More of the same won’t speed up invoice payments

Many of the most progressive businesses are enthusiastic about putting an end to slow supplier payments.

Many of the most progressive businesses are enthusiastic about putting an end to slow supplier payments. However, says Paul Christensen, CEO and co-founder of Previse, the archaic nature of current payment practices mean speeding up existing processes is a futile task. Firms need to completely rethink how their suppliers get paid to meet their ethical obligations.

Business leaders will be well aware of the mounting pressure to speed up invoice payments. It isn’t only the fear of negative headlines or the requirements of the prompt payment code, there is increasing awareness of the business logic of cracking down on slow payments. Many CFOs realise that, when a firm stretches its payment terms, it is destabilising its supply chain and driving up its own costs.

The desire to make change is there, but the question then turns to how to actually speed up payments. This is not so simple.

The current payments process in most firms is extremely manual, paper-based and time-consuming. A tier one bank we worked with processes 760,000 invoices per year, that’s more than one invoice every minute of every day. According to a 2018 study by Concur[1], the average invoice requires 21 people in order to be checked and approved. It requires six people to raise an invoice, a further six to approve it and then nine people to process an invoice on the back end.

Invoices require such careful consideration because 21% of firms have received fraudulent invoices and a further 41% have received duplicate invoices. Despite firms’ best efforts, 33% of firms admit to having paid a duplicate invoice.

Alleviating some of the burdens of invoice processing through basic automation would also be difficult. Currently, 57% of invoices are submitted to firms by post (this is fairly consistent across businesses of all sizes). Of those invoices submitted electronically, most are a PDF attached to an email which, for most simple automation processes, is barely more useful than a paper invoice. This is because, without very advanced artificial intelligence, it isn’t possible to take the unstructured data in a PDF invoice and convert it into a machine-readable format or automated processing.

To enable some automation in the invoicing process, many firms have turned to technologies such as e-invoicing. An e-invoicing system records invoice data in a standardised format which allows for greater levels of automated checking and can significantly speed up processing times, as well as cutting down on mistakes.

However, these systems have been met with mixed success. The complexity and cost of including suppliers in e-invoicing are significant. One study estimates the cost of onboarding a single supplier to an e-invoicing system is $925, requiring 17 hours of work and taking 16 days to complete. We are also aware of recent cases where suppliers were asked to pay over $4000 to join e-invoicing platforms. Given these costs, it is hardly surprising that the vast majority of e-invoicing systems are processing invoices for only the largest few suppliers. While these may be the invoices with the highest value, it leaves a very long tail of invoices which have to be processed manually.

Given these challenges, it is perhaps less surprising that firms are struggling to speed up their supplier payments. Increasing payment speeds by simply speeding up internal processes would likely take a vast outlay of additional manpower while automation is difficult to implement for all but the largest suppliers.

It is a cruel irony, however, that those suppliers which are left out of attempts to speed up payments are those which are most in need. It is the smallest suppliers which often have the weakest cash flow and are, therefore, least able to sustain slow payment terms. Not only that, but smaller suppliers are also often paid later than their larger counterparts. Our analysis shows that buyers pay their smallest suppliers, on average, 57 days later than the largest.

With hundreds of thousands of businesses across the world going bankrupt each year as a result of slow payment and Governments from the UK to Australia adopting increasingly tough stances on the issue, the current situation is clearly unsustainable. Nor is it a status quo that business leaders wish to maintain. However, as shown, speeding up payments by simply increasing investment in current processes, or by adopting technologies like e-invoicing, are not economically viable solutions.

We need a radically different approach if we are ever going to make a meaningful change to the way SME suppliers are paid. Fortunately, new advanced artificial intelligence (AI) coupled with powerful cloud computing, mean that it is now much easier to deal with the kind of unstructured data in paper and PDF invoices. This opens up real opportunities to process those invoices using automated processes without requiring suppliers to adopt expensive new technology or, in fact, change the way they invoice at all.

AI can process thousands of invoices before the human worker has poured their morning coffee. With these new technologies, we have an opportunity to completely rethink the way invoices are processed and, finally, bring an end to slow invoice payments.

[1] Concur Supplier Invoice Benchmark Report: Exploring the AP Landscape in the UK2018

Related articles