AI Increases the Value of Order-to-Cash Automation

If you search “for want of a nail” on your browser, you’ll discover the age-old wisdom that seemingly trivial things can have a far-reaching impact. It’s a parable for artificial intelligence used in business. Deconstruct the imagined big-picture impact of AI and there are thousands of minor tasks that soon will require little or no human involvement in the interstices of an end-to-end process. Humans will still be indispensable, but they won’t be doing predictably repetitive work. Collectively, this will produce step-function improvements in organizational productivity. It’s productivity, not efficiency, because rather than making small gains in the time or resources required to perform some step, AI eliminates the need for a human to do the work altogether. I covered some aspects of building AI from the bottom up earlier.

Order-to-cash covers all steps from customers placing an order to the receipt of payment. It includes order management and fulfillment, shipment, invoicing, payment receipt and application as well as the analytics and reporting associated with monitoring and managing the task. Key performance metrics for the process include time required, order fulfillment, days-sales-outstanding and those that measure execution effectiveness.

The last-named recognizes the importance of follow-on impacts of a better executed O2C process, such as customer satisfaction and removing barriers to a timely financial close. Ventana Research asserts that by 2027, almost all order-to-cash software providers will use GenAI to facilitate training and provide in-context help to boost productivity and finance and accounting departments.

There is a strong business case for digitally transforming O2C. Finance and accounting departments must turn to technology to digitally transform core processes and achieve greater productivity to minimize administration costs because simple efficiency is no longer a useful strategy. Rather than benchmarking individual steps to find ways to shave time and costs, departments need to eliminate the need for a human to do the work in the first place. They can achieve this by organizing work around end-to-end processes rather than siloed steps while automating repetitive, rote functions, streamlining actions and eliminating the need to perform unproductive process steps. AI makes the transformation process easier to achieve with limited risk.

The business case for AI-driven O2C transformation goes beyond the finance department because of the impact of cuts across multiple silos in an enterprise. It includes marketing, sales, customer service, warehouse management, finance and accounting. The investment is strategic because, beyond cutting administrative overhead, conceiving the O2C processes from a customer-centric perspective also helps organizations become “easy to do business with.” One aspect is in the humdrum step of accounts receivable. ISG-Ventana Research asserts that by 2027, at least one-half of enterprises will use embedded AI in managing accounts receivable to increase productivity while reducing transactions frictions for improved customer experience. Straight-through processing of all forms of payments, including paper checks, automated clearing house payments, credit cards via email and accounts payable portals also reduce transaction friction. AI facilitates cash application motions, ensuring that customers can use their full credit line instead of having purchases held up because already-received funds were not immediately matched to outstanding balances. And using AI can make accounts receivable a customer-facing organization.

Broad adoption of AI is underway to execute pieces of O2C. The impact of each AI-enabled individual change may be trivial and barely worth mentioning, but that’s an advantage of AI transformation. Rather than requiring a great leap with massive computing resources and a degree of faith, small-scale AI involves replicating easily understood motions with limited and easily verifiable outcomes. The supporting technology isn’t a general large language model prone to hallucinations but an orchestrated set of narrow, small and mid-size language models as well as domain-specific LLMs trained on enterprise-specific data. Such systems will never be completely infallible, but suitably reliable ones are within easy reach.

Increasingly, business software is designed to work with any number of applications or data sources, using application programming interfaces to automate the integration of processes and data. This arrangement is frequently referred to as a “platform.” From a technology standpoint, the application of AI to the order-to-cash cycle militates for the use of software that manages the process end-to-end on a single platform. That’s because of the need to use a single data set to continually train and manage O2C AI models as well as support consistent and accurate end-to-end process execution. This high-availability data set also supports better situational awareness and near real-time reporting and analysis.

The order-to-cash cycle is rarely thought of as strategic, and for decades the emphasis has been on squeezing costs out of the process. How it’s performed deserves a fresh look because it’s one of the tasks where AI can increase the business value of a cost center. I recommend that enterprises that do not manage O2C end-to-end do so and quickly adopt AI-enabled capabilities as they become available to support a more customer-centric approach to the receivables process.

Regards,

Robert Kugel