Data storytelling: A key skill for data-driven decision-making
IT and analytics leaders seeking to convey actionable insights from their organization’s data must learn how to tell compelling stories with data, emphasizing context and narrative.

To be a truly data-driven enterprise, organizations today must go beyond merely analyzing data. Rather, business experts and IT leaders must transform relevant data into compelling stories that key stakeholders can readily comprehend — and leverage to make better business decisions.

This vital skill is known as data storytelling, and it is a key factor for organizations looking to surface actionable information from their data, without getting lost in the sea of charts and numbers typical of traditional data reporting.

Following is a look at what data storytelling entails and how IT and analytics leaders can put it to work to make good on data’s decision-making potential.

What is data storytelling?

Data storytelling is a method for conveying data-driven insights using narratives and visualizations that engage audiences and help them better understand key conclusions and trends.

But that’s often easier said than done.

“Telling stories with data can be difficult,” says Kathy Rudy, chief data and analytics officer at global technology research and advisory firm ISG.

For Rudy, data storytelling begins with knowing your audience.

“Remember to start with who your main characters are, that is, the audience for your data story. What information is most important to them? Structure your data story so you anticipate the next question the audience will have by thinking like the reader of the story,” says Rudy, adding that, in her 20 years in benchmarking and data analytics, she has had to learn to tell a clear and concise story using data to validate ISG’s recommendations.

The first hurdle most data storytellers face is gaining acceptance for the validity of the data they present, she says. The best way to do this is to hold data validation and understanding sessions to get the question of data validity out of the way.

The goal of the data storyteller is to clear up all questions as to the source of the data, the age of the data, and so on, so that in subsequent views of the data, the storyteller isn’t continually defending the data, Rudy says.

“Don’t get overly technical or you will lose the audience,” she advises. “In the case of IT benchmarking, they don’t want to know about the technology stack, just that the data is relevant, secure, current, comparable, and accurate.”

Elements of data storytelling

Data storytelling consists of data visualization, narrative, and context, says Peter Krensky, a director and analyst on the business analytics and data science team at Gartner.

“With visualization, a picture is worth a thousand words,” he says. “How are you making the story visually engaging? Are you using a graphic or iconography? That doesn’t mean it can’t be a table or very dry information, but you’d better have a visual component.”

The narrative is the story itself — the who, what, where, why. It’s the emotional arc, Krensky says. “If it’s about sales forecasting for the quarter, are we doing great, or are people going to lose their jobs?”

Context is what the people hearing this story need to know. Why one sales representative is always outperforming all the other sales reps is an example of the context for a data story, Krensky says.

Grace Lee, chief data and analytics officer at The Bank of Nova Scotia (known as Scotiabank), says blending context and narrative requires a keen understanding of what makes a story compelling.

“The way that we think about stories, if we remove the data term, it needs a plot that you care about, it needs characters that you root for, and it requires a destination or an outcome that you believe in and aspire to,” she says.

Being able to put the data into context in the form of a narrative allows people to care more and to understand what the action is that comes out of it, Lee says. In addition to focusing on storytelling as a discipline, Lee’s team is also working to create more storytellers across the organization.

“The way we’re educating people around storytelling is really around action orientation, helping people create those narratives, providing more of the context, and allowing people to see the clear line between the data, the insight, and the action to come,” she says.

Lee sees the role of Scotiabank’s data and analytics organization as the storyteller for the enterprise because it’s only in the data that some of the insights about what customers need and want appear.

Key steps in data storytelling

Lars Sudmann, owner of Sudmann & Co., a Belgium-based consulting and management training network, offers insight into the steps that go into data storytelling.

  1. Identify the ‘aha’ insights: One of the greatest pitfalls of data-based presentations is the “data dump.” Rather than overwhelm the audience with data and visualizations, CIOs and data analytics officers should identify one to three key “aha” insights from the data and focus on these. What are the surprising, absolutely key things one needs to know? Identify them and build your presentation around them, Sudmann says.
  2. Share the genesis story of the data: To tell a good story with data, a good starting point is the genesis, i.e., the origin of the data. Where does it come from? This is especially important when storytellers present data sets for the first time.
  3. Transform surprising turning points into engaging transitions: When storytellers present data and facts, they should share where the data/graphs/trendlines make “surprising” moves. Is there a jump? Is there a turning point? Doing so can provide compelling transitions to deeper analysis, for example: “Normally we would think the data does X, but here we see that it declined. Let’s explore why this happened.”
  4. Develop your data: One of the biggest issues in giving presentations today is that people throw heavy data on the screen and then play “catch-up,” with words, such as “This is a crowded slide, but let me explain.” “This might be difficult, but…” Instead, storytellers should develop their data step-by-step. “I am not a fan of fancy animations, but for instance in PowerPoint there is one animation that I recommend: the ‘appear’ animation,” Sudmann says. “With it one can harmonize what one sees and what one says and with that a data story can be built step-by-step.”
  5. Emphasize and highlight to bring your story to life: Once storytellers have identified the flow and key aspects of their data stories, it’s important to emphasize and highlight key points with their voices and body language. Show the data, point to it on screen, walk to it, circle it — then it comes to life, Sudmann says.
  6. Have a ‘hero’ and a ‘villain’: To make stories more engaging, data storytellers should also consider developing a hero, e.g., the “good tickets,” and a villain, e.g., “the bad tickets raised because of not reading the FAQs,” and then show their development over time, in different departments, as well as the “hero’s journey” to success, Sudmann advises. 

Data storytelling tips for success

Rudy is a firm believer in letting the data unfold by telling a story so that when the storyteller finally gets to the punch line or the “so what, do what” there is full alignment on their message.

As such, storytellers should start at the top and set the stage with the “what.” For example, in the case of an IT benchmark, the storyteller might start off saying that the total IT spend is $X million per year (remember, the data has already been validated, so everyone is nodding).

The storyteller should then break it down into five buckets: people, hardware, software, services, other (more nodding), Rudy says. Then further break it down into these technology areas: cloud, security, data center, network, and so on (more nodding).

Next the storyteller reveals that based on the company’s current volume of usage, the unit cost is $X for each technology area and explains that compared to competitors of similar size and complexity, the storyteller’s organization spends more in certain areas, for example, security (now everyone is really paying attention), Rudy says.

“You have thus led your audience to the ‘so what’ part of the story, namely, that there are areas for improvement,” she says. “The next question in your audience’s mind is mostly likely, ‘Why?’ And finally, ‘So what do we about it?'”

The rest of the story leverages a common understanding of the validity of the data to make recommendations for change and the actions necessary to make those changes, according to Rudy. Data in this story created the credibility necessary to establish a call to arms, a reason to change that is indisputable.

And taking the old adage “if a tree falls in a forest and no one is around to hear it, does it make a sound?” into consideration, it’s crucial for data storytellers to consider the medium various individuals are using to consume information and what times they’re accessing this information.

“The pandemic has definitely helped in the shift of allowing thought workers to work from home,” says Kim Herrington, senior analyst for data leadership, organization, and culture at Forrester Research. “And a lot of times you’re communicating with thought workers that are across the globe. So it’s important to think about the communication software that you’re using and the communication norms that you have with your team.”

IT leaders get creative to fill data science gaps
The ongoing data scientist shortage sees enterprises reconfiguring data teams, upskilling promising employees, and partnering to improve talent pipelines.