In the past decade many businesses have undertaken measures to digitally transform their operations. As companies move away from outdated analog processes, an acute pain point has been realized: paper does not scale in a new fast-paced digital organization.
Paper is expensive and slow, creating manual labor, and bad for the environment. To deal with the paper problem, some organizations have turned to scanning invoices with Optical Character Recognition (OCR), only to encounter the same issues, but in different forms. In this post we’ll examine the problems with OCR, and how enterprise organizations can fix them with digital processing technology to achieve accounts payable automation.
Let’s start with the basics. OCR is a technology that scans paper documents into a digital image file, then uses a character recognition program to convert the image into digital text. Tracing its origins all the way back to 1974, OCR hasn’t changed much in the past couple decades. Scanning has gotten quicker and character recognition has become more accurate, but the underlying technology is still the same.
For AP departments OCR was a boon during the digital transformation process, enabling paper-laden teams to begin digitizing and indexing all of their files. Many ERP systems still bundle OCR software with their offering because it became so ubiquitous in the early to mid 2000’s. However as the speed of business has increased, and the needs of the leading organizations have evolved, the once useful OCR technology has become a liability.
OCR has become a liability to businesses because it is an imperfect technology. The character recognition programs that power the image-to-text digitization often return incorrect information. Small font size, bold, italics, colored paper, handwritten text, blurry or dark copies and strange formatting can all contribute to incorrect assumptions by OCR software. However, even in ideal scenarios where the scanned image is at the correct resolution, perfectly straight, and perfectly lit, the technology is still only (best case) 95-99% accurate. In many use cases, 99% accuracy is acceptable, however, for AP professionals and their colleagues in finance or procurement departments, even one wrong number can cause huge issues.
Many AP departments began leveraging OCR because they wanted to reduce the amount of manual data entry required to process invoices. However, because of the limitations with OCR technology, the resource burden has shifted from manual data entry to manual exception handling. When OCR scans an image and outputs incorrect information, the invoice in question is automatically marked as an exception and must be reviewed. Plus even in the rare situation that the OCR conversion works perfectly, eliminating exceptions is near impossible because there is no electronic connection between AP and the supplier. All resolutions for incomplete invoices, missing fields, invoice to PO discrepancies, or any other issues must be resolved over the phone or email.
62% of AP professionals report that invoice exceptions are the biggest challenge that they face on a day-to-day basis*. Invoice exceptions cause so much difficulty because they must be manually resolved. Resolution often requires the AP professional to dig up the original invoice to see which characters were improperly translated, or reach out to the supplier for a new invoice resulting in lengthy email chains or phone tag to determine the issue. This manual effort slows down AP and makes it impossible to process invoices straight-through. To learn more about the benefits of straight-through invoice processing, click here.
Even in 2020, AP professionals report that 50% of invoices their invoices are still sent manually, via email or postal mail*. OCR causes many of these invoices to become exceptions, but the good news for AP and Procurement professionals is that there is a much more efficient way to process incoming invoices that virtually eliminates invoice exceptions.
Enter true electronic invoicing (or e-invoicing). E-invoicing removes the issues of OCR because invoices are transferred via end-to-end digital, meaning there is no opportunity for information to be degraded or misinterpreted. Instead of an invoice arriving via postal mail as a piece of paper, or by email as a PDF, all invoices arrive digitally so they can be immediately processed without any manual intervention. Many AP professionals may be thinking, “that sounds a lot like EDI,” and you would be correct, with one big difference. Typically EDI integrations can be incredibly complex and resource intensive because they require a custom data integration. Each integration is unique to each supplier because each data field must be mapped individually.
However, Modern digital platforms like Transcepta are purpose-built to quickly and painlessly connect any supplier regardless of EDI capabilities. These modern technology platforms also come standard with a service component that connects and onboards all suppliers to a supplier network without additional effort from the enterprise. The supplier network enables instant two-way digital connectivity, ensuring that all invoice data is received and processed automatically with no scanning or conversion. Plus because there is a digital connection, invoices can be matched to a PO or to other required documentation, and any discrepancies that require resolution can be automated without AP intervention. In other words, no AP exceptions.
For suppliers that insist on paper or PDF invoices, Transcepta has a unique ability to convert PDF invoices to digital without using OCR. You can read more about this in this short whitepaper.
Digital invoice processing has many benefits, both to the customer and its suppliers. Whereas OCR relies entirely on the ability to reliably convert images to text, digital processing simply transfers data from one system to the other. This process unlocks some truly transformative benefits:
First and foremost, digital processing is 100% accurate. Because invoices are end-to-end digital, there is no opportunity for data degradation or human error. Because of this accuracy, invoice exceptions can be drastically reduced or completely eliminated.
Invoice data arrives from suppliers “ready to pay,” meaning no manual intervention is needed from the AP team. Using OCR the invoice would need to be reviewed to check for errors, or marked as an exception which would trigger a lengthy resolution process.
The result of fast and accurate processing is reduced costs. Less time and resources are required to check for errors and handle exceptions meaning less full time employees are needed to process invoices. Reducing invoice payment times can also unlock supplier discounts which help to offset internal costs and increase profit margin.
If you are questioning the utility of OCR, there has never been a better time to explore the benefits of digital invoice processing. Many ERP systems have OCR built-in, or offered as an add-on service, but they come with the same limitations that create obstacles to straight-through invoice processing. Platforms like Transcepta integrate with your ERP and can help your team eliminate invoice exceptions and enable accounts payable automation. If you would like to learn more about the benefits of straight-through invoice processing, click here. You can also request a demo of Transcepta by clicking here.