The word data is thrown around a lot, and nowadays there’s no part of a business that doesn’t use it. The rate that companies gather data today is bigger than anybody could have imagined a couple of decades ago. But having a big collection of data alone is not enough–it’s what, how, and why you use your data that really can make a difference.
With the ever growing list of Business Intelligence (BI) and AI tools, data is becoming increasingly accessible and easy to gather, and dashboards and spreadsheets are quicker to build. But companies are still struggling to harness the power of their data–why is that?
What they lack is a key storytelling component. Dashboards can only ever tell you what is happening; they can rarely tell you how, or why, or what this means for a business. As awesome as its power is, data alone will never be enough.
In a time of information overload–with screens on our desks, walls, and even our bodies–storytelling to cut through the noise is more important than ever. And this is where data visualization comes in.
In this article, we’re diving deep into:
Data visualization
Using data to tell a story
How to create a compelling data story
Using narrative as a tool
…with some real life examples to tie it all together. Ready? Let’s dive in.
What is Data Visualization?
Before we dive into storytelling, we should have a base understanding about some key principles in data visualization.
Our eyes are naturally drawn to colors and patterns. It helps us identify and focus on key points. Simply put, images speak louder than words and data visualization helps tell a story by curating information into an understandable form, highlighting hidden information and guiding the audience through a path that leads to a conclusion, or to discover something that was not so obvious from raw data.
The key elements that are played together to help data be more clear are: size, color, prder and scale:
Size can help emphasize information and add context to the user. Size is easier to be adjusted to the values and has a more direct connection to the information itself.
Colors are another element that can help the viewer understand the context and make associations between the elements. It is especially good to categorize your different labels or to emphasize the intensity.
Order should help reduce cognitive overload. When data points are pretty close together, ordering, for example from largest to smallest, will help you automatically see the most important points.
Scale is important to not give misleading information. We tend to associate the magnitude of the data by its scale, so keep that in mind when building your visuals.
One thing to keep in mind is that your visual will be useless if only you can understand it. You need to have a good picture in mind of who you are designing the visual for, and keep your audience in mind when creating it. If your viewer is not so familiar with the data, or it's their first time seeing it, be more instructive and give directions to guide the user experience.
Besides this product mindset, there are 4 commandments that every visual must follow. Let’s dive in.
Up down, left right
This speaks to the order that information should be presented in your graph. All of Western culture reads and writes in a very specific way, that is, starting from the top left corner and going from that direction downwards. Keeping this in mind when building a dashboard can be useful to guide you selecting which visuals are going to be present first, and then prepare your story so that your audience reaches the right conclusion.
Color correctness
Our brain is so powerful and it is built in a way that makes us look for patterns and associations all the time. Using this in your favor can make your visual easier to read and coherent; failing in color correctness can lead to misleading analysis. A good example can be making all data related to monetary features green or dividing your labels into different colors and keeping this consistency throughout your different visuals.
Filtering
Filtering is specific to interactive visuals (e.g dashboards). This can be a powerful tool to present more data without necessarily having it displayed all at once. However, be very careful where you put your filter, how well you signal it, and where the data is being filtered.
Granularity
This is the amount of detail in your visual. Is good to start with low granularity and end with high granularity. Something that is very usual to happen is when you reach the end of your presentation and some colleague asks you for the Excel spreadsheet of the data. Some people tend to need to see a table to believe what you’re saying, so it is a good idea to have this high granularity element at the end of your visual.
There are 2 main forms of delivering a data story. It can be explorative or narrative:
Explorative forms are delivered usually in Powerpoints, PDFs, or dashboards.
Narrative forms can be delivered in presentations, talks, meetings, or in dashboards as well.
Today we will focus more on the narrative type of data storytelling, but keep in mind that exploring your data is necessary and you might need to do it first before building your story.
Telling A Story With Data
A good data story is composed of trustworthy data, well-designed visuals, and a compelling narrative. The data aspect is straightforward–we must have accurate data to reach trustable insights. The visual elements help us visualize (duh!) the data better, finding trends and insights that are not easily seen in the rows and columns of a common spreadsheet. The narrative part comes into play to give voice to the data. Looking at raw data to get information or to prove a point is just the hardest thing to do. We use data visualization because we want to help the viewer understand the message that we want to pass on.
Each data point can give a message and the combination of them creates a story that voices the insight that you are looking for. Data storytelling is an attractive way to communicate the story that you found inside your data. It also improves the credibility of your data, as you look to find connections. You can have a brilliant idea or message that you want to transmit, but if you don’t know how to do it in an effective manner, your audience will counter argument you or create other narratives to explain to you the things that you have found.
“People need a narrative, and if there isn’t one on offer, they make one up.” Jean Hanff Korelitz
You don’t need to be the best storyteller that there is. It is a question of finding the right elements, understanding the simple “Pixar” formula and remembering the most important concept: “Show, don’t tell."
Storytelling is the art of delivering, developing and adapting creative stories utilizing specific elements: characters, context, conflict and message. Translated into the data world, these elements look like this:
The characters are your data points, your features or your KPI’s. They are the ones that go into the journey.
The context is your business setting. This is your start point and should be present at the beginning, as in “Once upon a time…”
The conflict is the reason that you are doing the story. No story is good without a conflict, people must be presented with a challenge to make the story exciting. In data this is the problem that you are trying to solve, or give insight on.
The message is your insight in itself, or the key takeaway you are trying to make your audience understand. At the end of the day, your audience may forget you, or the title of the presentation, but they must remember the message.
Define the objective of your story since the beginning. Do you want to uncover a threat to your company, reveal an interesting insight or just tell a funny story?
Creating a narrative
As we said before, a complete narrative contains characters, obstacles, conflict and a well-defined journey of transformation for the protagonist. Translating this into a data story is tricky because things may not seem so obvious at first, so let’s think of it this way: the data isn’t the story, but the story is the structure that your data will fit in. Following the “Pixar” model, you will find these sections in your journey.
Presentation
“Once upon a time…”. Insert your audience into the context. Their minds are not yet fully prepared to receive all the information, so start slowly introducing the different features that you are going to work with. For example, if you are doing a Sales presentation and your company has different departments, start by showing them to put this in their mind as you advance in your story.
Conflict
After that, it is not good to keep the viewer waiting for some action, or people might start to sleep. Present your conflict, or the point of interest that motivates you to follow the direction that you are going. Because of this conflict, you will need to show a bunch of different other features that along the way will help you solve or give an insight about this. For example, if your Sales team is struggling recently in profiting, this is the struggle that you want to show.
Journey
Focus on compelling visuals and comparisons that will show the different facets of your conflict and give basis for the next steps.
The change
We are now presented with the takeaway, or the pinpoint data point that demonstrates the reason for the conflict or the tool that can help solve it. Use that moment to deliver your message and convince people that the journey that you took is justified. You can show, for example, that the state of Tennessee has had a significant increase in costs and selling there isn't the best option right now.
A second challenge
It is common that we are faced with another challenge that we will have to deal with. This happens because nothing in real life is so simple that with only a change on a lever will solve your problems. Ok, we can stop promoting sales in Tennessee, but maybe we will lose future customers there. Be ready for these counter arguments and use past data points to address these questions.
The conclusion
Returning to the present and present recommendations for the future; this is very interchangeable and is difficult to put it in a formula. This is your chance to put the whole story together and leave your presentation on a high note.
How Real Companies Use Data Storytelling
Spotify “Wrapped” is a great example of a compelling data story that has the unique element that every company looks for: it’s personal. It was presented in 2006 for the first time and it's been a sensation since the beginning. As a data enthusiast myself, I can tell that it was one of the things that motivated me in subscribing to the paid version. It shows you how powerful big data can be, and at the same time how simple you can be with it.
In the past, it came in your email and now it is present as “stories” inside of your app, which makes it easier to post on other social media apps. They put together interesting statistics for each user such as the number of minutes they’ve listened to music, their favorite artists, podcasts and many other data points. This is very engaging because the direct reaction that most people have is sharing this with their friends, promoting the brand in an organic way.
Another cool example is Microsoft's “Anatomy of a breach.” This data story guides readers through a data heist to show how prevalent breaches are and viewers are encouraged to explore the data to draw their own conclusions, but the provider has full control of the flow that the user is taking. In the end, you are taking the journey that they’ve built for you.
LinkedIn's “Data and Insights” page is trying to promote data into the job seeking world. By presenting stats like “Most in-demand job” and “Most confident markets,” LinkedIn builds a story with their own data, gathered from their website and, as it has nowadays the status of a mainstream tool for job seeking, you can expect that the data is somewhat reliable.
As Plato said 2400 years ago: “Those who tell stories rule society.” Want to meet others like you who love data, and want to use it to make beautiful things? You need to join the Ironhack family!
Our Data Analytics Bootcamp empowers you with all the knowledge you need to launch your data career. Learn part time or full time, live online, or on campus. We've also got financing options to make investing in your career as easy as possible.
