Data visualization: It’s a strategy, not just a tool

By Ritika Puri, Contributor

It’s no secret that data is one of today’s most valuable business assets, and with technologies like social login and pixel-based targeting, companies can collect a seemingly infinite amount of it. The challenge, however, is that much of that information remains underutilized because most organizations don’t have the resources to process — let alone understand — it all.

Data visualizations present one solution. Using the principles of graphic design, marketing and IT leadership, for example, can provide innovative ways of representing data to spot previously invisible patterns in information or effectively communicate key trends to stakeholders.

While data visualization is far from a new concept, improvements in visualization technology and new ways of thinking about information have combined to raise the importance and impact of visualization for businesses with strong data stories to tell. Many companies are crafting visually powerful data narratives, and those that are doing it effectively wield visualization not just as a tool but also as a strategy.

Defining objectives

Like so many things in business, compelling data visualizations begin with clearly defined missions and goals — a point sometimes lost on technical leaders who, seduced by the graphical possibilities, may sacrifice impact for aesthetics.

“Often, those seeking visualizations are flattered more by flashy showmanship than by how the visualization draws true insight from data,” said Jeffrey Baumes, a data scientist at open-source software support company Kitware. Baumes encourages people to initially look past design and to focus instead on the core message the visualization is intended to communicate. It’s an “art,” he explained, to make the most interesting parts of a data story stand out.

Defining objectives can be challenging, especially because there can be many different ways to represent any one pattern. Data analysts and visualization experts will often go back and forth, making trade-offs to pinpoint the most appropriate and effective approach for the situation.

Colin Zima, chief analytics officer at Looker, agrees that those who create data visualizations should always begin with the objective — for instance, to emphasize the importance of a revenue or cost target. “Too often, charts or graphs are created to show some trend or a series of facts,” he said. “It should instead be a way to see the end goal.”

Baumes encourages creators of data visualizations to look for models as well as inspiration in historical and recent examples of great visual communication; he cites a map drawn by French civil engineer Charles Joseph Minard in 1869 of Napoleon’s march in the Russian campaign of 1812-13 as one of the more outstanding. It shows the diminishing size of Napoleon’s army as the campaign progresses.

“In this visualization, a map showing Napolean’s troop strength is connected to both geography and time, effectively conveying the disastrous effect of his march on Russia,” Baumes said.

Planning ahead                

Data analysts can make their visualizations most effective by fully assessing the business contexts in which they’re operating and understanding what results the presenter is trying to achieve.

“Before creating a dial on a dashboard that shows the inventory level of a warehouse, determine the decisions that this dial will help you make,” said Trevor Lohrbeer, co-founder and CEO of heat-mapping software vendor Lab Escape.

Lohrbeer encourages companies to develop a data collection and planning strategy and to give themselves time to collect and organize the information in order to generate a clear and concise set of insights. Presenters may not be able to achieve their objectives if they don’t have the right information to enable the decisions that need to be made, he said.

“If you don’t have the right data, find a way to collect it,” Lohrbeer said. “Consider the impact and timeliness of the decision … Match the investment in data preparation with the impact of the decision — make less of an investment in data preparation for small decisions and more for larger decisions.”

As Zima explained, visualization is the result of an extensive, purpose-driven process of data collection, analysis and reflection: “Good visualization allows business analysts to present analyses in a way that is accessible to anyone in an organization,” he said. “This crystallizes the intelligence that can be derived from information and can influence decision-making inside a business.”

About the Author: Power More