Banner

September 24, 2019
by Tim Ray
Stop me if you’ve heard this one:
Electronic payments would solve the challenges of cash allocation. 
Accounts receivable professionals have long been told that electronic payments would facilitate the seamless receipt and flow of payments and remittance data, without human operator intervention. Things haven’t worked out that way for most accounts receivable departments. The problem is:

  •       Remittance data is not standardized 
  •       Remittance data is frequently incorrect, if it’s provided at all 
  •       Staff must search remittance advices for critical data such as invoice numbers or PO numbers – and there’s no guarantee that important data will even appear on the remittance advice
  •       Remittance advices might include data such as reference IDs that make no sense to the biller
  •       Remittance advices often arrive decoupled from electronic payments
  •       Global businesses must manage multiple currencies, banks, systems, payment types and more

As a result, accounts receivable departments spend a lot of time hunting down remittance advices, matching payments with remittance advices, manually keying data, juggling multiple workflows (for lockbox payments, checks, and ACH and wire transactions) and resolving exceptions. When businesses receive remittance advices with their payments, there’s no guarantee that they will post cleanly. Receivables departments must rely on time-strapped IT resources to map the data on remittance advices, ACH payment records and the file formats delivered by bank lockbox providers. 

All this creates downstream problems, particularly in collections and customer satisfaction.  

It is for these reasons that more accounts receivable departments are planning to deploy artificial intelligence (AI) to automate their cash allocation process. AI uses computer systems to automate repetitive tasks and to collect and analyse large data sets to uncover insights and make decisions. The intelligence gathered by AI also enables the technology to learn over time to improve results.   

AI-driven solutions initially use intelligence, user behaviours the technology has observed, and the data from remittance advices to suggest options for allocating payments. But as the technology observes user behaviours over time (such as the customer numbers to which payments are allocated, parent/child relationships, customer payment timing and common deductions) AI posts a greater percentage of payments automatically. AI cannot allocate all payments automatically. But it is not uncommon for businesses using the technology to allocate more than 85 percent of their payments straight through.  

One healthcare organisation was able to allocate 79 percent of its payments – at a 99.88 percent accuracy rate – soon after beginning live production with an AI-driven cash allocation solution.  

The same intelligence that enables AI-driven solutions to learn how to allocate payments also is used to recognize and forecast customer payment behaviours (e.g., when customers will pay and how much). 

All this helps organisations address the challenges of allocating electronic payments. 

Common misperceptions about artificial intelligence

Accounts receivable professionals are understandably excited about the potential of AI to solve their cash allocation challenges. But many of them also are confused about how the technology works.

Half of the participants in a recent webinar sponsored by Rimilia said their accounts receivable department has little understanding of AI and that they are not currently using the technology. One quarter of participants on the webinar said their organisation “somewhat” understands AI.    

Regardless of where your organisation is in its understanding and use of AI, below are six common misperceptions about AI and how the technology can automate the cash allocation process:

  1. AI is bleeding-edge technology: AI is not new technology. Businesses have a long track record of using AI to learn the actions required for various business allocations. 
  2. AI the same as robotic process automation (RPA): Lots of accounts receivable professionals mistakenly use AI – technology that is concerned with “thinking” – and RPA – technology that is adept at “doing” – interchangeably. While the technologies complement one another, they are very different. RPA replicates manually intensive, rules-based tasks that are normally performed by a human (things like retrieving documents from a customer portal). AI takes automation to the next level with advanced automation and data analysis.
  3. Everyone is using AI: Some technology providers are using AI as a tagline to describe how their solution works, regardless of whether the solution actually uses the technology. Drill down into how solutions providers are leveraging AI, as well as their plans for the technology.
  4. AI solutions require remittance advices in order to be effective: Remittance advices are inherently flawed; data can be incomplete or inaccurate or the scan may be bad. Advanced AI solutions don’t rely on unreliable data from remittance advices. The technology can allocate most payments without the need for remittance advices. Moreover, the technology learns customer behaviours to deliver better results over time.    
  5. There won’t be any exceptions: AI-driven solutions won’t eliminate exceptions. But the technology can provide users with suggestions for matching payments. 
  6. AI initiatives can wait: Eleven percent of the participants on a recent webinar sponsored by Rimilia said AI is already widely used within their organisation. Cash allocation represents a perfect storm for AI: a business allocation that is ripe for automation and a technology that has evolved to the point where it can address these needs.   

Ready to automate your matching of payments and remittances? Click here to listen to our recent webinar, “Conquering the Chaos of Applying Electronic Payments through Artificial Intelligence.