My Process for Building Data Visualizations

Story by Data offers weekly challenges in the area of data. I am excited to find out about the challenges and participated for the first time this week.

Weekly challenge: Describe your process or work-flow for creating data visualizations / dashboards

Story by Data

My process of creating data visualizations is still evolving and iterating, however there are common threads in the development process. The user should guide every step: research, exploratory analysis, identifying trends, understanding the story, designing for the audience, and iterating on the final product. 


The research phase involves asking questions such as: What does the client need? What is the purpose? What do they want to know? What are the expectations for the final deliverable? These questions identify the requirements for the visualization. Research on the dataset(s) is all about building context and includes: reading reports, studies, and documentation regarding the data, defining key metrics, understanding the data collection process, and reading about the research trends for the topic.

Exploratory Analysis, Identifying Trends

At this point the context around the data and user needs should be defined. Next, I go through exploratory analysis and start building many, many graphs. Building the many visualizations is exploratory because I ask myself — What is significant? What is surprising? What trends are visible? What questions arise? Questions that come up after the analysis bring about further research, which will hopefully lead to answers. At this point the trends are established.

Understanding the Story, Design for Your Audience, and Iterate

Understanding the story and designing for the audience often go together during my viz development. I ask myself What does my user need? This helps me identify which graphs I will use because the visualization should match the need and purpose. Does the viz show the most important metrics/top trends? Is there context provided about the metrics? Answering these questions helps shape the story and shape the communication on the findings. Next, I sketch the dashboard on a grid. I move around the various elements of the dashboard on this grid and repeat this process. Once I feel comfortable with the sketch I begin to build in Tableau. Ideally, during this phase, the user will provide feedback through the development process. Feedback is important and allows me to iterate the visualization/dashboard.


This process of development is always iterating and I love to read about how others create visualization.  1) The client guides every step of development. (2) Research to understand user needs and build context for dataset(s). (3) Exploratory analysis to identify top metrics. (4) Understand the story and design for the audience. (5) Iterate on the final product.