As the role of the CFO evolves to focus on operational competitiveness, finance and accounting is no longer considered a back-office-only function; it also delivers end-to-end commercial value streams, e.g., quote to cash. As such, there is an increasing demand to eliminate organizational silos and drive process transformation across the value chain.
In a recent survey conducted by the Genpact Research Institute and analyst firm HfS Research on finance in the digital age, about a third of the finance leaders said that robotic process automation (RPA) tools and applications are having an impact on their operations today. Indeed, RPA is the finance and accounting technology they expected to show the greatest growth in impact over the next two years – from 34% to 58% – a massive jump.
RPA is not new but its definition is redefined every day. Finance processes have used rudimentary forms of RPA in transactional activities for many years, such as optical character recognition to read paper invoices.
What is exciting now is the availability and coming of age of new robotic software that enables automation across a spectrum of processes and works across different software platforms and systems. This change has been further catalyzed as digital technologies like machine learning, cognitive computing, natural language processing and generation, and advanced visualization mature.
As a result, both software developers and specialist business process service providers are mashing up these technologies to create fit-for-purpose solutions for specific areas of the value chain and to drive discontinuous performance improvements.
The benefit equation has also changed dramatically, from a pure labor/software cost arbitrage and 24 x 7 capacity augmentation to fundamentally redefining process outcome expectations in terms of speed, agility, customer experience, and the use of analytics to drive business strategies. All this of course comes at a fraction of the cost and with dramatically less rework.
RPA is a continuum: from simply mimicking repetitive human actions at one end, to, at the other extreme, seamless collaboration in human-digital interactions to perform judgmental work with superior speed, added value, and analytics.
Let’s take a few examples.
At the simple end of the spectrum, consider a payables process in which electronic data interchange penetration is still low and requires a great deal of manual effort to key data into different systems and route exceptions through workflows. This is where RPA is often used to deliver, in this example, a 2.5 to 4 times cost arbitrage. A principal advantage here is that RPA can link existing systems — such as enterprise resource planning (ERP) systems, documents, and databases — without requiring direct integration.
Now let’s take a more evolved example. At manufacturing companies, the order management process typically requires manual intervention to manage kickouts or exceptions. These are largely driven by missing information or discrepancies between the customer taxonomy in the orders and the company’s master data models. Machine learning can help automate the data-capture process and manage exceptions with built-in analytics, visualization, and notification processes to highlight and resolve issues. The benefits move beyond cost reduction to fundamentally improving the customer experience and supply chain performance.
This mash-up of technologies sits on top of the ERP as a system of engagement. Cloud-based advanced technologies overlay a company’s existing system of records/ERP to transform business operations and provide quick, easy implementation and scalability without heavy investment.
Let’s take another example, a cognitive buying assistant in the procurement process. One of the biggest procurement challenges is how to make it easier for people find the exact goods or services they need from the company’s catalogs. This difficulty often leads to maverick buying, which reduces managed spend and increases cost. An advanced robotic chat model using natural language processing and a search functionality on the catalog’s underlying databases can simplify the procurement process, improve the end-user experience, drive adoption, and increase managed spend.
Business performance reporting sits at the complex end of the spectrum. In most companies, the best and most expensive finance professionals spend two-thirds of their time compiling internal data from a variety of systems and databases, cleansing and repurposing it with very little time to incorporate significant external information, and then struggling to create insightful commentaries within set time frames that can be used to drive the business.
New technologies provide the power to significantly increase satisfaction with these processes by driving more value through a different human-digital collaboration where a combination of big data, analytics workbench, visualization, mobility, natural language processing, and natural language generation work iteratively with finance experts to redefine performance, capacity, and capabilities in business performance reporting.
This is all great news. Yet dramatic performance change in finance and accounting processes through RPA is not yet as prevalent as one would expect.
There are some critical dependencies for generating real value from RPA:
We are at the start of an exciting journey with RPA changing the finance and accounting landscape, and combining RPA with other advanced digital technologies is becoming more prevalent. For example, our research shows that more than two thirds (67%) of finance and accounting executives project that cognitive computing will have some or a major impact on their processes in two years, compared with only 43% who say so now. Five years from now, we will look at some of today’s processes and probably shake our heads in amazement that we managed to run businesses with our current ways of working.
Shantanu Ghosh is senior vice president, CFO Services and Consulting at Genpact, a provider of digitally-powered business process management and services.