Before you start creating your data visualization, you need to have a clear idea of what you want to achieve and what you want to convey. What is the main point or takeaway you want your audience to remember? What is the best way to support your message with data and evidence? How do you want your audience to feel or act after seeing your data visualization? These questions will help you define your purpose and message, which will guide your design choices and storytelling techniques.
Data visualization should be presented in such a way that the audience should be able to decode in the first five seconds. It should not be clumsy or overcrowded but contain a straightforward message without ambiguity. A simple message drives home the point when there is no distraction.
The next step is to understand who your audience is and what context they are in. Who are they? What are their backgrounds, interests, and goals? How familiar are they with your topic and data? How much time and attention do they have? How will they access and interact with your data visualization? These questions will help you know your audience and context, which will influence your data format, level of detail, tone, and interactivity.
Know your audience and context' is a vital part when it comes to how we are going to represent our data. The functionalities and interactivity of the dashboards will be purely based on the audience. - You can define the type of charts/tables in the dashboard based on the audience and their suggestions. - Functionalities like data downloading, types of parameters and filters, formatting, etc., can be provided according to the end user's choice. We are actually providing a tailored product to our audience to help them analyze and utilize the data in the best possible way. Therefore, ensuring alignment between the requirements of our audience and the end product should be a top priority.
Creating an audience segmentation is a good idea to know about your audience better. This will help you to choose the appropriate format for representing a data set. Let's say you have developed an app which assesses various health parameters of an individual and generates a report for overall health status. Your objective is to promote this app in terms of specificity and sensitivity. Now, you want to create an ad copy or a promo integrating the data analytics of this app. The choice of data representation formats will be guided by the type of audience. In this case you can decide depending on whether your audience is general public, key industry stakeholders with whom you want to partner, academic collaborators or a mixed audience.
Once you have defined your purpose and message and understood your audience and context, you can choose the most appropriate data visualization type and style for your project. There are many types of data visualizations, such as charts, graphs, maps, tables, dashboards, and infographics, each with its own advantages and limitations. You need to choose the type that best fits your data and message, as well as the style that best suits your audience and context. For example, if you want to show trends over time, you might use a line chart or a bar chart, but if you want to show geographic patterns, you might use a map or a heat map. Similarly, if you want to appeal to a professional audience, you might use a minimalist and elegant style, but if you want to appeal to a casual audience, you might use a colorful and playful style.
When choosing your data visualization, the type of visualization of the data does matter. You wouldn't use a pie chart to show trending data or do a line chart to show % of volume as an example. So choosing your visualization style isn't just about what you are comfortable, it also matters to choose what displays that data in a way that creates the story without you having to lay out additional context to that story. Think of data charts as the same purpose as having pictures in a picture book - if the kid can't read, they still have a concept and understanding of the story thanks to the pictures provided. Not everyone can read and speak data.
The final step is to test and refine your data visualization to make sure it is clear, accurate, and effective. You can test your data visualization by asking for feedback from your intended audience or from experts in your field. You can also test it by using tools and methods such as data quality checks, usability tests, eye-tracking studies, and A/B tests. You can refine your data visualization by making adjustments based on the feedback and test results. You can also refine it by applying best practices and principles of data visualization, such as choosing the right colors, fonts, labels, legends, titles, captions, and annotations.
One of the ways to present the message clearly is taking a minimalist approach. For eg if you have data labels, get rid of minor axis or even consider if you need the vertical axis. If there are multiple data points in a line chart, add data labels only when there are changes to trends. And don’t shy away from adding message boxes with arrows to explain the trends. In a time series line chart if there are more metrics to compare than the number of time periods, consider spreading them out by elongating the x axis and having only one line chart at a time to avoid any overlapping them. Generally stacked chart templates do not allow totals above them, you can add them by adding a line chart on a secondary vertical axis.