The perceptions and fallacies of OCR and e-invoicing
Perception is reality right? In terms of how people make decisions and form opinions, yes, perception is reality. And that’s why it is important to question what is perceived to be the truth – because getting it wrong could cost you dearly.
Here are just a few of the fallacies about AP automation that lead some organizations to make the wrong decisions:
Electronic invoicing is expensive
It is often said – unfairly, of course – that accountants understand the cost of everything and the value of nothing. Sure, there’s a cost to e-invoicing. There are often transaction fees and network membership fees. There may be costs to suppliers as well as buyers. And then there’s the implementation and change management costs. Whether you look at the costs in terms of hard cash, or in terms of time and effort, there are costs. Like any business case though, taking the costs in isolation only illustrates one half of the picture. These costs need to be compared to the costs offset by moving away from a paper process. That process isn’t free.
e-invoicing is unfair to suppliers
The view of many organizations who started e-invoicing programs years ago, was that the only way to make the business case for e-invoicing is to make the suppliers pay for it. It was symptomatic of the rush to embrace the digital world, and in the enthusiasm to implement B2B techniques and technologies, the bigger picture was sometimes overlooked.
Imposing big costs on suppliers is indeed unfair but there’s a wide range of options available now. There are free-for supplier models out there and even some of the networks that do charge, the costs are often in line or less than the equivalent paper and postage costs.
The supplier networks do throw up some anomalies, however. For some suppliers with relatively low invoices volumes, combining the costs of transaction fees and network fees can make the cost per transaction inordinately high – in some cases, quite ludicrously high.
OCR is highly accurate
In an age when we can identify people in fractions of a second by scanning their retina, of course the best OCR technologies are accurate. The best platforms available today can clean up dirty documents and even work in a mobile environment. Sales teams will claim close to 100% accuracy and, while we can allow them some license to exaggerate, these are largely accurate claims. But we need to be careful what we mean by “close to 100% accuracy”. Does that mean that nearly all of your invoices will be scanned at 100% accuracy? Or does it mean that 100% of your invoices will be nearly accurate. That makes a big difference.
And then there’s the problem of trust. If the accuracy isn’t 100% then there’s no confidence in the electronic document and many organizations simply don’t trust their scanned documents enough to not review every single one.
It is possible to get OCR set up to deliver good results by scanning and interpreting accurately but this will often involve a great deal of effort to “train” the system and compromises are often required, such as dealing with invoices at header level only.
The number of AP staff can be reduced by scanning invoices rather than manually keying them
This is no fallacy. Of course, by scanning invoices, time is saved by not having to key invoice data, but while the number of AP staff might reduce, the total number of staff dealing with invoices may not. A scanning and OCR setup doesn’t run by itself. The format of paper invoices is always evolving as new suppliers are brought on board and there is a level of ongoing maintenance required. If you reduce the headcount in AP only to see it increase in order to support maintenance, it delivers a zero sum result – or worse.