Data Storytelling: Transforming Data into Decisions

There’s no doubt that stories have a profound impact on those hearing them. Try to convey an insight to someone using only statistics or facts, and it’s unlikely that they’ll fully comprehend the significance of the information, let alone remember it in detail. However, if those stats and figures are turned into a story that takes the person on a journey through the insight and context around it, they’ll gain a deep understanding of a complex issue that might have otherwise been bewildering to a non-data scientist, had it come in the form of lines of data.

The memorability of storytelling goes far beyond what’s experienced with the average piece of information. In fact, Jennifer Aaker from Stanford’s Graduate School of Business found that people remember stories up to 22 times more than facts alone. Defining moments in stories make for “short experiences that are both memorable and meaningful,” according to “The Power of Moments: Why Certain Experiences Have Extraordinary Impact” by Chip and Dan Heath.

The potential for data-powered decision-making is endless. A recent survey by KPMG found that most of the senior business executives surveyed use data and analytics to make faster (86 percent) and more accurate (80 percent) decisions. Coupled with amazing storytelling abilities, companies have an invaluable way to communicate insights that are immediately understood, actionable and remembered thereafter. Let’s take a look at why data storytelling is crucial in deriving value from businesses analytics, from the first conceptualization of an analysis to it being presented to C-suite management, and even straight to the decision-maker. 

Starts with the Data Scientist

The value of data storytelling is demonstrated in all arms of the business, so it’s vital that data scientists hone this skill, as they’ll be the ones that conceptualize the insights they uncover into a clear and captivating story. And while programmers, developers or data scientists might not be used to focusing so much energy on creative pursuits – with many thinking it’s their technical ability that’s of value – it’s a skill that they’re going to need to develop. Insights that have not been expressed in a comprehensible, contextualized or interesting way are going to result in no action taken.

Data scientists need to remember to make the insights they find as clear as possible to other nontechnical readers. While the analysis may be easily understandable for them, they have spent time submerged in the data, and what’s obvious to the programmer will not be as clear to the CEO, or the client for that matter.

For example, a government health promotion agency came to us wanting to know how they could best divide their users, and then provide focused marketing initiatives for respective clusters based on user behavior. By conducting a data analysis, we found that there were three distinct clusters that can be used to characterize users of the fitness watch. These included a “cheating” cluster (users who had an excessive amount of points, likely accumulated through unfair means) along with the “non-starters” (users who were not very active in terms of step count, and did not take part in many challenges) and the very active users (these people were good at completing challenges, but their height/weight ratio was not ideal).

However, without turning these insights into a contextualized visual narrative that told a story, the client would have had no idea as to the distinctive characteristics of each cluster. We created a color-coded grid, where each square representing a characteristic of each cluster faded in and out as the user slides the bar along. This allowed for the person using the visualization to quickly gain precise insights into the various characteristics of each cluster, and how they interact with each other. 

Making Insights Clear for the Whole Team

A lot of what data storytelling is about is accessibility and its ability to reach a wider audience. This is no more applicable than in the communication between data scientists and the rest of the business. Team members, from sales executives all the way to C-suite management, now have a vehicle through which they can communicate the importance of these data insights.

Storytelling helps not only in the last mile connectivity, but also to map the mental models across all team members. Without it, the data scientist’s analysis can achieve nothing by keeping the insights to themselves, as their lack of accessibility keeps them from reaching a wider audience.

For example, a data scientist may uncover insights and present them in traditional rows and columns spread over multiple pages, which does not make it easily digestible and understandable for the decision-maker. However, if multiple pages of output can be shortened into one page that leverages the appropriate data design, the team member interpreting the data can turn it into actionable insights in a much easier and more effective way. 

Using Data to Connect with the Client

Once the client services or customer management team have the insights and data stories, they need to connect with the client. Once the agency is armed with the analysis on how the users can be grouped, it can adapt its strategy to suit the needs and wishes of each user cluster. In fact, in our example, with these insights the client services team decided they did not need to remind a certain group to exercise, instead treating them as the “champions” of the user base.

Before an analyst tells a story with data insights, the insights are useful yet non-obvious facts. After the storytelling process is done, they remain not only useful but have become memorable and viral. Telling a story with the data means that the client is likely to first know what action they need to take and how to prioritize it, all the while making the insights memorable – meaning they’re more likely to act on it in the future.