Inform Business on Findings: Data Storytelling – Why Do We Care?

Once businesses have started collecting and combining all kinds of data, the next elusive step is to extract value from it. Collected data may hold tremendous amounts of potential value, but not an ounce of value can be created unless insights are uncovered and translated into actions or business outcomes.

Data storytelling is the process of translating data analyses into layman's terms to influence business decisions or actions. The idea is to connect the dots between sophisticated data analyses and decision-makers, who may not have the ability to interpret the data. To date, there is no set of best practices on how to tell compelling data stories, but experts often describe data storytelling in traditional storytelling terms, which include a "hook" or a device to draw the listener or reader in, themes, the use of emotion and a conclusion or a set of conclusions.

Data storytelling weaves data and visualisations into a narrative tailored to a specific audience in order to convey credibility in the analytical approach, confidence in the results, and a compelling set of insights that is actionable to the audience.

Data storytelling is not ‘just’ data visualisation

Contrary to popular belief, data storytelling is not simply data visualisation, analytics reporting, or a handful of stats sitting in a PowerPoint somewhere. Data storytelling is the blending of two worlds: hard data and human communication. It is a compelling narrative crafted around and anchored by compelling data. The narrative is the key vehicle to convey insights, and the visualisations are important proof points to back up the narrative.

Data storytelling – What it really is?

Data storytelling is a structured approach for communicating data insights, and it involves a combination of three key elements: data, visuals, and narrative.

It is important to understand how these different elements combine and work together in data storytelling. When the narrative is coupled with data, it helps to explain to the audience what’s happening in the data and why a particular insight is important. Ample context and commentary are often needed to fully appreciate the insight. When visuals are applied to data, they can enlighten the audience to insights that they would not see without charts or graphs. Many interesting patterns and outliers in the data would remain hidden in the rows and columns of data tables without the help of data visualisations.

Finally, when narrative and visuals are merged together, they can engage or even entertain an audience. It is no surprise that people collectively spend billions of dollars each year at the movies to immerse ourselves in different lives, worlds, and adventures. When storytellers combine the right visuals and narrative with the right data, they have a data story that can influence and drive change.

Why data storytelling is essential?

Since the beginning of civilization, storytelling has been an integral part of humanity. Even in this digital age, stories continue to appeal to us just as much as they did to our ancestors. Stories play a vibrant role in our daily lives—people remember stuff that they actually enjoy, not what they deem important. Storytelling transcends industries and is often a focus in business books, conferences, sales training, and much more.

TED talks include a lot of storytelling. If you listen to the most popular TED talks, you'll feel that majority portion of the talk are in fact stories. Throughout time, storytelling has proven to be a powerful delivery mechanism for sharing insights and ideas in a way that is memorable, persuasive, and engaging.

Many of my colleagues feel that crafting a story around the data is unnecessary and a time-consuming effort. They argue that the insights or facts are sufficient enough to stand on their own as long as they are reported in a clear and concise manner. They may be right about their own way of data-driven thinking. But the revealed insights alone will influence the decisions and drive their audience to act—is based on the flawed assumption that business decisions are based solely on logic and reason. In fact, neuroscientists have confirmed that decisions are often based on emotion, not logic.

How to tell compelling stories with data?

Storytelling, specifically as it relates to data and analytics, is a routine challenge for many businesses. Often businesses find themselves in a position where they need to explain complex trends and data points to their clients. Organisations need good data storytellers to clearly and concisely convey why their clients should continue to invest in their services to continue to grow. So much of what they do is technical and behind-the-scenes, requiring familiarity with tech and a long attention span to grasp. This is getting increasingly difficult in today’s world, where people have a shorter attention span than goldfish.

To assess the extent of business needs, organisations should ask themselves the following questions:

• Are we currently using data to tell stories to clients? 
• Where can we improve our data storytelling?
• How do we educate ourselves and test these concepts across our company?
• How can we use storytelling to improve communication with our clients?

Considering human behaviour and psychology, here are some of the consideration when designing data stories:

• The use of colour is scientifically proven to draw attention to patterns more effectively than callouts or size.
• People typically scan rather than read, and when they do, they do so in F patterns.

• How pre-attentive processing, the subconscious and preconceived ways the brain filters and manages information could prevent people from successfully telling stories with data visualisations, need to be understood.

Data Storytelling Framework

Though there is no set standard for telling compelling stories with data, digital marketing company Two October compiled a list of 8 best practices or commandments for data storytelling, which is quite elaborate.

Begin with a question. To succeed at something, a goal is required – every goal is tied to a purpose, typically driven by a business problem or objective. It is the reason for doing it in the first place. Data within a story needs a purpose to meet the goal that is set for an audience, either in the form of a question or as a hypothesis. At the end, this is what the audience is going to learn.

Being clear with the answers to the following questions is crucial, and should be asked before processing the data:

• Why has this exercise been undertaken?
• What is the data that we are dealing with?
• What information are we trying to uncover?

End with an insight. Every analysis must end with some actionable insight, a solution to the business problem or ways to meet the objective.

Good stories conclude with actions to take and the predicted consequences of those actions. Of course, that means that analysts should consult with key stakeholders in advance to discuss various action scenarios.

If there is no actionable insight coming out of an exercise, the story is not worth telling.

Tell a compelling story. Analysts who can tell a story with data are the more effective ones. Regardless of the details of the analysis method and the means of getting it across, the elements of good analytical stories are similar.

To tell a compelling story one needs to think outside the box – combine words with images and data visualisations:

• They need to keep in mind what is interesting to the audience, not what’s fun to do as analysts. That can be very different.
• Putting it on paper allows them to start fixing it. If the findings stay only in the head, it is never going to be shared with anyone.
• What is the essence of the story? Most economical telling of it? If they keep that in mind, they can build out from there.

Explain with visuals, narrate with words. Though people understand metrics, trends and patterns better with visuals, one need to use words to add voice to the data. The visualisation must be functional, and that it tells the story effectively. So, the narrative and visuals complement each other. The visuals should not distract the audience from the story.

To achieve that, one needs to make sure that the visualisation has all the necessary information on it, clearly visible. In terms of putting the whole story out there for the reader to get, it is better to put that on the graph itself: title the graph, and make sure information is properly labelled. One might also want to include a line of text that reiterates the visual’s main point.

There are ways to make sure the visual is appealing, too. A clear, sans-serif font should be used, so that even on a small scale like a social media post, the reader still has a chance to see the text. Big blocks of bold colour and sans-serif fonts tend to go well together. The right kind of chart for the situation is also something to consider.

Data visualisations are not just for presentations. Data can be incorporated as part of content pieces on many different mediums. One can even go so far as to animate the graphs or make it interactive. Data can be used in videos, infographics, social media posts, posters, GIFs, or any other visual medium.

Be honest and credible. One of the most important aspects of working with data, but least appreciated, is how the findings are communicated. Even when all of the right steps are precisely followed, if the insights are poorly communicated or mislead the audience it can be worse than not using data at all.

Data can be used to create stories that lie, mislead, or inform correctly. To have a credible data story, the followings must be done:

• Start with the big picture. Frame all of the data communication within a bigger picture. This makes everything easier for the audience to understand and gives a strong start to the story.
• Show context. The more context we provide, both visual and narrative, the less likely the audience will be to jump to mistaken conclusions.
• Highlight hidden insights. Many of the most important insights in the data will be hidden below the surface. Highlighting these and contrasting them with the overall metrics is necessary to make a powerful story.

Here are a few things that must be avoided:

• Manipulate Scale. The scale of data and the units must be clear. Charts with multiple axes must be created with due care. It is better to over-communicate context about the charts than having the audience misunderstand.
• Cherry-pick data. The full breadth of data in communications should be used, not just the data points that help make an appealing case.
• Be inconsistent. The same colours, labels and conventions should be used across all the statements and visualisations in the story. Doing so creates a natural language for the data, one that the audience can learn.
• Lie. Any of the above should not be done purposefully.

Be clear and concise. There is such a thing as too much data. When analysts are telling a customer story, it is important to remember that the readers/viewers only have so long of an attention span to devote to reading or viewing that story. When making data points, keep them concise and clear.

Information overload is a common pitfall with data reporting. Data analysts can feel that all numbers are important, wanting to include as much data and evidence as they have uncovered. While that is not to say that all numbers are not important, there is a time and a place for each and every one. Data storyteller’s role is to focus on simplicity, taking complex or disparate information and making it tangible, understandable, and, importantly, more human. Where simplicity reigns, the user understands.

It is also natural to feel compelled to share everything; including the path we took to get there, but it should always be avoided. It impedes decision making and at the same time, unleashes a host of unnecessary questions. One should start with the end in mind – what action or outcome is expected? From there, data points that paint a clear picture should be highlighted.

Know and cater to your audience. Storytelling is not a one-time event. It is common to present the same findings again and again to a different group of audience. Perhaps different business units or different level of stakeholders. Catering the analysis to suit the preferences and expertise of the audience is a critical step. Structuring the story around what the audience know, do not know, need to know, and do not need to know is a good way to begin.

Here is a list of helpful basics when it comes to appealing to the audience:

• Listen to the audience. It is important to understand what each group wants to hear and how they want to hear it. The more analysts listen to the audience, the better they can craft an interesting story that speaks their language while also answering their questions and concerns.

1. What are their main goals or priorities? Can the findings, impact, or success be related to their goals or priorities?
2. Were there similar challenges in reaching analytics objectives? Even distant ones such as limited access to data, manual reporting hiccups, or navigating bureaucracy may help attract and keep their attention. 

• Make the story relatable. Once the findings are tied directly to the audience, analysts need to consider how they will share the story. Depending on whom they want to reach, determine their delivery, context, and/or wording so the story directly relates to specific stakeholders. One must remember what works for one group may not work for another.
• Keep it simple. When the story is easy to understand, analysts are more likely to get exactly what they want. So, in most cases, they need to avoid complicated terminology and insider lingo. Otherwise, they risk being interrupted every five minutes for an explanation or, worse, alienating the audience altogether.

1. The easiest way to keep it simple is to make sure that the analysts are prepared. Winging it wastes time and complicates the story as they try to find the right words.
2. If it is simple to deliver, it is easy to understand—meaning it is easy for the audience to remember.

Provide context. Every set of data should relate to the organisation’s larger goals, and analytics dashboard or data visualisation should highlight the correlations between each metric and these KPIs.

In order for the audience to care about the data, analysts need to establish context. Analysts must clarify why they should care, how points relate to each other, and how their (audience) actions or inactions will impact the company’s overall end goals.

Context can help establish a relationship between the data and the audience. Once the audience becomes invested in the data on a deeper level, they will then be able to think in terms of the “bigger picture”—as in, “if I do this, then that will increase these metrics, but if I do not do that, it may have a detrimental effect on those metrics.”

Concluding remarks

These connections, simple though they may be, help to drive informed decisions that take into account the long-term vision as opposed to the short-term buzz.

In addition to establishing connections between the data and the audience, context helps to establish connections between seemingly random datasets. If analysts were just to present two sets of numbers without giving them context, the audience might be confused, if not a little overwhelmed. However, when a backstory is provided for each set, they’ll be able to see the relationship and understand how each is affected by the other. This is important in helping the entire organisation understand how each department directly affects the others, and how they can all work together to achieve a greater vision.

Acknowledgement

Part of the content is sourced from Brent Dykes book Effective Data Storytelling: How to Drive Change with Data, Narrative, and Visuals (2019)