Home > Blogs > Creating Data Stories: A Critical Skill for Every Data Scientist
When you consider that humans are visual creatures, the implications for today’s data-driven businesses become clear. Given that the world generates 2.5 quintillion bytes of data every day, deriving real value from it requires businesses to present the right information in a visual format that is easily understood by end users. While graphs, pie charts, bar diagrams are good, they do not connect the different parts of a business problem to present a complete picture to help users arrive at an insightful solution. LinkedIn’s recent survey reiterates the dominance of data analysis as the hottest hiring skill, yet organisations often miss the last mile of analysis i.e. converting data into easy-to-understand insights. That’s where data storytelling comes into the picture.
Telling the story with numbers goes beyond simple visualisation
Data scientists proficient in programming, modelling, extracting, cleaning, and other data crunching tasks are a great find. But if they are not good storytellers, there is a good chance that their work will gather dust on the shelf. Storytelling is a critical skill for data scientists as it not only gives them the power to influence stakeholders and business outcomes, but also helps them sharpen their data presentations capabilities.
Remember how Sheryl Sandberg adding her personal story to her TED talk at the last-minute sparked a movement? Now 65% of content on TED Talks comprise stories. Stories work wonderfully – not just on stage but in business too. However, too often data storytelling is associated with visualisation, with a focus on creating visually engaging, static or interactive infographics. But can a series of rich visuals, stats and facts engage users meaningfully? Think of data visualisation as a photo portrait – there are many facets of a picture a user can gather from a portrait but not a complete story.
What it takes to create compelling data stories ?
The idea of data storytelling is to go beyond visualisation - to take the viewer on a data-driven journey. It begins with the process of formulating a question and culminates in arriving at a desired conclusion, in a step-by-step manner. The most important aspect of a data story is the narrative or the story arc. It should begin with a clear ‘data ask’ which leads the direction of the story, much like a business problem leads the work of a data scientist.
A good story arc begins with exposition, followed by rising action, climax, and finally the conclusion. The exposition sets the context and defines the universe of data being examined. The rising action builds up on the ‘data ask’, explores the facts, and asks relevant questions. The climax centers on the critical discovery through data that provides answers to all the niggling questions that a user is trying to solve. Lastly, the conclusion refers to the important takeaway(s) that a data scientist wants to leave the viewers with. To keep the story narrative tight, data scientists must ensure that the plot is built backwards to support the conclusion and take care to remove any extraneous tangents along the way that can confuse end users. Supporting story arcs with fictitious characters, building an analogy, or throwing in real world examples helps spice things up. Google Trends 2017 video does a great job of combining data from graphs and charts into stories about what people across the globe search for.
Emotional connection, more than logic, drives business decisions
Creating data stories is time consuming but science has proved that business as well as consumer decisions are made on emotions, and justified with logic. Take for instance, Apple’s emotional branding strategy that inspires people to become part of a lifestyle movement, hinging on humans’ basic emotional need to belong. Statistics, facts, and figures are what people hear and see, stories are what they feel. While uncovering data insights can be incredible, for data scientists to really usher transformation, it is critical that they communicate their work through impactful stories. The good news is data storytelling is as much an art as it is science - it can be easily inculcated through robust and well thought outlearning and development programs.