There’s been plenty of hype over artificial intelligence, no question. But there are highly practical ways that CFOs can use AI right now to bring new efficiencies to the enterprise. Here are five of them.
Customer data and predictive pricing
The CFO sits at the center of customer data flows: sales data, pricing information, receivables updates – the list goes on. This puts the CFO in a powerful position to link predictive analytics with customer behavior.
Here’s how it works. The controller’s team joins forces with data scientists to run correlations between price behavior and receivables. They overlay transaction data such as payment or customer demographic patterns.
They then add third-party data on weather and location at the point of purchase. The result: a projection of the optimal pricing for an age 18-to-24-year-old female repeat customer who is using credit cards to buy gifts for the holidays when the weather is poor.
Example: the holiday special. A major U.S. retailer combined point-of-sale, inventory, and pricing data with AI to create workflow efficiencies and drive dynamic, to-the-minute pricing in order to maximize sales during the busy holiday season.
Looking beyond book value
One of a CFO’s greatest challenges can be assessing the true value of assets. Uncertainty in valuation can add or subtract millions from the bottom line.
The most effective method for asset valuation is an examination of vast asset datasets. AI can help here.
CFOs in the real estate business are using AI to correlate thousands of housing variables – mortgage rates, quality of schools, number of bedrooms, local employment – to build models that predict home asset prices. That not only helps the real estate firm; it provides valuable information for buyers, sellers, and lenders.
Example: instant valuation model. A Dutch startup uses a model that leverages AI algorithms to instantly perform property appraisals. The CFO uses the model to estimate asset values at fair market prices for acquisitions and for formulating the company’s own balance sheet.
According to the U.S. tax authorities, bad debt equates to 0.5% of U.S. companies’ revenues. In 2018, that was more than $100 billion in missing money, reducing profit margins by as much as 5%.
A multivariate analysis of business-to-business customer data – such as industry type, credit rating, product purchases, and salesperson – can forecast the probability that a business will pay its bills and therefore should be extended credit. Identification of likely non-payers also helps in customer qualification and credit approvals.
Example: the rogue salesperson. A financial services firm was plagued with high mortgage defaults. The finance team built a predictive model of sales force compensation and found that the highest correlation was with loans approved by salespeople operating under aggressive sales incentives. Predictive analytics were used to understand the correlation between these compensation structures and bad debt, providing a long-term forecast of defaults under different scenarios.
Embezzlement and expense fraud
Internal fraud is particularly hard to detect. It is episodic and usually doesn’t leave a clear data trail. It’s often executed in small increments that escape detection. The perpetrator may intentionally distort the data trail.
Enter artificial intelligence, allowing the CFO to analyze and interpret expense data and detect suspicious expense claims. One can explore spending patterns and employee behaviors in various roles. Machine learning technology can identify and predict common behaviors of employees who falsify or exaggerate claims. This makes it possible for a CFO to forestall potential expense fraud before it happens.
Example: setting expense policy. A CFO’s team reviewed a historic set of proven expense abuse cases and captured data on expense type and vendors (for example, certain airlines or hotels). But rather than go after the employees, the finance function launched an aggressive program of education and monitoring of areas of abuse. AI allowed the team to target the training on the abuses that were most likely to happen, reducing future fraud in the workplace.
Detecting money laundering
Many banks have created alert mechanisms that predict money laundering based on known patterns of abuse, some of them mandated by law. The problem is that these systems can generate a flood of alerts and leave the CFO’s team wondering where to investigate.
Artificial intelligence can help programs recognize suspicious behavior and classify alerts as high, medium, or low risk. Applying rules to these alert classifications can automatically close false alerts, freeing up staff to focus on the most likely culprits.
Example: ranking alerts. A European bank faced the challenge of skyrocketing “false positives” in its anti-money-laundering system. The finance team created a predictive model for illicit transactions based on proven money-laundering transactions, using an algorithm that ranked alerts on a 1 to 10 scale. Scores above 8 opened an immediate investigation, while 10s triggered a freeze in assets. Regulators cited the bank’s predictive modeling as a standard for other banks in the fight against money laundering.
Artificial intelligence is fast moving from the experimental to the operational. Forward-looking CFOs are integrating AI into almost every part of the finance function, from receivables to treasury and from budgeting to pricing.
As they increase their powers of prediction, CFOs are finding that their roles are becoming more proactive, more strategic, and more valuable in the enterprise.
Want to find out how to build, integrate, deploy, and operate an intelligent application with SAP Data Intelligence? Join us on September 17 and get to know the capabilities of SAP Data Intelligence.
This article originally appeared on CFO.com and is republished by permission. EY is an SAP global partner.