Three Ways AI and ML Can Make Payments Data Work for You

If data is the new oil, here’s how billers can refine that resource for the best revenue results.

In 2006, mathematician Clive Humby coined the phrase, “Data is the new oil,” and that couldn’t be truer today. Like oil, data is a valuable commodity that must be refined to make it useful. Data can also serve as an asset and investment that can put power and influence into a company’s hands, just as oil has done for more than a century.

But how can data be transformed from an overwhelming mass of statistics and percentages into actionable insights that can support real-time decision-making? The answer is twofold: artificial intelligence (AI) and machine learning (ML).

AI and ML work especially well in the payments industry because of the large troves of data collected each day related to customers’ payments history and behavior. If you work with a technically advanced payments provider, it can apply AI and ML algorithms to this data that identify trends and patterns in real-time, make logic-based predictions, and automate actions to avoid or proactively resolve common problems.

In this article, we’ll explore three ways your payments provider can use AI and ML to “refine” your data, turning it into a more valuable asset for your business:

1. Enhance customer support and drive self-service

AI and ML can be used to improve self-service adoption. For instance, an ML model can identify customers waiting in the call center and send them messages with personalized links to pay. A simple tap or click will take them directly to their payment screen to encourage self-pay. AI can also pinpoint reliable payers who are not using auto-pay and send them targeted engagement messaging with links for easy sign up.

For customers with a history of late payment, AI can automatically send a message suggesting self-service ways to reconcile, such as moving their payment date to coincide with their payday, or breaking payments into multiple payments throughout the month. This gives customers buy-in into how they want to proceed, which ensures compliance. And it empowers self-service rather than having those customers call a service agent to hear their options.

Very soon, your payments provider may also be able to create chatbots powered by machine-learning algorithms that assist customers who have standard questions or payment issues. This could provide fast and efficient service for customers and nudge them toward payment.

The benefit of AI is that it conducts data analysis with minimal human intervention—getting better over time through machine learning—and then automatically initiates the most logical interventions. This saves customer service centers from getting bogged down with low-level tasks so they can focus on more complex or specialized issues. At the same time, payors gain more autonomy through self-pay and have a more efficient and satisfying payment experience.

2. Forecast payment shifts for better planning and protection

Your payments provider can use AI techniques to track and compare large sets of payments data, flagging any unusual patterns or behaviors that could indicate potential fraud. If a series of unusual events indicates a potential threat, such as anomalies in the data of a disbursement recipient, AI tools can detect it in real-time and notify the payments provider so it can authenticate the transaction and increase surveillance and security measures.

AI also can be used to analyze cohorts of customers based on hundreds of data points—from third-party data sources to online activity—and assign them a risk score. Data from government sources, such as inflation, unemployment, and gross domestic product reports, can also help predict macroeconomic downturns that could impact payments. Based on these risk profiles, the payments provider can help businesses determine which customers are more likely to pay late or miss payments, and create outreach tools that alert those customers when a payment date is approaching or already passed.

AI is perfect for this task because it can analyze large amounts of data from multiple sources in real time and make logic-based predictions at a granular level. This can help billers prepare for changes in payor behavior that could impact revenue streams.

3. Automate fixes to common payment problems

AI and ML allow you to automate solutions to common payment problems, such as frequent automated clearing house returns, chargebacks, and NSFs. Your payments provider can develop ML models that initiate a change to business rules when a customer hits a particular milestone. For instance, once a customer has two or more ACH returns in less than six months, AI can apply a rule requiring the customer to pay with cash or card only.

AI also can be used to identify customers with a pattern of late payments and initiate automated engagement messaging. The first message might be a payment reminder via text message, email, digital wallet, or push notification. A second message could be a personalized email outlining the customer’s payment schedule, payment amount, and any additional fees or charges, so the customer is clear on the obligation. For the most chronic late-payors, AI might offer solutions like splitting payments, adjusting the payment due date, or initiating a conversation.

When AI and ML identifies and automatically fixes payment issues, companies can reduce the time staff members spend manually resolving payment problems. And, because the ML model continues to learn and apply its findings, it can identify the solutions that best accomplish your goal and automatically applies those to future responses. In other words, if one message, strategy, or business rule improves on-time payment the most, it learns the nature of that response and applies its findings to future decisions.

AI and ML make data useful to billers by taking large amounts of raw information and turning them into useful insights and practical applications. But the other necessary step toward refining data is data democratization, where payments providers and billers work together to put data in the hands of all stakeholders who can benefit.

Step one of data democratization starts with a payments provider committed to making payments data accessible to its clients. To do that, the payments provider should have a data warehouse where it compiles data in one place, scrubs it of protected personal information, and offers it to clients in raw form or in an array of helpful reporting dashboards.

The payments provider can also provide benchmark data in aggregate, including payments platform data, anonymized client data, and companies of comparable size, type, or regional location. This data can be compared side-by-side with your own data to reveal any outliers that may need improving—for instance, higher rates of ACH returns or lower adoption of autopay.

Finally, the right payments provider should be able to provide external data that offers additional insights. That might include data from credit bureaus that can help your company determine the risk profile of an individual or cohort, or identify fraudulent activity that may warrant extra caution.

Even weather-forecast data can prepare you for the impacts of a major storm, when customers might lose power or have to evacuate, impacting the timeliness of payment and the type of customer messages you distribute.

Partnering with the right payments provider makes all the difference. When that’s in place, billers can make that data available to everyone in their organization who can use it, from sales and marketing to finance and customer service.

Just like oil, data is most valuable when fully accessible and available to fuel the decisions your team makes every day. Those who achieve that goal will have a competitive edge in understanding and supporting their customers and achieving operational efficiencies both now and in the future.

—Roger Portela is the senior director of product at PayNearMe..