This content is part of the Essential Guide: Guide to telling stories with data: How to share analytics insights

Compelling data story starts with effective visualization of data

Data visualizations can transform analytics data into actionable business information. But remember to keep things easy to understand and remember your audience, an expert cautions.

Data visualization software empowers data analysts and business users to develop charts, graphics and presentations that convey insightful information about business performance and opportunities. An emerging practice called data storytelling has evolved around the use of visualization tools to translate the results of analytics applications so they can be more easily understood by corporate executives and other decision makers in organizations.

But you have to be careful when attempting to craft a story based on data visualizations, especially when you've been steeped in the analysis work for so long that you become blinded by the familiarity of the content. To ensure that you're delivering the intended message to the appropriate audience, you might want to begin by asking some questions about the objectives of a data story and the process for disseminating it within your organization, for example:

Are you trying to openly convey a particular message, or are you looking to let the data speak for itself? In some cases, a data visualization is simple enough that it can be unambiguously interpreted by anyone who looks at it. In others, the complex integration of data and multiple visualizations is needed to drive home the point of analytical findings. Sometimes the members of the audience for such visualizations are savvy enough to interpret the data on their own. But if they aren't, it's up to you to provide an easy-to-understand narrative to accompany the visualizations -- one that doesn't get bogged down in talk about the analytics process that detracts from the message of how the results should influence business actions.

Are you able to directly engage the audience, or will the information be delivered through a proxy? If you're presenting the narrative for a set of data visualizations to a group of business execs in a meeting room, you have some freedom and flexibility in how to structure the presentation that you might not have if it's being handled in some other way. For example, if a data story is being presented in a more structured format -- such as a video that will be sent to various execs for individual viewing -- you might have to adjust the visualizations and the accompanying narrative so they more clearly convey the intended message.

What kinds of visualizations will help support -- and not overwhelm -- the points you want to make about the analytics results at hand? There's a danger of getting mesmerized by the potential flashiness of data visualization methods to the point that a presentation becomes littered with beautiful but meaningless graphics -- decorations, essentially. You need to think upfront about designing relevant and practical visualizations, not works of art with no real informational value.

To avoid 'over-decorating' data visualizations, effective organization and graphical consistency are both called for.

Those questions suggest that there's more to data storytelling than just creating a pretty picture. When shaping your data visualization plans, to ensure that a data story presentation has the desired effect you should also consider including content density, abstraction and dimensionality.

An individual graphic might integrate multiple analytics results. A simple example is overlaying multiple line graphs to suggest a correlation between their values. But the denser the data being presented, the more thought must be invested in the design process to ensure that the audience won't be confused by a chart that hides the intended message.

Our familiarity with simple graphics, such as bar, line and pie charts, allows us to immediately interpret the data in them -- if they're well designed. On the other hand, layering multiple dimensions of analytics data within the same visualization lets the data storyteller share more complex and deeper information with business decision makers. To avoid "overdecorating" such visualizations, effective organization and graphical consistency are both called for. And the data story narrative, whether it's spoken or written, should be crafted to complement the visualizations in a clear and precise manner.

Ultimately, the effectiveness of data visualizations and data stories needs to be measured on practical grounds. Such assessments could start with examining whether business executives and others on the receiving end accepted the information presented in visualizations as believable and trustworthy, and whether they were able to quickly grasp and absorb it. Another good question to ask is whether audience members were motivated to share the information with others in your organization. In the end, of course, the biggest question to answer is whether they were persuaded to take action as a result of seeing the data.

Effective visualization of data and data storytelling can be a compelling way to transform analytics data into valuable business information -- but misuse of visualization and storytelling techniques can lead to confusion and inaction among execs who are counting on BI and analytics applications to help guide their business decisions. To give them what they're looking for, consider how the data visualization medium can best be adapted to support the message that the data you're working with is trying to send.

About the author:

David Loshin is president of Knowledge Integrity Inc., a consulting and development services company that works with clients on big data, business intelligence and data management projects. Email him at [email protected].

Email us at [email protected] and follow us on Twitter: @BizAnalyticsTT.

Next Steps

More on the need for simplicity in data visualization

Data storytelling is in the spotlight, but are people actually doing it?

Open source data virtualization tools starting to catch on

Dig Deeper on Data visualization software