Once an enterprise has its new document imaging software and scanning equipment up and running, it’s business as...
usual, right? Wrong. The only way to maintain the investment of time and money in the imaging and scanning system is by constantly revisiting it.
But few organizations continue to benchmark the performance of their document imaging systems or fine-tune them after the go-live date, and that can be costly, said Alan Pelz-Sharpe, a principal analyst and director at Real Story Group, a content management consulting firm based in Olney, Md.
“In the world of capture, [program success] is highly dependent on how you scan the document – no two implementations are the same, so there’s no way to say how long it will take or how complicated it will be,” said Chris Riley, senior enterprise content management and document capture architect at consulting firm ShareSquared Inc. in Pasadena, Calif.
Companies would do well to set goals for getting a return on investment (ROI) and reducing the costs associated with working on and maintaining paper documents, according to Riley. “The ROI could be to process an invoice in a day versus two weeks,” he said. “Determine how much money you want to save by using this technology.” But any ROI target should be set high, he said, especially if the imaging and scanning technology is being used to help eliminate manual data entry and reduce staff.
Plan for exceptions with document imaging software
Something that’s often neglected after implementation is the reasonable expectation for exceptions – documents that don’t conform to predefined rules and have to be entered manually.
“This is the destroyer of capture projects,” Riley said, adding that optical character recognition (OCR) and document conversion tools “are interpretive technologies, so they’re never 100% [accurate].” In addition, the core accuracy level of these technologies is already at or near its peak, according to Riley: “It’s not going to get much better, so companies need to consider the exceptions.”
Many companies start to see an ROI right away from automated document imaging management processes, and there’s a window where they’re enthralled with the technology, Riley said. But then an invoice or bill of lading comes in that breaks all the rules that have been set up in the imaging system, and the company has to start over.
“What happens is that they get an immediate ROI with the technology, spin several sessions and add exceptions to the [imaging and capture] logic, and then lose the ROI,” Riley said. “They have to be willing to process some oddities manually.” By modifying the system to handle the exceptions, “you jeopardize all the logic you have,” he added. “It’s as much of a science as an art.”
As part of the implementation process, it’s important to determine what percentage of exceptions you are willing to accept. Riley recommended that companies examine all the types of documents they intend to scan and create logic for 80% of the volume.
Tracking and ensuring document imaging success
In document imaging and capture, accuracy is paramount – so it’s important to set up and implement imaging and document scanning systems with anticipated levels of accuracy and the metadata values you want to capture. For example, the values might be invoice number, total cost, purchase order number – whatever is pertinent to the particular workflow.
“It’s all based on strategy and the definition process of the workflow and the document,” said Alan Weintraub, a principal analyst at Forrester Research Inc. in Cambridge, Mass. “It’s not a one-size-fits-all. That’s part of the up-front definition process you go through when you’re defining the system.”
In addition, determining and adhering to checks and balances along the way can help to ensure that the information is readable and indexed correctly. “Companies often take the time to define the metadata, but they don’t put enough checks and balances in to make sure they’re reaching that quality level,” Weintraub said. “You need to make sure you can find the documents.” An internal quality assessment process can help ensure that the document management team or external imaging and scanning provider maintains the service level a company requires.
“Keep in mind that in some OCR implementations, making changes may demonstrate improvement in one area and detriment in another,” Riley said. “The most likely case is settings for one document type are not working well for another.”
Finally, “always refer back to your goal list to see where your production is and what you would like to improve,” Riley said. Once a document imaging system is up and running, he added, close monitoring, soliciting feedback from the imaging team and fine-tuning the system as needed can help enterprises get the ROI they were expecting on an ongoing basis.
ABOUT THE AUTHOR
Catherine LaCroix is a freelance writer based in Portland, Ore. She covers technology used in business, education and health care.