Organizations are collecting data about consumers, businesses, and markets regularly. Every time I ask my Alexa Dot to turn on lights or order paper, data is collected about my power usage and product choices. These data can be used by organizations to provide tailored customer service or products that meet my needs. However, collecting data is only useful if it can be analyzed and transformed into information used by decision-makers.
How are these decision-makers transforming and using data? In accounting, it can be said that "Accounting is Big Data" – in fact, the American Accounting Association annually hosts its Accounting Is Big Data Conference with a goal of connecting professionals engaged in incorporating data into the decision-making process.
One use of data analytics is to streamline the process of testing internal controls and compliance. For example, internal audit groups can use a full population of employee reimbursement expenses and using pre-determined algorithms in analytics software or regression analyses to identify outliers in employee reimbursements or identify commonly used vendors. A company's external auditors may use data analytics to build efficiencies into their audit planning and testing.
For example, external auditors may be able to examine full populations of their client's transactions to identify unusual journal entries. Alternatively, external auditors may be able to use analytics to test transaction-based internal controls.
When I teach data analytics to accounting students at TCU, we focus on four elements:
Asking data-driven questions; extracting, transforming and loading data into analytics software; analyzing data using software and statistical tools; and communicating results.
What is the problem? Before engaging in data analytics, one needs to understand what is the problem that needs to be solved. Often, this means we need to consider if the problem should be solved using data analysis, or if there are other qualitative aspects that must be considered to get a full picture of the solution. It also requires that we think about segregating business dilemmas into smaller "bite-size" questions.
Data are messy! Before we analyze data, we have to be confident that the analysis will result in a solution that is accurate and complete. Sometimes this means that we need to verify that we have access to all the data and that the data used in the analysis is accurate.
Not every data point necessary for analysis is structured. Some data are in the form of video or audio files. Not all data is complete – often fields of records can be missing or in a format not appropriate for analyzing. Therefore, it is important to have the tools to examine data for completeness and accuracy.
Spinning data – this is the fun part. Examining large data sets for patterns and relationships provides an opportunity to answer those data-driven questions.
A variety of software can be used to analyze data – and it does not have to be complex software. Software programs such as Excel, Tableau, IDEA, Qlik, IBM SPSS, and SAS are a few software packages that provide the tools to visualize patterns and statistically analyze relationships.
Finally, communication is ultimately the most important element of the data analytics puzzle. I teach my students that if they can't communicate the first three steps to decision-makers, then the analysis is less powerful.
Renee Olvera, CPA, PhD is an associate professor of professional practice at Texas Christian University. She's writing this column for the Fort Worth Chapter of the Texas Society of CPAs, a regular contributor to FW Inc.