As explored in our FORsights™ article, “The Importance of Data Literacy: Understanding Your Company’s Data,” the first step to improving data literacy is to gain understanding. In the second of our three-part series, we’ll focus on the next step, which is to question and validate that understanding through analytics.
At its core, analytics involves leveraging data and visualizations to examine information and provide informed answers to specific questions. Data scientists engage in extensive inquiry when reviewing the data for patterns, continuously refining their questions as new patterns and insights emerge. They seek to understand the underlying causes of observed phenomena, identify the factors influencing changes, consider related implications should one variable shift, and determine actionable steps to achieve specific objectives.1
Primary Types of Analytics
Descriptive Analytics
Descriptive analytics most commonly utilizes business data to answer the question, “What happened?”2 This could look like a mapping of payments to vendors over time, a chart of medical sales per region, or ages of new subscribers. Descriptive analytics shows the storyline of the business; what has happened and what is happening. It can bring to light questions that need to be answered regarding company data.
For example, Pulse™ for CHCs (community health centers) can help CHCs better understand financial and operational factors in their organization and how this compares to their peers. When businesses want to understand the cause of the descriptive history, they turn to diagnostic analytics for solutions.
Diagnostic Analytics
Diagnostic analytics seeks to answer the question, “Why did this happen?”3 For example, from your descriptive analytics, you may discover an unusual increase in a vendor’s payments over time. One possible explanation is the double payment of invoices. Our Payment Risk Analytics product looks for this pattern, among others, using the following approach.
First, all invoices paid by the business are laid out. Then, logic is applied to identify shared attributes (such as vendor name, invoice amount, and date) between invoice records. These similar records may indicate an invoice was paid more than once but was entered slightly differently. With this added layer of logic, the data has undergone a specific treatment in the search for potential duplication of invoices. The end-user can then inspect the results to determine whether the invoices were paid more than once. Recently, this analysis aided a client in identifying $213,000 in overpaid invoices.
Predictive Analytics
Predictive analytics uses learned factors and trends to draw logical conclusions on what may happen next if given a similar set of variables.4 This analysis relies heavily on accurate historical and current data and may utilize competitor and industry trends to fuel the process. Machine learning and artificial intelligence (AI) are large players in this type of analysis.
Another use of predictive analytics is the “what if” analysis, which is part of the Program Economic Analysis (PEA) product from Forvis Mazars. This form of scenario analysis enables decision makers to adjust key variables to forecast potential outcomes should certain factors occur.
Prescriptive Analytics
While predictive analytics identifies what is likely to happen given a set of scenarios, prescriptive analytics suggests an action to be taken based on the same information.5 We have used prescriptive analytics techniques in tools built for our clients to provide suggested routing for an outside sales force to maximize the time spent with customers or prospects who are likely to purchase the company’s products.
How Forvis Mazars Can Help
Products and services from Forvis Mazars often contain more than one type of analysis to provide helpful insights for clients. Analyzing data can provide valuable insights but is incomplete without the communication of the findings. This can be done through presentation, visualizations, and more. Look for information about effective communication of data findings in an upcoming article.
If you have any questions or need assistance, please reach out to a professional at Forvis Mazars.