Making faster decisions with AI

Like many midsize, privately owned businesses, Apex Parks Group, which owns 16 amusement and water parks and family entertainment centres across the US, generates an abundance of financial and operational data that readily tell how the company is doing. What's more difficult to figure out is why the company is performing the way it is.

Data are generated by each park but also outside the business; for example, by social media. When all the data are combined and visualised in a user-friendly way and when artificial intelligence (AI) such as machine learning and predictive analytics is leveraged to provide insights that are forward-looking, Apex Parks senior management can make better decisions faster.

Along the way, however, Apex Parks faces common challenges: Financial and operational data are kept in different systems. Operational drivers aren't mapped to financial metrics. Many processes must be performed manually, and data are kept in silos.

To be able to identify what drives performance more effectively and more quickly and improve forecasting reliability, Richard Fox, CPA, Apex Parks' vice-president of data science and analytics, is working on an AI pilot project with AlignAlytics, an international data analytics consultancy.

A data-driven, analytical approach to finance management relies on AI and machine learning to answer three key questions: How is the company performing? Why is it performing the way it is? And what should be done to make it perform better?

"Reports and spreadsheets may tell you the first, how are we performing, but it's analytics that help you understand the why, the root cause of what's going on in your company," Fox said.

Knowing the "why" allows you to decide quickly and effectively on how to improve performance, he added. The "why" gets you to the root cause of what is driving performance.

Analysing customer data with AI and machine-learning tools connects operational drivers with financial metrics to help the company target its marketing to a one-on-one level, fine-tune products and services, or pursue innovation. For this to work, however, it's important that the data are modelled and enhanced correctly, Fox said.

In a hypothetical example, he visualised the results of a business that sells several brands with higher and lower selling prices in multiple markets. Status and variance analyses comparing the different brands by selling price and revenue showed which brands dominated in which market. Applying a machine-learning algorithm went a step further. It showed why actual revenue was below plan: Customers shifted to brands with a lower selling price, which generated less revenue.

Once it works properly, Fox said, "it's embedding machine learning underneath the hood that then delivers the result in a dashboard for improved decision-making".

Senior leadership and the board of directors can use the dashboards to plan different scenarios, model them, and make smart bets on how to improve company performance.

Some companies go a step further, Fox said. Amazon sells books, shoes, and sweaters. Netflix is a video streaming service. And Airbnb offers holiday rentals. These companies have grown not only because of their products and services but also because they've made analytics their strategy, Fox said.

"They are changing prices in real time, based on activity on their website," he said. "How many people are looking at a product? How many people put it in their cart and didn't purchase it? How many were purchased in the last 24 hours or the last seven days? Those algorithms are running in real time and adjusting the prices."

As more data become available, from social media, the internet of things, and other sources, decision-making will speed up further and increasingly rely on real-time, streaming data, he added.

"We finally have enough data to be able to use algorithms like neural networks, and the processing power in the cloud is making it much easier to run these algorithms on large data sets," Fox said. "That's why AI is now the new buzzword."

4 objectives of financial management analytics

AI and machine learning can be used as tools in financial management analytics to interpret more and more data, make data analytics repeatable, and provide insights that help address strategic questions.

To stay on track, Fox and Roland Mosimann, president of AlignAlytics, suggested following these objectives:

• Design the data analytics and interpretation to understand why the organisation performs the way it does. Monitor high-level key performance indicators. Analyse leading, not lagging, indicators, and drill to detail. Use machine learning to look at the root causes of performance.
• Focus more on strategic questions and performance management, such as financial budgeting and planning, than on reporting.
• Deepen and broaden your line of sight. Combine internal and external marketing, sales, and operational data to determine what drives profitability, revenue, expenses, and balance sheet performance.
• Drive value creation. Make sure the business is running efficiently. Develop future business through a growth and innovation portfolio. Establish and manage resources and capabilities that are most critical for sustained high performance.

Richard N. Williams is a freelance writer based in the UK. To comment on this article or to suggest an idea for another article, contact Sabine Vollmer, an FM magazine senior editor, at Sabine.Vollmer@aicpa-cima.com.