Is a capture rate of 99.5% accurate?
Before Capture Rate…
Before we talk about capture rates we should first discuss how the information from an invoice is entered into a finance system. There are many options but the most common three being manual entry by an accounts payable clerk, outsourcing your invoices to be processed off site and a finance automation solution. Each have their benefits and pitfalls as we’re discussing the importance of capture rates and process management, we will be focusing on finance automation solutions and the way that they capture and input the information.
How do you capture information?
OCR (Optical Character Recognition) is the most common capture technology found in finance solutions, but you will also find barcode recognition and template based intelligent capture. OCR technology captures every piece of information on the document then uses robotics and machine learning to determine what information is displayed to the user.
Template-based capture works very similarly but first a template must be created by the user, the template is created with boxed parameters and any information inside the boxes will be captured. If the location of the information is changed on the document then a new template will have to be created. This is fine if you are dealing with one or two suppliers, but it can quickly add up and the automation you were trying to achieve, quickly turns back in to a manual process.
A barcode on a n invoice works the same way a barcode would work at a supermarket, the bars in the bar code represent numbers and letters. When the barcode is scanned the information is captured and sent to a computer. The computer then makes sense of the information where it is converted for ease of view.
What are capture rates?
So now you know how the information is captured, what are the capture rates? Well, put simply the capture rate is the accuracy of the information pulled from an invoice by the OCR technology. Most, if not all finance automation solutions that use OCR can, and have claimed a capture rate of 99.5% and above, with the technology ever improving will we ever see capture rates hit 99.99%? Yes, but will it be worth the cost? Is it the most important part of the automation process?
The effectiveness of process management and rules.
Arguably as import as capture rate, is the process management of the information . How the solution automates the information received through the capture piece. It’s all well and good that you can capture information from an invoice but what do you do with that automation? Most solutions if not all use a combination of OCR, robotics, machine learning and AI.
An important factor in process management are the rules and workflows you have set up to indicate where a piece of information needs to be sent to for checking and/or validation, and how flexible these are for your business needs.
A good example would be an invoice coming into the business, the OCR captures the information and the solution will recognise the information from past invoices using machine learning. If the invoice is over a set amount, the rules you set in place will indicate if it must be sent for validation. If so, the invoice is then sent to the relevant person for checking and approval. Once they have seen the invoice, it will be approved and sent into your ERP to be paid.
What happens if that person is on annual leave? The invoice will just sit there waiting for the them to get back but how long could the invoice be waiting? Implementing rules means the solution would recognise that the invoice hasn’t been touched in a set time period and then the invoice would be sent to a second validator for approval. This ensures that people aren’t waiting on the invoice, they can see where the invoice is at any given time using analytics provided from the finance solution. This frees up accounts payable clerks to do more important things, rather than wasting their time talking to suppliers who are following up on the late.
Importance of a clean master data.
Another common factor that a lot of people forget about is data maintenance, this requires a person to take time out of there day to make sure all the supplier information is correct. This takes up a lot of time which is overlooked as busyness or laziness. If the data isn’t accurate this can cause the Automation rate to slow down or even grind to a holt. Most companies plan many days or even weeks to clean the data every couple years.
Depending on quality of the of OCR, machine learning, robotics or AI, the rules and processes set up by the business, as well as maintenance of the supplier master data dictates the automation rates. All solutions and businesses are different, so no process is ever going to be the same which means it can be hard for many companies to guarantee a capture rate percentage.
Is one more important than the other?
The goal isn’t to have to choose between these benefits, but to find a provider that offers strong solutions for both. If, and when, you start you to looking at automating your invoices remember not be impressed by the guarantee of a 90% capture rate but to dig deeper and find out how finance solutions use the information. How they process it, how the capture piece works with the workflow tools, make sure rules and processes are understood by everyone. When it is implemented right and understood truly is when automation can bring true benefit and allows robotics to work the mundane tasks while humans can worry about the more important things.
Look out for my next blog where I will be introducing ITESOFT’s new exclusive technology Shared Supplier Master Data, this new innovative technology “Capture as a service” or CaaS, which will help to increase the thought-put of invoices to a higher level as well as checking for potentially fraudulent activity.
In the meantime, check out: What to Look for in an Automation Solution?