Data-driven storytelling — from evidence to impact
The power of data lies in the stories it tells
Walk-through of the process of transforming raw data into effective narratives and visuals.
Irrespective of the field of work you are in or how data-savvy you are, chances are high that you had to use data to get your point across to a broader audience. Data backed ideas have become an absolute norm in almost every industry, and everyone is almost required now to be able to use data, if not to analyze then at least to be able to tell a story with it.
Unfortunately, what is also common is seeing a slide which is full of data that you can’t wrap your head around (I have been both a culprit and a victim).
Although, most of the slides are not as bad as the one above, you get the impression. Over the last year, I have been curious at times to understand what makes a data story great? What makes it effective?
In this short article, I want to describe many of the principles leading storytellers use — a narrative arc, visualization, interaction, and more — so you can more easily create persuasive and engaging stories. I want to go through some quick principles that allowed me to get better at telling a story from the data at hand.
Note: In no way, is this a comprehensive list and an airtight one but just what allowed me to think about the concepts in a clearer manner.
The principle covers three broad areas:
- Data to understand the scenario
- Data to form a narrative — exploratory or explanatory?
- Data to visuals —Let the data dictate the kind of visual (charts and graphs) you are using
1. Data to Understand
First of all, I think the core purpose of data is to really try and understand what’s happening. With more and more data at hand though, there are many patterns that start emerging, as soon as you start diving deep. The challenge is then to select a few data routes that you want to really explore. It’s about picking the right lane.
An easy checklist would be to force yourself to answer Why, What, Who, How, When, and Where — for each time you start digging deeper. Generally trying to answer most of these questions can provide some great directional information on what you are looking for. It’s not important to answer ALL of these. Try and put this down for any time investment you are going to make with collecting and analyzing your data, an upfront 2-minute exercise can help you save a lot of time down the line.
For instance, for this article I could think about what data analysis here could help me make my point that data storytelling is crucial but is often done wrong?
Why — This could help readers know that its an issue many people face.
What — A survey that identifies how many people complain about data presented in the wrong way
Who — People in a business context
How — By not doing this exercise
When — In business presentations or quarterly meetings, etc.
Where — Business meetings, etc.
For instance, after having answered the above, I might look for quantitative or qualitative data (perhaps quotes from some of the famous leaders) that help me make the point.
A quick google search like this would tell me that about 70% people surveyed recently agree that most presentations with data suck (obviously, I would not take the first thing Google gives me, but only from a reliable source and after carefully making sure this was done in the context that I am looking for, etc.).
The key point here is that I know what kind of data I could look for that helps me build the context for the audience reading the article.
2. Data to form a narrative
Once, you have the key data that you want to use to understand the problem, you then start digging in that data to find a narrative for the point you want to make. For instance, in the image above The Pudding is very clearly demonstrating a narrative based on the graph we are seeing (more on that later). It’s very clear to any observer that the next slide is going to build on that narrative — are we at peak beer hype?
3. Data to Visuals
The final part, is to actually visualize the data in the context of your narrative that you want to present. Determine what you want to show; relationship, comparison, outliers, composition, and then use the charts that demonstrate the point and generate the desired effect.
For instance, look at the visualization below from Mckinsey which illustrates which functions have faced the major changes due to data and analytics.
The visualization here presents a complex idea in a very easy to read format. From this graph, it seems like sales and marketing irrespective of the type of organization have been majorly affected.
Using bar charts, pie charts, and stacked columns is a decision that can help you elevate your narrative and make it clear to the audience what you are showing them. The audience ideally reaches the same conclusion as you want them to reach and the one that adds to your broader narrative.
In Conclusion
Data storytelling is not as. easy as it seems on the surface. It gets tougher, especially when you are the one who has worked on the analysis behind what you are presenting. Taking a step back, using the three-steps laid above i.e. data to understand, data to form a narrative and data to then visualize and guide your audience to that narrative can help you get better and effective at storytelling through the data.
Recommended material:
- Develop Your Data-Driven Storytelling Chops with Four Simple Principles
- “Storytelling with Data” | Talks at Google
- Data-Driven Storytelling
