With its global presence, Springer Nature handles invoices in many di fferent currencies, not to mention di fferent VAT rates. Prior to Xelix, it relied on two invoice checking systems, plus manual checks.
Springer uses SAP, which only identifies basic duplicates, so the business built its own tool in house. Springer Nature’s outsourced BPO team used the tool weekly, and it would pick up duplicates after processing payments. The problem was it needed constant development. Aside from system processes, there was a human checking element too.
“Whenever we had a payment run, many people reviewed it for errors. They’d check for currency errors and invoice numbers that didn’t look normal for a certain supplier. But when you’re processing the level of invoices we handle, errors are inevitably going to slip through unnoticed. For duplicates created on a subsequent day, for example, this was always true. If our vendors spotted a double payment, they’d contact us. In fact, that became a common duplicate checking method when our tool didn’t pick it up,” says Duncan.
If the basics were already in place, why not develop the in-house tool further? Springer Nature had a development team at the ready, so this might have been an eff ective route towards more accurate checks. Duncan explains why that's not the case: “Building your own tool is great because you can focus on what’s important to your specific business. You can also exclude what you don’t need. But we also discovered many limitations. To remain effective, any finance tool needs constant maintenance and development. And yet, just a small number of people hold your in-house expertise. What if one leaves? What if there’s a higher priority project to focus on? We could have developed our tool’s capability further, but it would have taken resources needed elsewhere. So, I wanted to compare our tool with a system we could buy into, and use ourselves.”
Another frustration with Springer Nature’s in-house tool was it could only check for duplicates after invoice processing. That meant having to recover money already paid, which is always harder than preventing the duplicate in the first place. Duncan could have chosen to engage third-party service providers to identify duplicates, but he pointed out they’d still need to revert to his team with their findings and that would drain just as much time as doing it themselves.
