Data visualization and data storytelling are crucial aspects of being able to fully understand your data. If you’re in a role working with data, it’s important to recognize the steps that it takes to build and tell one.
According to Tableau, data visualization is a graphical representation of information and data. Without a data visualization, data is usually displayed in a spreadsheet or list. Visualization takes data analysis a step further, making it easier to read and understand.
Data storytelling isn’t just data visualization or analytics in a dashboard – it’s the combination of human connection and hard data. Data storytelling is taking data visualization a step further and communicating insights. The audience needs context behind data visualization and being able to communicate that is data storytelling.
Data storytelling is an incredibly effective way of communicating data to an audience. It’s been proven that humans are better at interpreting visuals than they are at spreadsheets or numbers. The human brain processes images 60,000 times faster than text. A study performed by researchers from the University of Minnesota in 1986 found that presentations using visual aids were found to be 43% more persuasive than unaided presentations. Data storytelling, with the aid of data visualization, is a great way to explain, analyze, and interpret data.
Before building your Visualization
Planning your visualization to aid your storytelling is important. So, where do you start? What do you want to answer? Before you start creating your visualization, ask yours “what question(s) am I trying to answer with this visualization?” Often, stakeholders have a vision of what they’d like to see with little knowledge of whether the data actually exists and is accessible. If the data isn’t available internally, it’s okay to leverage public data that lives outside of the organization.
There are several preconceived notions that people have of data storytelling. It’s easy to get bogged down in the bigger picture or prioritize certain aspects of visualization over another. To have a more holistic understanding, there are certain things that will hinder you from effective data storytelling:
- Assuming data fluency
- Believing it must be aesthetically pleasing
- Falling victim to the “fun facts factory”
Assuming Data Fluency
Data fluency is the shared “language” related to the data. Data fluency connects people across roles through a shared set of standards, processes, tools, and terms. Those who work closely with data like Data Scientists and Data Analysts often incorrectly assume a higher level of data fluency than is true, especially when working with tenured audiences.
To combat this, take a step back and honestly consider your audience. How educated are they around data? Airbnb manages this assumption by offering a “data university” to all employees, not just those in data or technology departments.
Believing Visualizations Must Be Aesthetically Pleasing
People tend to associate complex visualizations as high-quality visualizations and disregard functionality and readability. In business, functionality reigns king. Does the visualization show what it’s supposed to in a simple way? Can the viewer quickly and easily interpret what’s being displayed? Is it working as expected? Once functional, making the visualization aesthetically pleasing takes extra time and effort, which may decrease the ROI. Deeply consider how important aesthetics are to your audience.
Falling Victim to the “Fun Facts Factory”
Beware of the “fun facts factory” – people love the “fun facts factory.” Analysts enter the “fun facts factory” when they start generating data “fun facts” for stakeholders who then don’t take any action on the information they asked to see. The “fun facts factory” exists in every industry, company, department, and team – data visualization is trendy right now and everyone wants to be involved.
For example, an HR leader might request a visualization of employee retention rates. And assuming that employee retention rate is probably important to that HR leader, you do the work. And guess what? That HR leader never takes any action on the information you provided. Stakeholders shouldn’t ask for visualizations just to know a number or a “fun fact.” The stakeholder should always commit to an action tied before you seek out the data and make the visualization, otherwise, it’s everyone’s time wasted.
A great first question is “if I tell you or if I show you this visual, what will you do with it?”
While Creating Your Visualization
You constantly be thinking about how you’re going to be communicating this data to your audience while you’re building your visualization. Keep it as simple as possible and don’t include visuals that don’t add any value. Ask yourself: “if I exclude this visual from the dashboard, can I still arrive at the same conclusion?” This ensures you’re simplifying your visualization and making it easier for your audience to understand.
Follow an Iterative Process
Visualization development is an iterative process. Keeping regular communication with key stakeholders can ensure that you’re building the most functional, effective visualization, and you can tailor the story to them accordingly. Feedback from stakeholders should always result in better functionality or readability – not changes that don’t add value, like font and color changes.
While keeping your audience in mind, you also want to use your dashboard space effectively. Tableau says to keep your most important numbers, your KPIs, out of a chart. They’re often referred to as BANs, or Big A** Numbers. These are the numbers that are the most important that your audience should see right away. Keep them visible and make them big! Tableau did an eye-tracking study to prove that dashboards with large numbers showed a concentration of visual attention directly on those numbers.
Charts should be picked carefully and thoughtfully. The most important but simplest chart is the bar chart. Bar charts are generally a safe bet because they clearly convey information and are easily understood by all. Pie charts used to be very common but can be difficult to read especially when you have similar numbers next to one another or similar size “slices”. A better alternative to the pie chart is the donut chart. This is a pie chart with a text ”hole” in the middle that adds some additional context.
At the end of the day, whatever chart you decide on, make sure it is easily understood by the audience. A simple, functional chart type beats a fancy, difficult to interpret chart any day.
Another common misconception about data visualization is that everything and every chart has to be colored. Be selective with color. Color can help differentiate numbers and make things pop – but you don’t want to overload the audience’s brain with too many colors. Typically, one or two colors is recommended.
A good, basic template uses a gray gradient (white to black) with the addition of red or orange to make things pop (i.e. showing negative numbers in red).
Do a walk-through. Carefully go through all the features, filters, and functionality with a critical eye. Then present your visualization to a neutral, third party who has limited knowledge of the data. When you’ve spent so much time with the data developing the visualization, the curse of knowledge creeps in. You comprehend what’s going on in your visualization, and you incorrectly your audience does too. This presentation ensures your story is engaging and helps you prepare for possible questions from stakeholders.
Sometimes, you’ll have to deliver bad or unexpected results to stakeholders. It’s important to be cognizant and careful about how you deliver the message. Be prepared to answer a lot of questions about your data and visualization and try to deliver any bad news in the most direct but constructive way possible.
- Data visualizations are a powerful way to tell a story and have a lasting impact on your audience
- Humans process and retain visual information much faster and better than text-based information
- Assume your audience knows nothing about data or data concepts
- Beware of the “Fun Facts Factory”
- Data visualizations should be functional first, aesthetically pleasing second
- Tailor your story to your audience the best you can
Need help getting started with data storytelling? Reach out to one of our consultants today.