By Chris Kosty, CPA, CIDA

All companies look for ways to gain an advantage over their competition. That advantage can come in many forms, such as top-of-the-line machinery, a new marketing plan, or a strategic hire. Interestingly, in the current age of technology and business, each of us are already in possession of a competitive advantage, we just may not know it. What I’m referring to is data and data analytics.

Data analytics is the discovery, interpretation, and communication of meaningful patterns in data. Management can then apply those patterns toward more effective and informed decision making. Data analytics can enhance oversight, create efficiencies, and optimize your processes to help differentiate your business from the competition. Companies across all industries have the ability to unleash their data to help them become more efficient, effective, and profitable. Data, however, is similar to oil: it’s a valuable resource, but not necessarily in its natural state. Data needs to be discovered, extracted, and refined to make meaningful use of it. Not to diminish the importance of the upfront structure, planning, and processes needed to make the data analytics environment a smooth operation, but, to me, the clear communication of findings is integral to using data effectively.

Have you ever performed some data analysis, provided a report, and then received a reply such as, “This looks great, but can you just tell me what is says?” Getting to results is a challenge in and of itself; communicating those results to others who are not as familiar with the data as you are and empowering them to answer those questions themselves is even more of a challenge. However, this should always be the goal. Data visualization is the most effective vehicle to deliver those insights.

Data visualization is the representation of information in the form of a chart, diagram, picture, and so on. My personal definition also includes “to gain insights and convey results.” Otherwise, we are just making a pretty picture-book with the same questions arising as laid out above. Interpreting results of an analysis using only a report filled to the brim with numbers typically requires a strong understanding of the data, with context, to understand any trends, outliers, or abnormalities. Graphically representing data normalizes the interpretation and allows the end user to arrive at conclusions themselves, efficiently and effectively.

I was recently tasked with performing an analysis of trends of cost types for four separate jobs over the course of a year. My intention was to identify and target-test abnormalities contained within that data set to understand the scope and timing of these cost types at micro and macro level. The underlying data used to complete the analysis was within a single Excel sheet, however it did fill over 64,000 rows and 22 columns. Each individual row represented a single entry-point of a certain cost, and using Pivot Tables, sorting, and filtering would have been time consuming and ultimately had me fall short of my initiative. Further, my allocated time to complete this analysis would have been primarily filled with data manipulation instead of data interpretation. I decided to use data visualization to perform my analysis, which resulted in the picture below.

Almost immediately, we can see abnormal trends on both an individual job level as well as comparatively between the entire portfolio of work. Each color represents a different cost type, which allows us to also evaluate the allocation of cost types at the same time. As intended, and a natural byproduct of data visualization, when we hand off the results of this analysis the end user can understand where the anomalies reside in the data set for themselves without having to regurgitate the analysis and talk through the numbers. Importantly, the numbers are always there to investigate the abnormalities, but now we have a targeted approach to reviewing the raw data as opposed to looking for a needle in the haystack.

When starting on your data visualization endeavors, it’s important to have the end in mind. The type of analysis you are performing will typically steer what type of chart or graph to use. Are you showing a trend over time? Use a line graph. Trying to evaluate the composition of a data set? Use a pie chart.

Another strategy I deploy is being mindful of my chart titles. Simply restating the x-axis and y-axis in your title provides little additional value. For the chart above, titling it “Cost Types Over Time” is insufficient. However, titling it “Is Our Company Experiencing Abnormal Cost Accumulation?” invokes a sense of investigation and allows the user to read and interpret the chart to answer that question themselves.

I've often heard the saying, “Data doesn’t have an agenda, only people do.” Keep this in mind as you work through your analytics. Don’t let personal biases tell the story or answer the question: the data will do it for you.

Christopher T. Kosty, CPA, CIDA, is a member of the Automation and Data Analytics Process Team (ADAPT) of Schneider Downs & Co. Inc. in Pittsburgh. He can be reached at ckosty@schneiderdowns.com.

Sign up for weekly professional and technical updates in PICPA's blogs, podcasts, and discussion board topics by completing this form.