How Intelligent Automation Will Transform the Finance Function
As businesses seek to rebound from the global pandemic, the rise of intelligent automation will change the way organisations operate forever. This article examines how technologies such as machine learning will reshape the office of the CFO.
The need for finance to deploy more efficient, dynamic ways of working predates the global pandemic, yet the events of 2020 are proving to be a significant catalyst for technology transformation. For finance, that means embracing digital technologies, such as machine learning, that can be applied to core processes.
CFOs have long been looking to reduce the time spent on processes such as close, consolidations, reporting and payroll, and the COVID-19 pandemic and changes to where and how businesses operate have made this shift an imperative.
Thomas Willman, principal, finance advisory global practice leader at The Hackett Group, shares, “Finance has had to transform in so many ways in 2020. What hasn’t changed is that all of this work still has to be done; what has changed is that it has to be done away from the office. Finance professionals are exploring ways to increase automation and apply machine learning to identify patterns and make recommendations that previously would have required manual intervention.”
What Intelligent Automation Means for Finance Day-to-Day
In the right hands, digital technologies and greater automation can be a fantastic combination for CFOs to transform the finance function. However, success will depend on identifying and prioritising tasks that will deliver the greatest value. When it comes to automation, the first goal for a finance team should be to automate the repetitive and transactional tasks that require human effort or manual intervention. Doing this will free up a significant amount of finance’s time to be more of a strategic advisor to the business.
The second goal is to identify where digital technologies, such as machine learning, can be applied to detect, predict or recommend, ultimately infusing a greater level of “machine” intelligence into a transaction or process. Once the machine sees a pattern, it’s capable of applying the same result over and over, and as the machine continues to learn, it gets smarter and smarter.
The result – automation paired with machine intelligence – creates intelligently automated processes, thus eliminating much of the time that has been previously spent on traditional transactions and processes. A Workday Adaptive Planning survey found that over 40% of finance leaders say that the biggest driver of automation within their organisations is the demand for faster, higher-quality insights from executives and operational stakeholders.
The research in Accenture’s “Charting a Path to Intelligent Automation” report states, “About three-fourths of CFOs surveyed say they are helping to drive business-wide transformation, so getting things right in the finance function is critically important. Thinking through the end-to-end strategy, methodology and deployment of intelligent automation tools in the context of shaping the organisation rather than fixing a specific pain point is essential.”
To finance, of course, numbers matter, and when you put automation under the spotlight from a cost and efficiency perspective, the evidence speaks for itself. Research from an Argyle webinar featured in CFO Dive states, “A company with a 20-person finance team typically loses the equivalent of 1,920 hours annually, or an estimated $124,800 in costs, to manual tasks. A big company with a 100-person finance team might lose 9,600 hours, at an estimated $624,000 a year.”
Where Machine Learning Can Drive Finance Transformation
Despite the obvious financial and operational benefits of machine learning, many finance functions have been slow to adapt. Accounting, supplier management, procurement and auditing are all key areas that are ripe for automation, yet the risk – particularly for large companies operating across multiple geographies – can act as a barrier to innovation. Teams in each of these areas are also immersed in “keeping the lights on” – often at the cost of transformation.
Transaction processing is another barrier that prevents finance from achieving transformation and ultimately delivering a better business partnership. It's not surprising that it’s the first port of call for CFOs looking toward automation.
“Automation provides finance leaders with a great way of optimising the way they manage their accounting processes. This has been a painful area for finance for such a long time and can have a direct impact on an organisation’s cash flow,” says Workday’s Barbara Larson, general manager, Workday Financial Management. “Finance spends a huge amount of time sifting through journal entries, invoices and other documentation to manually correct errors while machine learning could automate this, helping to intelligently match payments with invoices.”
Machine learning can also mitigate financial risk by flagging suspect payments to vendors in real time, enabling a much more effective and efficient process. Internal and external fraud costs businesses billions of dollars each year. The current mechanism for mitigating such instances of fraud is to rely on manual audits on a sample of invoices. This means looking at just a fraction of total payments, and is the proverbial “needle in the haystack” approach to identifying fraud and mistakes. Machine learning can vastly increase the volume of invoices that can be checked and analysed to ensure that organisations are not making duplicate or fraudulent payments.
“Ensuring compliance to federal and international regulations is a critical issue for financial institutions, especially given the increasingly strict laws targeting money laundering and the funding of terrorist activities,” explains David Axson, CFO strategies global lead, Accenture Strategy. “At one large global bank, up to 10,000 staffers were responsible for identifying suspicious transactions and accounts that might indicate such illegal activities. To help in those efforts, the bank implemented an AI system that deploys machine-learning algorithms that segment the transactions and accounts, and sets the optimal thresholds for alerting people to potential cases that might require further investigation.”
Improving Financial Planning and Analysis
If you subscribe to the view that the role of financial planning and analysis (FP&A) in the future will be to deliver data-driven decision support for the business in real time, then it’s clear that finance must transform its processes to meet this vision – and automation is a central component in this transformation.
Research from McKinsey states that on average, approximately 60% of finance activities can be fully (40%) or mostly (17%) automated with technologies available today. Where FP&A sits on this spectrum is open to debate, but the same study claims that many tasks in this category have the ability to fully (11%) or mostly (45%) be automated.
Few could argue that there’s a transition going on from a spreadsheet-based FP&A culture to a much more collaborative, automation-based FP&A culture. It’s hard to say where we are in that transition, but the desire to move toward analytics and technology skills in finance versus spreadsheet skills is a pretty dramatic shift. In a CFO Insights survey, 78% thought Microsoft Excel® skills were most important two years ago; that figure is now 5%. The automation in applications that has become available to finance professionals is driving that shift.
As embracing digital technologies to drive greater automation becomes ingrained in the finance function, manual data gathering, consolidation, verification and formatting will disappear.
Automating Reconciliations
Finance functions today spend far too much time reconciling data across various systems. Think of the transactions between internal and external systems, as well as across various ledgers. Manually based tasks mean mistakes are inevitable, with duplication or data-entry errors driving inefficiency.
Robynne Sisco, president and CFO at Workday, saw this firsthand in previous organisations where she worked. “Each month, finance would have to close the period, access the data, reconcile it, format it and analyse it. By the time we delivered the numbers to the business, it was two weeks after the period ended and too late to take action,” she says.
Using rules and patterns, machine learning can provide finance professionals with the ability to identify a large number of these reconciliations, understand the problem, and in some cases, correct the problem or flag it for human intervention. Finance should be enabled to automate the reconciliation, consolidation, reporting and close processes so tasks are completed accurately within a single system.
Closing the Books Faster
At most companies, anticipation of closing the books is enough to send blood pressures soaring across finance. That’s in large part due to the number of systems involved in the financial close process, with input from various functions across the business. For companies, including Aon, the onset of the pandemic meant having to deliver the company’s first-ever remote close.
For finance teams dealing with multiple, disparate systems, the new tools and resources available to help close the books more efficiently and accurately can be viewed from two perspectives: cloud-based systems and digital technologies. With cloud-based systems, a key advantage is the relative simplicity of deployment versus on-premise software. Updates are much easier to deploy, and the cloud makes it easy to quickly and efficiently scale up and add more services.
“Currently, much of today’s finance work is condensed into an intense period around each month-end, with many manual entries processed at this time,” says Workday’s Larson. “Intelligently automating core transactions and processes will address this inefficient working pattern and help ensure entries are posted correctly the first time, removing the need for a high degree of manual intervention. A good example of this is machine-learning–enabled anomaly detection, which will identify potentially anomalous transactions and automatically correct coding or surface them for review before the entries are posted.”
Security is also a strength of cloud-based systems, enabling enterprises to leverage expertise rather than develop it. Many vendors are strategising long-range plans to offer only cloud-based solutions – a major trend for companies looking to invest in technology for the finance function. For people driving transformations within finance, the increasing availability of cloud solutions is a significant opportunity.
Delivering Faster, Deeper Insights
While finance will get a huge boost by intelligently automating the processes mentioned above, it is the ability to meet the increased demand for insights, reporting and analysis– as well as the rising volume and complexity of data required in near real-time from key stakeholders – that will realise the most positive impact by intelligent automation. In fact, 26% of organisations in a global CFO study said that their primary reason for implementing automation within the finance organisation was to provide enhanced decision support that will make their teams a more strategic part of the business.
As embracing digital technologies to drive greater automation becomes ingrained in the finance function, manual data gathering, consolidation, verification and formatting will disappear. Today, these non-value-added tasks are enormously time-consuming, leaving the finance team little time for analysis. And, as manual, routine processes become more automated, finance teams will be able to focus on value-add activities, such as scenario planning, risk assessments, performance and predictive modelling.
“With new sources of data come new analytics techniques and a search for insights. Organisations will apply their usage of automation and data mining techniques over planning, delivery and outcome data to enhance visibility and tracking across those processes,” says Jason Byrd, partner, technology business management, KPMG. “New insights will allow teams to capture timely data to analyse velocity, deployment and customer response, creating a feedback loop of decision-making and course correction.”
While there is hope that the global disruption driven by the pandemic will start to dissipate in 2021, finance leaders must ensure they are prepared for anything – and that means finance must embrace intelligent automation to maximise the efficiency of its available resources.