4 world class GPOs from Cargill, Air Products, Keurig Dr Pepper and Danone explain how they changed the tides in their favour and prepared their A/R for future.
With great leaps in digital innovation, the emerging technologies have revolutionized the world and continue making a big impact in all spheres of life including businesses across all industries. Consumers and enterprises across the globe have a lot of potentials to look forward to, as frontier technologies like AI, ML, RPA, IoT and Big Data, are finally available to mass markets.
But the terms AI, machine learning, and robotic process automation are often used interchangeably when there are key differences between each type of technology. With the overuse of these buzzwords, it is critical to understand what these terms really mean. This section provides a brief sketch of these trending technologies.
Robotic Process Automation or RPA is an application of technology aimed at automating business processes. Using RPA tools, a company can configure software, or a “robot,” to capture and interpret applications for processing a transaction, manipulating data, triggering responses and communicating with other digital systems, according to the Institute for Robotic Process Automation and Artificial Intelligence. RPA scenarios can stretch across a variety of tasks from automating workflows, managing infrastructure to labor-intensive back-office processes. The main goal of the robotic process automation process is to replace repetitively, time-consuming, labor-intensive clerical tasks performed by humans, with a virtual workforce.
Artificial Intelligence is a way of making a computer, a computer-controlled robot, or a software think intelligently, in a similar manner the intelligent humans think. It is an umbrella term for a machine’s ability to imitate a human’s way of sensing things, making deductions and communicating. An example of this is machine vision, which allows for real-time traffic counting from the feed of a traffic camera, the anticipation of congestion, and alerts concerning potential emergencies.
Machine Learning is an idea to learn from examples and experience, without being explicitly programmed. Instead of writing code, data is fed into a generic algorithm, and it builds logic based on the data given. It is a method of data analysis that automates statistics and analytical model building. It is a branch of artificial intelligence based on the idea that systems can learn from data, identify patterns and make decisions with minimal human intervention.
While RPA involves processing rule-based, repetitive tasks in high volumes, it needs to be explicitly trained for new tasks and can only take inputs in predefined formats. Businesses leverage AI to perform complicated tasks that may have unformatted and unstructured inputs. As AI evolves with performance, it drives to find better solutions. ML, on the other hand, is used to identify and analyze trends to predict outcomes with self-learning capability.
With the extensive hype surrounding these new world technologies, readers would expect high adoption rates. As per a survey report by Statista, 84% of enterprises believe investing in AI will lead to greater competitive advantages. However, the reality is far different. The next section explores this paradox.
For the last couple of years, many companies of virtually all sectors of the business market — including finance, government, retail, telecommunications, utilities, energy, and transportation — have been enthusiastically considering the evaluation and management of AI-based projects. Some of them even consider this technology as the cornerstone of their digital transformation journey. However, as per a survey report by Peeriosity, only 13% of organizations are currently using AI. The low rate of AI implementation calls for speculation so as to evaluate why the AI adoption in AR is low, despite the sensation it has created. This section evaluates the challenges in AI adoption.
The root cause of resistance is information asymmetry. While there is a lot of chatter and hype around AI adoption in AR, businesses struggle with information asymmetry when looking out for options. The overuse of buzzwords has created a grey area in the minds of the A/R professionals regarding this technology and its applications. Credit and AR managers struggle to cut through the smoke and mirrors of buzzword-happy software vendors to understand what AI means for their processes. Moreover, the vendor landscape is highly segmented with different types of vendors offering different solutions and diverse portfolio – process automation vendors, custom AI shops and RPA vendors. Consequently, there is no clear understanding of what the expectations should be from an AI project leading to poor estimation of project ROI and benefits. Confused ROI expectations result in difficulty in stakeholder alignment and gaining buy-in from decision makers. The result is failed attempts at bringing in artificial intelligence to the A/R departments, therefore, the low adoption.
The first step to resolve the AI tangles is to understand the vast landscape of AI and automated solutions available today. The next section takes a deeper look into the fragmented technology vendor landscape and why is it difficult to choose an AI technology partner.
In the vast technology landscape with vendors advertising hyped jargon and offering a wide variety of features, evaluating all available options becomes extremely complicated for A/R managers. The most common automation technologies can be segmented into five basic areas and evaluated on two parameters:
1. the range of its application (from specific to broad)
2. its capability spectrum on the scale of human intelligence and judgment.
The five categories include:
• Macro or Scripted Automation: It consists of short sequences of code written to perform a single task or a series of tasks.
• Business Process Management (BPM): This category includes automation which implement various methods to discover, model, analyze, measure, improve, optimize, and automate business processes
• Custom AI: Automation customized for a specific activity or a defined set of tasks
• RPA: Automation of low-value, repetitive human actions with minimal human judgment
• Cognitive Automation/Strong AI: The ability of computer systems to learn, reason, think and perform tasks requiring complex decision making
The scattered graph is evidence enough of the wide-scale features and functionalities offered by different automation solutions. Further, if you dive deeper into AI, you would find not all AI are the same. There are different types of artificial intelligence capable of solving entirely different sets of challenges:
• Artificial Narrow Intelligence – The focus of Artificial Narrow Intelligence is narrow. It focuses on developing technology that is capable of executing a single task effectively, providing excellent service to the user in that particular area. This technology does not enable them to do complex tasks like a human brain. It can be applied where the work can be predefined and more information processing is required than complex decision making. Examples of this technology include customer queries, restaurant recommendations, and weather forecasts
• Artificial General Intelligence – Artificial General Intelligence focuses on a broader aspect and has a wider range of applications. This technology is enabled with the reasoning that is equivalent to or may even exceed human intelligence. It is expected to be capable of being applied in broader areas where complex decision making and in-depth analysis is required. Examples could be an artificial neural network that functions like an actual human brain.
Due to different features and functionalities that could be scaled with AI, it is imperative to understand and evaluate the current state of business and perform a thorough needs analysis to weigh the different options efficiently and identify the most suitable AI.
Against the backdrop of the hype and technology breakthroughs, the businesses need to distill fact from everything else. This section illustrates a simple and effective approach to shop for enterprise AI in this tricky market, by zeroing in on the right questions to consider. These questions should focus on the different facets of the problem and solution. This would help the readers to arrive at a checklist before they approach the vendors.
The following enlists the six essential questions and aspects of technology vendor evaluation:
1. Business problem: The first and foremost step is recognizing the fundamental need of technology for a particular organization. The key is understanding the real problem or opportunity you are trying to solve – is its growth, cost, compliance, customer satisfaction or something else?
2. Single solution: The second aspect is to look for a complete solution that addresses all the key concerns instead of identifying and resolving individual problems. This enables the firm to avoid the burden of completing incomplete solutions. Moreover, choosing from an inventory of systems and later dealing with integration problems is a nightmare that businesses would like to steer clear of.
3. Results timeline: The third aspect is the end result. It is important to look for true agility instead of misleading IT development philosophies. Definite timelines must be set up to achieve ultimate business outcomes and realize the true potential of automation.
4. Flexibility: Is the solution responsive and can rapidly adapt to change? The fourth aspect revolves around flexibility. To prevent frequent revisits and revisions, the solution must ensure scalability. Your team should be able to build further, go deeper, broader and expand into the features and functionalities offered by the automation without any fuss.
5. Non-intrusive: The fifth factor to consider is coherence. The automation should fit seamlessly into your existing technology landscape. The companies should be able to leverage the investments they had made prior to adopting the solution and not let the new solution render their current systems obsolete.
6. Fundamental change: The whole idea of an AI solution is to bring about a steady, scalable and effective change in the way organizations work today and not just bring in temporary, band-aid solutions. For this to happen, they need to understand what real AI features to look out for in the multitude of existing solutions.
With today’s aggressive economy, the A/R domain is all the more in need of the superpowers of AI. The next section explores how the AI and RPA features revamp the order to cash processes.
The credit and A/R teams across companies and industries are dedicated to evaluating all available solutions and leveraging the full potential of top-notch technologies in order to improve their order-to-cash functions. This section dives deeper into the four core A/R processes and gives a brief overview of how AI and RPA revamp these functions.
Credit management is an important business function in any organization as it drives revenue enhancement and risk containment. The process involves gathering credit information from different Credit Agencies, Credit Groups, Public Financials and Credit Insurance Bureaus followed by capturing credit application data from online or offline forms. A/R teams engage in credit scoring and risk assessment before onboarding customers and setting up credit limits. Credit management also involves other functions like periodic reviews and blocked order management.
Robotic Process Automation can be leveraged in credit management to capture credit application data by defining business rules and enabling template-based credit applications. RPA could also provide the functionality of auto-retrieval of credit reports from various sources. Further, it could also contribute to better collaboration for credit approvals, periodic reviews, blocked order release, customer correspondence by setting up pre-defined system responses.
AI, on the other hand, could provide deep-routed insights which in turn could help in improving credit policies and ensuring effective risk mitigation. It could predict blocked orders and potential customer default based on the analysis of a multitude of factors.
Cash application is another important A/R process that applies incoming payments to the correct customer accounts and receivable invoices. It involves downloading and aggregating payment and remittance information from a variety of sources including checks, emails, EDI files, websites or customer portals. This is followed by payment-remittance linking and three-way matching of payment, remittance and invoice information. Further, if there are short-payments, the reason codes are entered and discounts (if any) are applied. In cases of missing or incomplete remittances or any other discrepancy, analysts engage exception handling and finally, the cash is posted into the ERP.
For cash application, RPA can be implemented to aggregate remittance details from multiple sources including emails, portals and EDI files. With predefined rules in the solution, the payment and invoice information can be easily matched and short-payments can be auto-coded. Moreover, automated correspondence can be sent to customers in cases of missing remittance.
Artificial Intelligence enables data identification (example: identifying relevant details in check stubs) and accurate remittance information capture. Intelligent invoice mapping for unstructured or non-standardized data ensures that ‘all’ corresponding invoices, payments and remittance are matched irrespective of templates and formats. Auto-resolution of repetitive exceptions based on pattern recognition and trend analysis is another important feature provided by AI. Moreover, it can enable auto-triggered emails to customers with high probability of sending incomplete or incorrect information or altogether not sending the remittance.
Deductions Management consists of aggregating back-up documents like PODs, BOLs, and claims from different sources like carriers and warehouses, linking the documents to the corresponding deduction line item, followed by cross-department correspondence (example: with sales), verification and root-cause analysis. Based on the workflow and analysis, the deduction is approved or rejected and the correspondence is sent to the customer accordingly. As a result, the dispute is resolved and the system is updated.
RPA can be implemented for deductions management to aggregate back-up documents, enable automated correspondence for cross-department collaboration and support customer correspondence for approval/denial of deduction line items. AI can be leveraged to predict the validity of deductions which could, in turn, be used for prioritizing the deduction cases. This would ensure time and resource savings as the deduction analysts would spend more time on more critical research.
Intelligent linking of backup research documents with deduction is another crucial functionality that could be enabled by AI.
Collections are one of the most important processes in Accounts Receivables. For a business to have a leading growth graph, a business ‘needs’ to put efforts into Collections Management. It involves creating a prioritized worklist based on numerous factors such as credit data, payment history, correspondence information, invoice values, and aging data and utilizing industry best methods like customer segmentation. It further involves customer correspondence, logging correspondence activity, tracking payment commitments, reminders, and follow-ups. This is followed by payment collection and reporting.
RPA can be integrated into collections for generating prioritized worklist along with auto-recommended account level actions based on predefined business rules and strategies that take necessary factors such as credit data and payment history into account. It can enable automated collaboration with customers using templates and correspondence packages. Reconciliation of invoice payment and approval statuses from customer portals, third-party websites, is another feature provided by RPA for
AI could predict invoice payment dates at the customer/ invoice level based on historical data and trend analysis. Dynamic, real-time prioritization of the worklist can be realized with AI. Further, it could enable the prediction of customer preferences including correspondence time and mode, ensuring a high probability of collection success.
Since these millennial technologies are ready to storm the finance and accounting sector, a precinct guide to efficient implementation is a must-have for A/R managers today. The following enlists the five simple steps for an efficient A/R transformation with AI.
Artificial Intelligence is present all around. It is carried around in pockets, in cars and homes, and influences human lives on a scale bigger than people realize. While the smart minds continue to explore the seemingly infinite potential of this technology, the complex domain of accounts receivables makes true understanding a nightmare for credit and A/R managers. However, with a clear knowledge of these ‘pop’ technologies, the A/R managers can evaluate the functionalities of different solutions, select the perfect match, implement the solution to improve their process and ultimately secure their next promotion!
HighRadius is a Fintech enterprise Software-as-a-Service (SaaS) company. The HighRadius™ Integrated Receivables platform optimizes cash flow through automation of receivables and payments processes across credit, collections, cash application, deductions, electronic billing and payment processing.
Powered by Rivana™ Artificial Intelligence Engine and Freda™ Virtual Assistant for Credit-to-Cash, HighRadius Integrated Receivables enables teams to leverage machine learning for accurate decision making and future outcomes. The radiusOne™ B2B payment network allows suppliers to digitally connect with buyers, closing the loop from supplier receivable processes to buyer payable processes. HighRadius solutions have a proven track record of optimizing cash flow, reducing days sales outstanding (DSO) and bad debt, and increasing operational efficiency so that companies may achieve strong ROI in just a few months. To learn more, please visit https://www.highradius.com/.
Integrated Receivables is a solution to optimize accounts receivable operations by integrating all receivable and payment modules to work as a unified business process. At the core of the Integrated Receivables platform are solutions for credit, collections, deductions, cash application, electronic billing and payment processing – covering the entire gamut from credit-to-cash. The HighRadius™ Integrated Receivables platform is a stand-out as it enables every credit and A/R operation to execute real-time from a unified platform with an end goal of lower DSO, reduced bad-debt, faster dispute resolution, and improved efficiency, accuracy for cash application, billing and payment processing.
HighRadius™ Integrated Receivables leverages Rivana™ Artificial Intelligence for Accounts Receivable to convert receivables faster and more effectively using machine learning for accurate decision making across credit and receivable processes. The Integrated Receivables platform also enables suppliers to digitally connect with buyers via the radiusOne™network, closing the loop from the supplier A/R process to the buyer A/P process.
HighRadius Autonomous Receivables Software Platform is the world’s only end-to-end accounts receivable software platform to lower DSO and bad-debt, automate cash posting, speed-up collections, and dispute resolution, and improve team productivity. It leverages RivanaTM Artificial Intelligence for Accounts Receivable to convert receivables faster and more effectively by using machine learning for accurate decision making across both credit and receivable processes and also enables suppliers to digitally connect with buyers via the radiusOneTM network, closing the loop from the supplier accounts receivable process to the buyer accounts payable process. Autonomous Receivables have been divided into 6 distinct applications: Credit Software, EIPP Software, Cash Application Software, Deductions Software, Collections Software, and ERP Payment Gateway – covering the entire gamut of credit-to-cash.