5 Keys to Powerful Data Visualizations

Data models can distill powerful insight from raw data. Yet, this insight is only valuable when it is acted on, and it’s only acted on if it’s understood. Visualization plays a vital role in revealing key aspects of data within its overall context, enabling understanding for technical and non-technical decision makers alike.

At Elder Research, we go beyond static graphs to build data visualization applications that facilitate data exploration, providing users with the ability to select many different ways to view their data. We enable the user to take a compelling business question to the data as a query, and we translate that into a view where they can explore the data’s details.

To create powerful visualizations, make sure they are:

The first priority is to ensure accuracy (which includes clarity).

The primary purpose of a visualization is to convey understanding, so a visualization that confuses or misleads will engender business mistakes.

No matter how attractive a great feature or trick appears, don’t use it if tests show it misleads your intended audience.

Each visualization must be easy to read. It should never take significant effort to decipher, but be intuitive to understand and navigate.

Visualizations should help focus attention on interesting attributes of the underlying data.

For example, to represent risk scores we use a color gradient from green (low risk) to red (high risk) as shown above in the map view from our Risk Assessment Data Repository (RADR) tool.

The map is intuitive as the location shows where the contract is, and the size of the dot indicates the dollar amount at risk.

Another way we help users focus on risk is to use a polar plot, placing the high-risk points at dead center.

Specific events are represented as icons, as shown on the right.

An interesting way to give life to visualization applications is to make them interactive.

For example, if a user finds a set of points in a scatter plot that look interesting, allow them to draw a box around those points so only those values are detailed in a data table (as demonstrated below). Or, if the user finds an interesting value in the data table, the tool could allow them to select and highlight it on the graph in order to visualize where it falls in the overall distribution.

This back and forth data exploration enables users to connect the visualizations to the data details, and vice versa.

Customization is a powerful visualization feature.

People are more comfortable using things that they have control over, and allowing users to customize their visualizations facilitates data exploration at a much deeper level.

Although it adds complexity, allowing users to determine the type of graph they want to use, which fields to group, or whether the y-axis displays a plain value or a cumulative score, can enhance user experience and increase adoption of the tool.

Customization may increase the learning curve for users, but it provides them with intuitive tools to explore data and discover unique insights.

The ultimate goal is for data insights to be implemented in business infrastructure, resulting in action.

For each specific scenario, aim to provide tools to make that transition as smooth as possible for the users.

Seemingly small features — such as the ability to easily save pictures of graphs, or export selected subsets of data into formatted excel reports — can go a long way.

Focusing on data accessibility and ease of use from start to finish will increase user satisfaction and adoption, resulting in increased value from data-driven insight.

These five properties are the foundation of powerful visualization tools that help analysts transform data into actionable insights.

How can you make your data visualizations more accurate, readable, interactive, custom, and accessible?