Collins | AI & New Robotics: Tools For Credit Departments

The other morning, I took my daughter for an early dentist appointment. While getting out of the car, she mumbled something about it being time to have robots, not people, take care of our teeth. I’m not quite sure about that.

What I am sure about, is that robotic process automation (RPA) and artificial intelligence (AI) are changing the way we do business. It wasn’t long ago that media credit and collection functions were handled at the local, or perhaps regional, level. The team knew the customers and could predict when to expect payments or problems. Information about potential new customers came from local sources and respected credit bureaus such as MFM’s BCCA subsidiary.

Consolidation, along with process centralization, means that credit and collections professionals can be far away from the clients they need to manage. While experience still counts, good decisions tend to require more research and often involve other departments. There seems to be more process and less service. It may be time to consider adding some automation.

In a special report called The Future of Media for the November/December issue of MFM’s member magazine, The Financial Manager (TFM), Bill Weiss, vice president of business development, credit and collections at High Radius, talks about AI and RPA for credit and collections. He comments, “if you are doing a job the same way that it was done a decade ago, it’s very inefficient.”

Weiss begins by imagining a workplace in which a virtual assistant guides a credit and collection manager through the day. This assistant could provide information about newly submitted credit applications and note missing data, weigh risks and assign credit ratings based on company-defined parameters. With the manager’s approval, such an assistant would confirm a credit decision, contact the sales team, and update company records, all in the space of a few minutes.

This scenario requires a combination of RPA- and AI-enabled technology solutions. RPA can be used for repetitive processes. Weiss uses the example of counting money or dispensing cash, as is done by an automated teller machine. Such processes “must be rules-based and repetitive, with pre-defined input formats.” When a task changes, the code must be modified; each new task requires new code. For media credit departments, RPA can be used “with high-volume, well-defined processes, such as pulling documents from web portals.”

In fact, several BCCA members use such processes to pull credit data from its Media Whys credit report database.AI systems, on the other hand, learn on the job. Such systems can “handle more complex tasks by ‘learning’ and adapting to changes.” In the world at large, it is AI that powers self-driving cars, voice recognition, and suggestions for video content on YouTube and Netflix. For credit departments, AI can recognize a change in customer payment history or in the payment itself — say from check or ACH to credit card.

Onboarding New Clients

Weiss believes that onboarding of new clients is the first use case for automation. Initially, an RPA system can be used to ensure that the new client credit application is complete. Just as happens when someone tries to complete an online form and neglects to fill in a mandatory field, the client will get an error message when critical information is missing.

Then, keeping in mind that RPA speeds rote processes, such systems can be used to pull credit data from pre-set sources (like BCCA’s Media Whys database), compare pre-specified credit application or credit data fields against defined values, and plug the results into a formula that spits out a creditworthiness score. Such a score, when combined with data points from the credit report(s) and or credit application can result in recommended credit terms.

Going a step farther, a media company may add an AI solution to profile all of its customers. Such systems can help in the case of potential new customers without sufficient data for an RPA-assisted credit decision. The program should be able to compare data from the new client to that of other, similar, customers. Each new case will help the AI system learn and improve its creditworthiness assessments.

Blocked Orders

Blocked orders, those orders that typically require attention from a credit analyst, may also benefit from automation. Weiss says today’s manual process begins with a notification to the analyst. That person then checks the appropriate company program(s) to find information such as what the customer owes the company and when the payment or payments were due.

He says that what follows can involve two or even three different company departments and be quite labor-intensive. A company representative will have to contact the customer to find out when the company can expect payment for the outstanding balance. If the customer’s response is acceptable, the appropriate person or department will be notified to release the credit hold.

Additionally, someone will need to follow up to confirm that payment is made as promised. AI programs can be used to predict blocked orders before they happen. Weiss proposes combining such a system with other financial technology software to automate correspondence with customers approaching their credit limits.

Such correspondence would include a link to an online payment portal. Once the customer makes a payment, the customer’s available credit limit is updated. Such automated actions can reduce or even eliminate the need for human intervention.

AI and RPA are here, now. Weiss cites a 2018 McKinsey & Co. study that found about half of the businesses it surveyed had already embedded AI into at least one corporate process.

So, while I’m not ready to have a robot clean my teeth or predict where or when I will need a filling, we have already accepted such solutions into other parts of our personal lives. It may also be time to consider them to streamline our businesses too.

The November/December issue of TFM will be on the MFM website for a couple more weeks. Weiss’ article includes a sidebar about using RPA and AI to streamline media collections processes.

AI For The Ad Sales Process

MFM’s January webinar looks at another use for automation — ways to streamline the media advertising sales process. On Jan. 21 at 1 p.m. ET, Matrix experts Brenda Hetrick and Adam Gotlieb will provide case studies from companies that have deployed an open API infrastructure and centralized systems to gain a single, holistic view of all relevant data resulting in optimized operational efficiencies and allowing them to discover new revenue opportunities.

All MFM webinars are offered at no charge to MFM corporate members, $50 for MFM individual members, and $75 for non-members. Eligible participants can receive one Continuing Professional Education (CPE) credit.

More information and an online registration form may be found on MFM’s website.

Mary M. Collins is president and CEO of the Media Financial Management Association and its BCCA subsidiary, the media industry’s credit association. She can be reached at [email protected] and via the association’s LinkedIn, Twitter or Facebook sites.