Advances in technology will continue to impact the credit management profession. Combined with the emphasis the C-Suite has placed on the digital transformation of finance functions, expectations are for an even more automated future, supported by much deeper customer and process insights.

The current state of technology has brought automation and transactional transparency to large swaths of the order-to-cash (O2C) process. Artificial Intelligence (AI), Machine Learning (ML), and Robotic Process Automation (RPA) technologies have not only improved on that but are now being used to fill in the remaining gaps. As a result, much of the work that drives the O2C process can now be handled autonomously.

In other words, the machines process the workflow in the background, leaving credit and collection professionals to make any decisions that are outside of setting parameters and to handle other exceptions. Furthermore, decisions made using such a system will be supported by holistic insights gleaned from the order-to-cash process, your customer's procure-to-pay systems, and external market and corporate insights. Ultimately, this will facilitate corporate credit professionals playing an expanded role in their enterprises' third-party risk and working capital management endeavors.

To get a better sense of the direction technology is taking the credit profession, I spoke with Sayid Shabeer, the Chief Product Officer at www.HighRadius.com. Our conversation ranged from the effects AI and ML are having on credit and collection activities to how innovations in the payment space are impacting accounts receivables management. For your convenience, an executive summary of the interview is provided below, followed by an edited transcript of the entire conversation. We think you will find the discussion and insights intriguing.

• Executive Summary
  A combination of Data, Software, and AI is being used to create highly automated AR, Credit, and Collection processes along with enhanced decision support tools
  These autonomous receivable processes will collect and analyze much more internal and external data, and present it to the user with better context and deeper customer insights
• These autonomous receivable processes lift credit management to a higher level by enabling a focus on holistic decision making driven by insights gleaned across the entire order-to-cash process
• Natural Language tools will provide credit managers with customer intelligence derived from a wide range of unstructured data sources (periodicals, press releases, Internet searches, etc.)
• Blockchain technologies have the potential to be transformational, but despite progress in the area of international trade finance, have not yet demonstrated themselves through mass adoption
• The growth of Alternative Payment Methods (APM) has been accelerating, particularly in Europe and other parts of the world, and solutions such as Real-Time Payments (RTP) are expected to soon become commonplace within the US marketplace, accelerating cycle times and receivables processes
• Invoice Financing at the point of sale is expected to grow with products like B2B Buy-Now-Pay-Later (BNPL) and other technology-enabled solutions gaining traction in the marketplace, which makes it possible to essentially outsource credit for a firm's smaller customers in addition to first-party outsourcing of collections, which incidentally are facilitated by autonomous receivable processing
• Suppliers need to adjust to the continued growth of commercial Credit Card Payments by implementing systems that will help them contain the costs and manage the security and privacy risks
• As technology facilitates integrated receivables and in order to optimize receivables processing and risk, credit managers need solutions that provide them with 360-degree visibility of the order-to-cash process and all the participants in that process, including internal and external customers
• Credit Managers will need to play an expanded role as part of the office of the CFO, which itself is looking at the interactions between the different areas of the finance function (order-to-cash, procure-to-pay, record-to-report, and treasury) in order to derive greater efficiencies

Interview with Sayid Shabeer, Chief Product Officer, HighRadius

David Schmidt for Credit Today: What new order-to-cash tools should credit executives anticipate being available over the next two to five years?

Sayid Shabeer for HighRadius: That's a great question. Looking around the corner is part of our job. This might be new terminology that I'm using, and I'm happy to explain a little bit, but what we are betting on is what we are calling Autonomous Credit. I'll try to break it down using the analogy of three legs of a stool. One leg of this stool is data. But data, not in the sense of credit data that we are used to, from various credit agencies. We're also talking about data that is created as part of the buyer behavior, for example, how are these buyers paying? That set of data is going to the credit agencies, but once a client is with us, we're seeing how a buyer is paying over months and years. The data that we're talking about is not just external data, but data on the payment behavior of buyers over time.

David Schmidt: I assume you are seeing that through your EIPP (Electronic Invoice Presentment and Payment) platform as well as the collections module.

Sayid Shabeer: Exactly. Our integrated receivable suite's cash application is where we know through our integration with the banks, the payment activities of every company buying from our customers: who is not paying, who is late, how much are they paying, are they paying the full invoice…all of that kind of information makes for very rich data.

Now, the second leg of this stool is probably the most exciting part for us. Data has always been available -- people download credit reports – and then review them manually. The second leg is the software where we're talking about decision support for the credit analyst or the credit manager. It is not about downloading a credit report and then doing a bunch of things in Excel with some modules for scoring risk and so on. It's really about, for any approval, how can you automate a lot of things, then be able to do the things where the human has to make the decision and support them in making that decision. That's where software, for us, comes in. Automating where necessary and supporting decisions by credit analysts and credit managers will become the norm.

DS: As we translate that into practice, how do that impact credit organization and the routine of credit departments?

SS: There are a couple of things when I think about the impact on the departments. We know that a lot of credit departments still spend a lot of time trying to corral and bring data together. Whether it is a question of pulling data from credit agencies or analyzing how a customer is actually paying overtime, a lot of time is spent on pulling data together versus actually making decisions. The software includes using APIs so that data is automatically coming in, counting is happening, and it is supporting your decisions versus you spending time trying to bring data together in order to make the decision. One of the big changes that I'm betting on is that credit managers and analysts using autonomous credit will be spending a lot more time on value-added decision-making versus trying to pull a lot of the data together.

A second important change I anticipate relates to periodic reviews. We're going to move to more of a continuous review. Events are happening in real-time: bankruptcy, notices of all kinds of changes, COVID, and so on. Instead of the analyst or the manager periodically trying to find information for a periodic review or reacting to something that occurs, now there is going to be more data coming in, in real-time, being evaluated by the software with the AI (artificial intelligence) looking at the patterns, and then proactively saying here is an alert. This is something you need to take care of, or this is something for you to look at as an action item in your worklist. It is going to be more proactive in terms of your worklist versus a periodic review or a reactive mode. It becomes almost a continuous review.

The second aspect of that continuous review is that when it is manual you have to reactively do things, or when you are doing it periodically, there is only a certain percentage of your portfolio you can cover due to the manual capacity of your team. If you have 50,000 customers, you might decide to look at just 5,000. You might look at just your largest customers in detail. Now, because of the software, the AI, and the fact that data can be brought in by the software in real-time, now you can actually review all 50,000. So, it's not just about moving from periodic to continuous review but covering the full portfolio versus a limited part of it.

DS: That makes a lot of sense. Let's move on to the third leg of your stool.

SS: I've already talked about it a bit. The last leg is AI. It has become something of a buzzword, but what does AI really mean for me as an analyst or a credit manager?

Let's think about certain use cases. One of the biggest challenges is blocked orders. Credit managers are used to dealing with credit holds. When a sale comes in, you have a short time window to respond. A lot of releasing orders from credit hold. We are now using AI to predict when an order will be blocked with a credit hold. Given the historical data and patterns of orders coming in, we can predict that new order is expected in the next four days and that it will go on credit hold. Now you have the potential to look at getting paid ahead of time to reduce potential credit holds.

If an order gets blocked in spite of the prediction, now the AI tool can recommend the action to take. Should we release the order because we are likely to be paid shortly even if we don't do anything or should we increase the credit limit? These are the kinds of decisions that the software can suggest to humans.

DS: So, from the credit manager's perspective: I haven't been able to be proactive. I know these orders are coming through and I've tried to take care of as many as possible, but some will end up on credit hold. The system is now going to say, 75% of these holds are being paid within the next week, and things like that.

SS: Right. Just go ahead and release this order because you are likely to be paid. It might even suggest increasing the credit limit to unblock the order. The patterns are important. Even before a credit hold happens, predicting that given the pattern of orders, go ahead and increase the credit limit, and request payment four days or a week before the order is placed is a very powerful thing as well.

Making recommendations to proactively change credit limits is another use case for AI based on the historical pattern of how the customer has been paying and ordering. All kinds of credit limit changes can be recommended by the AI module.

There are even more examples that can add value, but I wanted to illustrate what it means when you talk about autonomous credit with data, software, and AI, and how these things translate into real benefits that will change how credit is managed in the future.

DS: Let's move along to the payments area. A lot of payment innovation is being brought to the marketplace. Credit Today interviewed Lorenzo de Medici, Chairman, and Edward Boyle, CEO, of Medici Bank International, a so-called challenger bank. Among other things, they are using blockchain technology to accelerate trade finance. The use of smart contracts, stable coins, and real-time payments shortens the settlement cycle, reduces costs, and provides transparency across the entire transaction. Are there other areas of business credit that can be affected by this optimization that they're looking to achieve?

SS: Blockchain is a technology that we've looked at. We've spoken to some CEOs of blockchain innovators and companies. In my mind, I am still looking for concrete, wide adoption use cases of blockchain technology that create value. Clearly, the currency case, whether it's Bitcoin or other use cases of blockchain technology is well-known. The other use cases, as you talk about for trade finance and for other cross-border payments, big players like Visa are doing things around that. However, I'm still looking for wide adoption and what is a clear use case of blockchain technology. A lot has been proposed, but what has taken off in terms of actual adoption I'm still waiting to see. We are not directly involved in those things at HighRadius, but I'm looking to see what really emerges there.

DS: Right, that's totally understandable. I'm trying to keep up on the innovations, and it seems like blockchain might have applications between large trading partners in terms of all the deductions and disputes, and putting all that on a ledger where everybody can see it and work together to resolve it compared to the way things are often handled now with passing spreadsheets back and forth. But it's a good point. Someone's got to prove the commercial use case, and we haven't seen that yet.

SS: In theory, it makes sense. In the consumer world, people are used to networks, whether it is Facebook, or professionally such as LinkedIn, where there is collaboration and communication, and so on. Although there is a clear value that has been talked about for blockchain and distributed ledgers and decentralized finance, why would the traditional technologies be open to this transformation? A single entity could emerge or these distributed entities could jump in and build the technology, but there needs to be enough who are willing to cross the chasm, as we call it. Before I can believe the theory translating into practice, and why couldn't a player solve many of these issues with traditional technologies, Blockchain must cross the chasm from theory to practice.

DS: One other thing that's going on in payments is the whole idea of buy now, pay later (BNPL), essentially autonomous invoice financing by a third party at the point of sale. Consumer adoption is growing quickly, and we see a lot of investment in B2B applications of BNPL as well as real-time payments. These things appear to have the potential to change credit management.

SS:BNPL is definitely making a big impact on the consumer side. In terms of B2B, which we mainly deal with, we've not yet seen much impact. We're attuned to what our customers are looking for, and we still have not seen BNPL move into the B2B world yet.

Internationally, what is important in the B2B world is that it is not just about ACH or wire or paper checks, as is dominant in the US as far as payment methods are concerned. As to B2B payments in other parts of the world, there are very many different modes of payment such as SEPA Direct Debit in Europe, Giropay in Germany, iDEAL in the Netherlands, and all the other payment methods across the globe. Those are sometimes dominant methods of payment in those countries. We have to support multiple payment methods for our multinational clients, so the growth of so-called alternative payment methods or APMs, has been significant.

DS: The Europeans seem to be well ahead of the US in real-time payments, which obviously has implications for getting orders released and shortening the trade finance cycle. For some reason, there seem to be a lot of BNPL firms out of Australia.

SS: Is that happening with B2B in Australia?

DS: They're doing both B2C and B2B. A lot of them have moved into the B2B arena. We're also seeing a company called Apruve in this country advertising the idea of turning over what they call "long tail" customers for BNPL and other next-day settlement solutions on a company's invoices. In other words, applying the 80-20 rule: give us the 80% of your customers that are smaller and let us handle the credit for them, so you can focus on the important customers.

SS: You're spot on. I'm familiar with the company you just mentioned, Apruve. Traditionally there has been receivables financing, and new forms are being enabled by technology, making it easier to do receivables financing. I do expect that there will be new players, especially with the capital behind them to enable those kinds of financing.

DS: And I think some of it's the democratization of payments, because obviously, the larger companies have access to different types of receivables finance, much more so than smaller ones, but as these tools, these different payment rails show up, I think smaller businesses will then have other options.

Let's move on from that. Are there any other drivers of change that you're seeing? One of the things we see is CECL regulations. Outsourcing is another potential catalyst and has been around a while, but technology now allows the outsourcer to work more closely with the client. Also, AP innovation continues to push changes in credit. One of the big complaints we always hear from credit managers is, "Now I'm entering my invoices on their portal." So, the AR manager has become an AP clerk. I think some of what you've talked about in terms of AI can solve some of these problems, but we'd like to hear how you see these things driving change.

SS: You brought up the AR manager having to upload invoices, and that's where the software part of it, through APIs, is pushing information over to the AP systems from the AR systems. Where APIs are not available, there are even possibilities for RPA bots to upload invoices into the AP portals instead of humans logging in and uploading invoices. These types of use cases are strengthening the reason for autonomous credit, where software enables automation and takes care of the manual work that is a pain for a lot of people. This will free up credit analysts to really be focusing on the decision-making.

When we think about that balance between AP and AR, there's definitely a growth in card payments. Although it's a smaller sliver than for checks and ACH payments, this 5-10% sliver is actually growing at a higher pace. Card payments are growing, mainly because of the cashback received for making payments on cards, which converts the AP department from a cost center to potentially a revenue center.

It is the supplier side that's ending up paying for that card in terms of the interchange fees and other processing fees, so that's definitely an area of tension there. And handling those card payments in a secure way is an issue. If they send an email to a supplier containing card data, suddenly you have PCI data in your system. We have to support PCI-compliant ways of receiving those emails, or other payments securely and processing them versus having to handle them in an insecure or a manual way, which is happening currently to a lot of suppliers.

DS: We recently conducted a survey on credit cards and alternate payments. A lot of Credit Today's audience is smaller credit departments, typically 1 to 3 person departments. There's usually not a treasurer, so they report to a CFO or a comptroller, and Credit ends up handling a lot of this payment stuff. There are different things they try to do to manage the fees, whether it's using surcharges or only accepting credit cards at the point of sale rather than 30 days later on the due date. The costs involved with accepting credit cards, over the last six years, have probably doubled. It's not a huge percentage, but it's now become a significant percentage.

SS: That's one of the other issues that's driving change. I do see that from an AR perspective, we need to help our clients be able to first of all securely handle credit cards, and secondly be able to offset some of the higher costs that, are imposed on the supplier, whether it be a surcharge or other mechanisms.

DS: To finish up, I have a few open questions for you. How can credit executives expand their role in the corporation? How do you see that happening? One area we know the treasurers are talking about is getting more control over AP and AR and the whole working capital situation at their firms. Another is third-party risk management. Our readers are more and more getting asked to provide credit evaluations for their own company's suppliers. Any thoughts along those lines?

SS: One aspect is, that credit executives now have to not just think in terms of the credit function alone. They really need to think about the entirety... what I call 'integrated receivables.' That's step one. It's no longer a silo of credit, and maybe even collections. But really being able to look at credit, collections, cash application, invoice presentment, payment, and deductions - the five pillars, as we call it, of integrated receivables.

They're all interlinked and interconnected, but how does that inform the other functions? And more importantly, how do the other functions inform credit? A simple example is a case where you might have high credit utilization, and obviously, you want to have collections be able to collect from those firms so that you don't have blocked orders. Of course, we also have the classic case of bad debt being a consequence of any credit decisions you make, which shows up more on the collection side. There are connections both ways. When a collector is making calls, they need to have visibility into the credit utilization, but this is not commonly available right now. It's all in silos.

One aspect of it for executives is that they need to have clear visibility within receivables or credit. What's happening in collections? What's happening with aging? What's happening with bad debt, what is the DSO overall? What is the cash application cycle time? How much in deductions is being written off? Credit executives are being asked to provide visibility on all these things and more. Now expanding further beyond that, when you think about other functions like treasury looking at cash forecasting and cash management, they need to see the entire working capital situation (AR, AP, and inventory) in order to optimize it. It is very common, for example, for the GBS (Global Shared Services) head to be asked to now play a role in those decisions because they have visibility of these things. It was very much highlighted during the early days of COVID.

Having that information and being able to be part of the overall solution is crucial in terms of being relevant in the future state that I talked about, autonomous credit. In fact, our thesis is that a lot of the CFO functions are extremely correlated, and there is going to be a consolidation within the Office of the CFO. It's no longer going to be silos of receivables or the procure-to-pay or the treasury or the record-to-report. We know that the Office of the CFO is really looking at interconnects between these and visibility across these. So, whichever area of the Office of the CFO you're playing in, it is very critical that you have visibility to other areas and we're betting that software and people will enable that. Software tools will enable this versus the growth of email exchanges or other communication platforms that are more common in the current world.

DS: Right, and that ties into what we talked about earlier, about the credit manager becoming more process manager than just an account manager. That whole integrated, holistic outlook, is not just procured to pay or order to cash. It's more like procure to cash, where both trading partners have visibility of each other's processes.

SS: From the credit perspective, how do we get to a single pane of glass allowing a 360-degree view into the entire receivables? We are moving from just having a view into what's happening in credit to more of an overall perspective on AR, which encompasses all the five pillars. A single pane of glass, a 360-degree view of what is going on with your end customers, and AR becoming an input into other functions. The classic example is that of the treasurer, where AR, the incoming cash, is probably the most important factor in and the most uncertain part of cash forecasting. AP and Inventory are more in the company's own control. That's where the cash forecasting pressure becomes relevant. Shortening some of these things, whether it is bad debt or reducing DSO became very critical for the working capital, which is obviously a key ingredient for the CFO.

DS: One final question. Five to ten years from now, how will credit departments have changed in terms of staffing and functions?

SS: Just to clarify, when I talked about consolidation, I mean consolidation in terms of tools that can break the silos and provide that 360-degree view of customers. I do expect that all the functions would remain. Whether it is credit or AR or the quarterly report, those are core jobs that will be remain, but there will be more consolidation in terms of the tools to enable the different departments to get that 360 views through a single pane of glass.

Beyond that, when you think about how credit departments change in terms of staffing and functions, I think it is going to be more strategic. These are strategic functions, and they're going to be more for folks who focus on decision-making, rather than on a lot of the mundane and manual tasks we now do. As companies grow, they can now focus on the right decisions versus scrambling to get the work done. They're going to be monitoring only a certain percentage of clients, they're most important because that's all they have the capacity for. Let's think of it more like a pilot who's getting all the alerts and has to take care of things to keep the plane flying on the course, versus worrying about a lot of the details of "How do I get this loaded?" or "How do I get that done?" Those tasks will all get automated, so that you can focus on the real decisions, becoming the human working with the machine, which is the reason we have autonomy, versus dealing with a lot of basic details.

DS: The analysis side of things ties in with what we're hearing elsewhere. The worker of tomorrow is going to have AI on their desktop so you can become more of a decision-maker, by taking the information and having the analytical tools to make good decisions with it.

SS: The right information being surfaced to you in context, versus you having to go find it.

DS: Something we haven't talked about is Natural Language Tools. I've seen companies that now provide sentiment analysis, from looking at publications and any other public information that's out there and then providing scores and summaries.

SS: One hundred percent. Part of our vision for autonomous credit I talked about is data beyond that from a credit agency. I talked about the buyer's own behavior, regarding how they are paid, but it goes even beyond that. As you were saying, there are other data sources that can be streamed in, because of APIs and other new technologies that can collect data that you would not normally have evaluated as part of your credit decision, which is where some of the AI also comes in, to interpret the data and surface what is relevant to your decision and suggest the right kind of decisions. That's truly the autonomous spirit.

When I say AI, it can involve simple statistical things, simple rules. That's one end of the spectrum. At the other end of the spectrum, there are sophisticated machine learning models operating on a large data set, including machine and natural language processing (NLP), that you talked about, to interpret the data and extract the intent or the summary of that information. So, everything in between is what I'm thinking of as AI. Data, software, and AI work together to enable better credit decisions.