What Happened to My Spreadsheet?
Data visualization. It's a phrase heard more and more within the analytical community to describe a new, radical way to understand and draw insights from the ever-growing mess of structured and unstructured data. But there's a problem here. Data visualization isn't new. It isn't necessarily radical. And, if you can't organize your data in a meaningful way, it's very difficult to gain any real insights. So why is everyone interested in data visualization?
The short answer is because it works. Like many analytical techniques, there are increasing levels of benefits that can be gained through more advanced application of data visualization. This article will examine how data visualization is being used today for reporting and analysis and look forward to what may be on the horizon for this field. You may recognize your company's usage of visualization and will hopefully discover some new ways you can change the way you look at your data.
Data Visualization for Reporting
The most common place for businesses to start using visualization techniques is through data reporting. Elementary school students learn the fundamentals of line, bar, and pie charts for summarizing numbers and often these same charts make up the bulk of a company's data visualization portfolio. Whether reporting financial metrics, project performance, or even operational goal achievement, simple charts can provide instant information in a clear and easily understood manner. Green numbers are good - red numbers are bad. An upward sloping line indicates growth while a downward sloping line makes everyone nervous. We all have seen and used these charts. But that's just the beginning.
Data visualization provides context for your data, prompting you where to dig deeper, highlighting outliers, or reinforcing assumptions
Reporting is a difficult word for many people. While "reporting" has the connotation of static, crowded tables of numbers that cross your eyes and obscure information, it can also be dynamic, insightful, and purposeful. Consider a simple dashboard, built in your favorite visual BI platform, updated regularly without manual effort that displays your company's 10 most important KPI metrics in a way that explains current performance within a matter of seconds. Who wouldn't want that?
Effective dashboards use interactive filters, drill through capabilities, and engaging visuals to create dynamic, self-serve access to information. Much like your car's dashboard, these visualizations serve as indicators to the status and health of your business, focusing your attention on where the problems are, not on determining the importance and context of each metric. These types of dashboards do have limitations, however. They generally concentrate on high-level, aggregated data and provide little in the way of actionable insights. But that's ok. That's what they were built for. That's why they must be accompanied by data visualizations for analysis.
Data Visualization for Analysis
If reporting is where most businesses start, analysis is where they must go next. Without understanding the drivers of your KPI metrics, you won't know what business levers to pull to drive change. Analytical statistics can be broken down into three major categories: descriptive (what happened), predictive (what will happen), and prescriptive (what should we do to make things happen). Visual BI platforms have been very good at the descriptive side of analysis for a long time. Twenty years ago, we were using spreadsheet programs to create those line, bar, and pie charts we all know and love to look at distributions and historical trends. As computing power and data available has grown, so has our application of data visualization. We are now able to use animation, multi-dimensional views, and augmented reality to better understand our data. We apply these techniques to sophisticated predictive models, machine learning algorithms, and unstructured data sets to forecast where our business is going and how soon we will get there. The advances are not necessarily with the visualization techniques, although modern visual BI platforms help, but rather with the underlying data we apply them to. The old saying "garbage in, garbage out" absolutely applies on the flip side—"treasure in, treasure out."
Reporting and analysis must complement each other. How many times have you heard "we can't show that number. If they ask questions about it, we won't have an answer."? Given the massive amounts of data most of us deal with every day, there is no easy way to understand it without help. Data visualization provides context for your data, prompting you where to dig deeper, highlighting outliers, or reinforcing assumptions. Questions will still come, but you will be prepared to answer them.
The Future of Data Visualization
This is a journey we are all on. It's going to be a fun ride.