The question remains: OCR vs digital invoice capture - what are you really getting?
Many organizations have discovered that their account payable processes do not work well in a “work from home” environment and have started to look at how they can digitally capture their supplier / vendor’s invoices to remove the paper and fully automate it.
The key objectives are also to improve productivity, speed up processing to ensure key suppliers are paid on time and enable more timely financial and cashflow reporting.
For basic digital invoice capture, the first step is often to look for solutions that incorporate Optical Character Reading (OCR) invoicing technologies, as this provides the capability to:
- scan a paper invoice and use the technology to “read” all of the relevant content
- convert each printed character into its digital equivalent,
- and then reformat the data so that it can be used by an accounts payable automation or accounting application.
That sounds simple enough, but is it?
If you take a few minutes to grab a handful of different supplier / vendor invoices and look at how the information is presented on each, you will start to understand the challenges. Almost every supplier/vendor invoice is fundamentally different in what information is provided and how it is laid out – and that is just for the invoice header information. When you start to include the invoice line-level information, the number of permutations becomes almost infinite.
To use OCR invoice capture, the very first step is to teach the OCR software where the relevant information is on the page for each and every supplier/vendor your business uses. Most OCR applications aid this process by taking the first stab at the process and then present the image of the invoice alongside the OCR-generated digital data to an operator to check.
This check is in two parts: the first is that the information is correctly identified, e.g. supplier, invoice number, invoice date, purchase order number, goods, tax and total values. Once that is done, then the second check is that the digital information created is accurate, as getting one character wrong (e.g. invoice number) will cause a lot of problems later on. If the data extraction includes the invoice line-level information, the time to check each invoice is more than doubled.
The reason that this level of validation is important is because even the best OCR invoice capture solutions available can only claim a 95% - 97% accuracy, so a manual check is fundamental to ensuring 100% accuracy.
The OCR provider will address this need for validation in one of two ways: either they will undertake the validation process themselves, usually using offshore labor (commonly in the Far East as it is considerably cheaper), or simply present the information to your accounts payable staff to undertake the checks themselves. Usually, the price you pay per invoice processed will reflect which methodology is used and also whether you just collect invoice header, or invoice header and line information.
Once the content is approved, the invoice can be coded for the general ledger. As most OCR invoice capture applications only extract the invoice header level details, any sub-analysis of the invoice value across multiple departments or at invoice line level has to be manually entered.
This entire process has to be repeated for every invoice processed, as each invoice is unique and often the level of accuracy achieved is a reflection of the quality of each paper invoice processed and the quality of the subsequent scan.
Once all of this has been completed, and depending on the functionality available within the accounts payable automation application, the invoice can matched against orders and deliveries, sent for digital approval, and/or just posted directly into the accounting / ERP software.
OCR has been around for a very long time, it does a job of work and for most businesses and it was faster than manually posting the invoice.
2020 – the world has changed!
If you ask any accounts payable professional, they will tell you that the vast majority of invoices they receive are now sent by email as a PDF, as this capability is now standard within almost every accounting / ERP application. So instead of printing them out and posting them, the invoices are delivered almost instantaneously for free… and the process is green.
Given the opportunity, accounts payable professionals will also explain in some detail how annoying this is, as they now have to print every one of these invoices out first and then scan them so they can be read by the OCR software. This additional workload has removed most of the time savings OCR offers in the first place – not only taking them back to square one in terms of productivity, but also with the additional cost of labor and ink to produce the paper copy.
What they probably don’t know is that same emailed PDF invoice is actually a digitally rich document (containing all of the data required to provide the human readable version), nor the huge commercial implications this simple fact has on automating the accounts payable function.
The sole purpose of OCR is to generate the digital data and put it into a usable format so that it can be checked before processing.
But if all of the data is already available within the PDF, OCR is now a costly, time-consuming, inaccurate and totally pointless technology.
Digital invoice capture
The technology to enable the accurate extraction of PDF invoice data has also been available for some time, what has changed, however, is the price: it’s now available to every business of every size.
The benefits of digital invoice capture are many. The supplier of this service usually undertakes the supplier “onboarding” process to accurately identify the information. Using advanced AI and machine learning, the process is extremely fast and 99.99% accurate, but with the added benefit that once each supplier has been onboarded, every subsequent invoice from that supplier (irrespective of the customer) is processed automatically, without touch. Furthermore, and just as important, the inaccuracy of OCR trying to read every character and generate the correct digital character is removed. Data is data and is always 100% accurate!
The invoice capture process becomes so simple. Just ask your suppliers to email their invoices to a nominated email address or simply forward their emails on after you receive them.
With this process, the invoice is captured and digitized (with the PDF, a copy of the originating email and any associated email attachments also included in the e-Invoice bundle) and automatically delivered directly into the accounts payable automation application invoice register, ready for processing.
Digital invoice capture also provides far greater data extraction. This can include the supplier’s bank information, which can automatically be validated against the data within the accounting / ERP software to reduce fraud.
It is also possible to extract key procurement information, including all line-level data – this includes product descriptions, units of measure and unit prices and quantities. It revolutionizes digital procurement, but that is another story.
Find out more about digital procurement
Digital invoice capture is simply faster, more accurate and considerably cheaper than any OCR solution.
There is still a requirement for OCR for the ever-diminishing number of paper invoices, but if you want the best possible experience, at a much lower price and with far higher productivity for your accounts payable team, digital invoice capture is the future for your business.
Want to find out more about our digital invoice capture (also known as e-Invoicing) and how it can transform your business’ accounts payable? Check out our guide!