A new business opportunity has emerged for banks that provide corporate cash management - the ability to offer data clean-up and data management services for transaction information that can be muddled and inexact.
"The data is often dirty and has inaccuracies that need to be repaired before a firm can make sense of it," says Neil Vernon, development director for Gresham, which is supplying the technology behind an initiative by Australia and New Zealand Banking Group (ANZ:ASE) to help corporates reconcile and clarify the data behind large volumes of corporate payments.
Data reconciliation for payments has long been a problem for corporations with multiple suppliers and vendors, and is a point of entry for bank marketers. Invoices and billing documents are often filled out the wrong way, lack full identification or have incomplete information. ANZ is using image capture and a business intelligence rules engine to correct and standardize payments information and documentation with as little human intervention as possible.
The bank, which is using software from Gresham called the Clareti Transaction Control (CTC) system, has launched a pair of new products called ANZ Cashactive Control and ANZ Cashactive Virtual, to aid in cleaning up data for cash management and visibility purposes. Cashactive Control covers compliance for accounting, legal, property, and government issues, while Cashactive Virtual addresses the management of company funds, assisting organizations with cash flow management. CTC is Gresham's reconciliation and compliance flagship, which allows business controls to be used to digitally correct documents and transaction records.
"For today's treasurers and CFOs to have accurate and timely visibility and control of their cash position, they require accurate and timely information, as well as efficient, flexible and reliable transaction processing. It ensures that cash is in the right place at the right time with the right information," says Anne Collard, global head of product and channel management, payments and cash management, for ANZ Institutional and International Banking.
The new ANZ products capture invoices and other billing documents, then use CTC to analyse the information in the captured invoice, comparing it to other, similar invoices that have been paid or received by the corporate client. A rules-based engine determines if the new invoice is "too different" from other invoices, which would require human intervention to either reconcile the invoices manually or check other details. "The incidence of needing a human to manually match the data can be drastically reduced," Vernon says.
Vernon says invoices often come in with partial invoice numbers, or other missing information that can be easily reconciled and automatically populated or edited to place that invoice in line with corporate and regulatory standards. "There are lots of clues that a human looking at the data would know based on past bills from that supplier. CTC has that [logic] in its rules engine."
At one client, Vernon says a staffer named "Jen" always signed invoices with her name, rather than an invoice number. "There were payments of 100 Australian dollars signed by 'Jen.' CTC was able to 'learn' by looking at other information that Jen was from the same supplier, and was able to identify the 'Jen' bills as a match. It's a pretty straightforward process, but the manual cost of matching a thousand invoices to payments can be a large and arduous process."
Large corporations are increasingly centralizing treasury and finance operations, a practice that at some firms had been handled on a department or regional basis. Collard says this trend, along with tougher regulatory requirements demanding strong risk practices, have corporates increasing their expectations regarding the quality of reporting and information. "There are two key dependencies to achieving all of this. First, organizations need deep integration between bank data and the systems they use to manage their business and financials. This is commonly referred to as straight through processing. Second, it requires reconciliation - or more precisely automated reconciliation," Collard says.