looking at a chart can find a clear pattern or trend, and when we see something, we can internalise it.

However, even the most colourful, fancy graph can fail at conveying the right message – and likewise a boring, plain graph may fail to engage the audience. It’s a balancing act between the quality of data collected and represented.

Common types of data visualisations are charts, tables, graphs, maps, infographics, and dashboards.

What is a data dashboard? A data dashboard organises multiple sources of information in an easy-to-understand format and in a single location. This can help the audience visualise and connect the data story and the hypothesis that you’re trying to prove or disprove.

By upskilling in data storytelling, teams can take advantage of any opportunities in markets, and solve problems in new, creative ways.

Learn more about how QA can help your business gain in-demand data skills, and develop a business boosting data strategy.

About The Author

Natalie Beecroft is a Senior Content Editor at QA and has written for several publications and organisations.

 

Learn more about our Data Training