Boost Your Bottom Line with Predictive Analytics

Wouldn’t life be great if we all had a crystal ball? While wizardry and fantasy may sound like a great way to see the future, no one would actually advise a client or employer on assumptions based in magic and hearsay.

Typically, we serve our clients and employers based on fact-based historical reporting, which tells us where we are and where we’ve been. However, there is another way to offer guidance and opinion: predictive analytics, a process-driven activity that combines facts about the past with inferences to anticipate the future.

What does this have to with accounting? Plenty. As a CPA, you are typically detail-oriented, trained to prove your assertions and document your work, strong in certain technical areas and knowledgeable in multiple aspects of business. Most importantly, you are the trusted adviser your stakeholders depend on to help guide them to bottom-line solutions.

Wouldn’t it be beneficial to know the answers to these questions?

• How much budget will be left in three months given that my business units typically underestimate Q3 spending by 10 percent?
• How much is portfolio volatility likely to decrease next month, given that volatility is at a historically high level and typically reverts to average levels?
• How will sales revenue of a core product change based on the combination of a recent slowdown with increased spending on marketing?

The simplest predictive model asserts that what happened in the past will continue to happen in the future—and even a simple model of this type can create powerful explanatory value. For example, electrical consumption, retail spending and the commute time between two addresses in a city follow consistent cycles. As a result, any change in these variables can be explained by the time of year (electrical consumption and retail spending) and commute time (time of day/week). Additional predictive or independent variables can be layered into a model to increase its stability and power, while specialized modeling techniques can be used to extract more value.

As you use predictive analytics to carry out your project, there are four steps to consider:

1. Once you have the business case, figure out how the case justifies the value of your project. Your understanding and knowledge of various factors that drive fundamental business performance helps create, and identify, new opportunities for the future.
2. Surround yourself with a team of people who have similar skillsets and possess a variety of perspectives that allow for growth and rational reasoning to deliver the best results.
3. Decide if the data you have is sufficient enough to support your project and which data is worth using, disposing of data that may cause some kind of doubt in your theory.
4. Predict the future based on your results, and be willing to make necessary revisions to perfect your model.

Creating the model is just one step; putting it into action is just as important. Don’t be afraid to suggest predictive analytics as a way to look at the future. Your clients or employer will appreciate your willingness to step outside of the norm and tackle something new or different.

For more information on predictive analytics, download and read “Why Predictive Analytics Should Be 'A CPA Thing',” a new white paper by the IMTA’s Business Intelligence Task Force.

Adam Haverson, CPA.CITP, CFA, Senior Manager, CapTech Consulting. Adam is an enterprise-focused management consultant specialized in advanced analytics. 

Related articles

Managing the Risk of Emerging Technologies

Bitcoin: Mystery Currency or Real Money?

How You Can Manage the Risks of a System Implementation

Does Your Organization's Information Architecture Support Effective Reporting?

Governing and Managing IT Investment and Spending