By Markus Noga, Senior Vice President, Head of Machine Learning, SAP and Martin Naraschewski, Vice President, Head of Finance and Risk Solution Management, SAP
Digital transformation is a phrase that businesses have been using for years. While it feels like we’ve been hearing about digitally transforming for quite some time, the reality is that finance organizations are only experiencing the first impacts of what will be a massive change in the way they work. As the world transitions to the digital economy, there will be a tremendous positive change in the way finance performs its function.
As enterprises become more automated and intelligent, finance is expected to become the knowledge hub of the organization, not only reporting but also simulating outcomes and predicting results. To achieve this, finance leadership should be spending a lot more time in business value creation and driving business strategy rather than providing inputs based on historical numbers.
Thanks to digital transformation, finance is now inextricably linked with (and dependent upon) technology to fulfill almost all objectives around operational efficiency, leadership, and fiduciary responsibility. Machine learning is further accelerating this digital revolution by allowing companies to extract insights, streamline reporting and fuel forecasting without having to program computer systems and set manual parameters. Finance and accounting teams leveraging this technology are now learning to operate intelligently and are moving outside of their transactional role to support timely business decisions with robust planning and simulated functionality.
Machine learning allows organizations to access, analyze and find patterns in Big Data in a way that is often beyond human abilities. While the algorithms that enable machine learning have been around for decades, advances in computing power, data availability and data accessibility are making it possible to incorporate machine learning in more enterprise functions. For accountants, this means simplifying day-to-day activities such as the payment clearing process, so they can focus on tasks where the machine learning models can’t provide exact matches.
WANT TO BUILD A FINANCIAL EMPIRE?
Subscribe to the Global Banking & Finance Review Newsletter for FREE Get Access to Exclusive Reports to Save Time & Money
By using this form you agree with the storage and handling of your data by this website. We Will Not Spam, Rent, or Sell Your Information.
Machine Learning in Action
To best illustrate how machine learning can support finance in tasks such as payment clearing, let’s take a look at energy management company Alpiq. Its finance team has been using a traditional rule-based approach for the payment clearing process, making maintaining rules a challenge as they are constantly changing. By moving to a single integrated environment that learns from accountants’ behavior and leverages both historical data and existing AR workflows they will be able to amplify the accounts receivable process by matching incoming payments to open invoices. With the help of intelligently extracted information from payment advice documents and historical data records of successfully matched payments and invoices, they will be able to reduce manual efforts, error-prone and repetitive tasks.
With advanced technology that enables innovative strategies and growth plans, accounting teams have a competitive differentiator that empowers better compliance and reduction in costs that can steer an enterprise toward success.
Overcoming Mundane Clerical Work
Many accounts receivable (AR) departments struggle to clear invoice payments in situations where customers pay different amounts, do not include sufficient remittance information, or combine multiple invoices in one payment. To clear the invoice, employees must either manually add up various invoices that might match the payment amount, or contact the customer to clarify. Also, an increasing volume of electronic payments means de-coupled remittance information sent separately. In the case of short payment, an employee can either ask for approval to accept the short payment or request the remaining amount from the customer.
With machine learning technology, employees can automate the clearing of payments and receive proposals to match incoming electronic bank payments to open receivables. With the technology, incoming payments are either automatically cleared or a short list of possible clearing matches are suggested that an employee can quickly investigate. Automating this process provides the flexibility needed to move beyond the process of matching and reconciliation of heterogeneous data for more rewarding and higher-value work.
Reducing Costs through Advanced Automation
The traditional rule-based approach to the payment clearing process has become challenging due to constant format changes and the addition of new payment methods. Maintaining rules effectively requires relentless maintenance and excessive costs. This is where using machine-learning-enabled solutions in payment clearing can be advantageous. The technology seamlessly adapts to changing conditions, as it is constantly learning from accountants’ actions, capturing much richer detail of customer and country-specific behavior, without the expense of manually defining detailed rules.
Accounts receivable employees have access to an integrated environment that leverages both historical data and existing workflows – with minimal maintenance required. Leveraging machine learning technology, it adapts to changes automatically, processing incoming payments faster, reducing days sales outstanding (DSO) and improving customer service. Instead of manually reviewing months of spreadsheets, self-learning algorithms can find patterns and solutions in data to make decisions easily, and with confidence. Shared service teams no longer need to spend time updating payment rules and regulations. This freedom enables them to process higher transaction volumes, focus on strategic tasks, and scale with the business by delivering insights and informing decision-making on demand.