How to Use Analytics to Find and Solve the Toughest Organizational Questions

The term analytics means different things to different people depending on the role and position in your business. If you are a CFO, it will mean something different than it does to a data scientist. If you are a financial analyst, it will mean something different than it does to a biostatistician.

I can remember in the late 1990s while implementing a financial consolidation and reporting solution in Vancouver, British Columbia. I was talking to the financial analysts of a major gas pipeline company. They talked about processing actual values, budgeted values, and a forecast that was updated monthly. Even though I have a financial background, because I was working for an analytic software vendor, I thought they were talking about analytical forecasting. What they meant was re-forecasting the rest of the year’s financial results based on the run rate of the most recent financial results.

Defining Analytics

Techopedia defines analytics as “the scientific process of discovering and communicating the meaningful patterns which can be found in data. It is concerned with turning raw data into insight for making better decisions. It is especially useful in areas which record a lot of data or information.”

While the definition gives an overview of how analytics is defined in the technology industry, it does not convey the same meaning to everyone who reads it.

• Does it mean calculating the average and standard deviation within the data and looking at the data points that are farthest away from the average as interesting?
• Does it mean looking at the last three-quarters of your financial data to see if you can accurately forecast the next quarter so that the shareholders are not surprised by the results? How would you do this? How much data is relevant in the forecasting process?

All of these exercises are involved in performing analytics at varying levels. Which technique you use will depend on what business decisions you are trying to make as well as the results needed to best support those decisions.

Descriptive Statistics – Identify the Tough Questions

You can look at what analytics means across different industries and to different roles in those industries. Let’s review an example in the higher education industry to demonstrate what I mean by this and establish a baseline understanding.

Suppose you want to analyze the Calculus 101 class within the math department to help make staffing and course time offering decisions. You decide to examine the last five years of class data. The following descriptive statistics instantly come to mind:

• What was the gender breakdown across the class?
• How were the enrollees distributed across the four major classes (freshmen, sophomores, etc.)?
• How many students enrolled in the 8:00 AM class? The 9:00 AM class?
• How many took the Tuesday/Thursday offering versus the Monday/Wednesday/Friday offering?

After you look at the results, you have a pretty good idea about the demographics at your institution and which days, times and professors are the most popular.

From those results, you can make some conclusions about the Calculus 101 class from the data:

• Analyze the trends in class size, gender makeup or grades across the five years.
• Calculate how many students withdrew from the class each semester.
• Review overall trends in class performance across various professors.
• Make decisions about the number of offerings in each time slot and across which days the class should be offered.

For those of you who look at this as analytics, don’t let anyone tell you that it isn’t. This is part of the analytic process. However, it is not the total picture. There is so much more to look at on the topic of analytics – like predictive analytics!

Predictive Analytics – Solve the Tough Questions

How can we gain further insight into the patterns and trends in our Calculus 101 class?

• Perhaps you want to calculate the probability that a student will withdraw from a class next semester, also known as the attrition rate?
• What if you want to compare attrition rates across class times?
• How about analyzing attrition patterns across professors or if you wanted to make inferences about the performance of students who sign up for the class with a given professor?

Can you answer these types of questions from the data and analysis listed above? The answer is “maybe” in some cases. However, to truly answer these questions, you would need to dive deeper into analytics to do so. This is where more complex techniques and data science usually come into the conversation. A word of advice – these more complex results and techniques are not readily available in the average spreadsheet package.

Analytics means different things to different people. It will take on varying levels of complexity depending on your role in the company, the industry in which you work, and the problems that you are trying to solve. The level of insight available to you will depend on your investigative curiosity and the inventory of analytic software and skillsets available in your organization.