Three Ways to Use Automation and Machine Learning to Resolve Disputes and Increase Cash Flow

Nicole Dwyer
Three Ways to Use Automation and ML

When an invoice or purchase order is in dispute, it puts the payment at risk. With the payment at risk, several factors are now potentially impacted:

  • How do you forecast cash flow?
  • How much time will it take to resolve this dispute? What will the impact be on your DSO?
  • Which resources (AR team, collections, legal or procurement) will spend time - and therefore money - resolving this dispute?
  • How will this impact the customer relationship and overall customer satisfaction?

There are a wide range of disputes from a simple inquiry or question, possibly needing more supporting documentation or information, to a disagreement with the invoice (pricing, terms, quantity, damage, returns and much more). It is also common for the dispute to be simply around not receiving the invoice. In all cases, disputes take time, money and effort to resolve.

Managing disputes takes up about 30% of the entire credit-to-cash process. That’s 30% of your time focused on reclaiming what’s at risk. Therefore, the faster a dispute can be resolved to the satisfaction of all parties, the more beneficial it is - to all parties. This is where leveraging a smart platform that offers automation and machine learning to support your processes can have significant and measurable benefits, such as a predictable, accurate revenue forecast, higher cash flow, lower DSO and increased customer satisfaction scores.

There are three key areas where automation and machine learning can help you drive cash flow and reduce the cost of managing disputes. 

Dispute Prioritization

When you are working with dozens, or hundreds, of disputes, it can be hard to determine which ones should have your immediate attention. All disputes require resolution, but knowing how to prioritize the work helps your teams work more effectively. They are quickly addressing the most important business needs and at the same addressing the customers with the highest level of stress. 

A smart AR system can help. A machine learning algorithm can look at the body of a dispute and evaluate the following factors to assign it a priority level:

  • What is being disputed?
  • What is the severity of the dispute?
  • What is the value of a customer to a company (based on the amount of business and open balance of a customer)?
  • What is a credit score of a customer?

Using all those factors to understand how urgently a dispute requires attention gives your AR or collections team criteria to make smart decisions.  The highest value and highest risk items are flagged for their attention, and your team is able to use both judgment and facts to determine whether lower level disputes are even worth going through a resolution process. 

Dispute Classification

Using available data to automatically classify a dispute helps assess the actual risk to the revenue involved, as well as the amount of time or resources that may be required to reach resolution.  The most common dispute claims are: 

  • Never received invoice 

  • Invoices don’t match or don’t reconcile to buyer records
  • PO number is missing
  • Other supporting documentation is missing - Proof of Delivery, W9 etc.
  • Missing tax information
  • Problem with the product or service (quality, shipping etc.)

Knowing this information and being able to filter that on the user interface allows the AR or collections team to put processes in place that can expediently manage the resolution. It is immediately evident how much time might be spent on quick, simple tasks such as resending copies of invoices or verifying a PO number, versus those that might require research such as reconciling records or understanding a service issue. Schedules and processes can be organized for the most effective and positive resolution.

Dispute Auto-resolution

Now that we understand the types of disputes that are open, we can use machine learning to automatically resolve them. This would eliminate time wasted by valuable team members performing manual, low-value tasks. 

For example:

  1. If the dispute is that invoice was never received, the system will automatically send appropriate invoices to a customer.
  2. If supporting documentation is missing, and this documentation is available in the directory - it will be delivered to a customer automatically. 

Consider the amount of time saved when your team members aren’t researching, copying faxing and emailing invoices, or appending documents that are already available in the customer’s files. 

In addition, consider the value to your customer when the response to their dispute is swift. They know their message has been heard and they have immediately an answer specific to their complaint. Instead of pushing the low-value disputes to the bottom of your AR team’s priority pile to be delayed or ignored, you have implemented a smart system that takes care of these issues with ease and speed.

Using the technology available today to create a smooth process and interaction for what can be a difficult customer touchpoint brings very tangible benefits to your business. The ease and speed that automation offers in these three areas of dispute management translates to your cash flow, your forecast accuracy, and reducing your DSO, on top of increasing your value to your customers. It’s just smart AR.

Nicole Dwyer
About the Author

Nicole Dwyer is Chief Product Officer for YayPay, bringing more than 10 years’ experience in accounts payable and receivable technology to ensure YayPay continues to meet the needs of its customers. Having spent her entire career in commercial payments, Nicole understands high- and low-value payment systems, the complexities of how businesses pay and get paid, and has worked with distributed teams spanning the globe. She is a graduate of Worcester Polytechnic Institute. Residing in New Hampshire with her husband, daughter, and son, they spend their time outdoors and creating new adventures.

Read more articles by Nicole Dwyer