OCR Streamlines Paper Processes

Many businesses believe they can solve all their paper problems simply by installing a few document scanners and converting physical files into electronic images. This is flawed thinking. While electronic images are easier and cheaper to store, getting paper into an electronic format isn’t always an easy task. Scanning can be a full-time job that requires a great deal of document preparation to ensure an optimal image is captured. In fact, establishing an internal imaging infrastructure often results in added overhead costs for dedicated scanning labor. Furthermore, while imaging offers productivity gains by providing users with simultaneous access to document images via a PC, these benefits are limited. Sharing the data contained within these document images is still a process wrought with manual data entry if other technologies are not integrated with the imaging system.

DON’T SETTLE FOR LIMITATIONS IN YOUR DOCUMENT IMAGING SYSTEM

PSCU Financial Services realized the limitations of its own imaging platform in early 2001. PSCU serves its credit union clients by mediating between them and credit and debit card companies. Members often call the PSCU contact center to challenge charges on their credit card bills. These disputes can often be handled over the phone, but more complicated ones require the cardholder to submit documentation to PSCU’s case management department including a dispute form issued by the financial services company, a copy of the credit card statement, and corresponding receipts. In addition to dispute resolution, PSCU’s case management group also handles other document-laden processes including address changes, fee renewals, and account closures.

The case management department began scanning all of these incoming documents and storing the electronic images in a system PSCU developed in 1995. This platform was initiated in an effort to reduce the volume of paper documents it needed to handle. Unfortunately, the process still required a lot of document preparation (e.g. adding separator sheets between document batches, manually sorting document types, and manually orienting pages in the same direction).

“Five full-time employees were needed to manually presort all the incoming documents into the proper work queues,” says Lisa Howden, software analyst for PSCU. “Once sorted, each dispute we received had to be scanned with a separate cover sheet and a specific bar code separator sheet.” Once dispute documents were scanned, data entry operators still had to manually key critical data off each image into the PSCU database. Once processing was complete, each dispute was sent to a PSCU work case representative for research and resolution. The problem was the scanning process required the case management team to work a great deal of overtime just to meet the three-day customer response time the company was contractually obligated to provide its credit union clients.

ELIMINATE MANUAL DATA ENTRY, DOCUMENT PREPARATION WITH OCR

The case management department needed to eliminate many of the manual data entry tasks that bogged down its scanning process and began to investigate technology options that would automate these steps. Three software solutions were evaluated, but in 2002, PSCU selected OCR for AnyDoc by AnyDoc Software to enhance its imaging efforts. OCR for AnyDoc is designed to locate and capture critical data off document images whether that data is presented in machine-print, handprint, mark-sense, or one- or two-dimensional bar codes.

Now, after PSCU’s case management department scans its dispute documents, OCR for AnyDoc extracts vital data such as account number, dollar amount of the dispute, and the date the disputed item was posted to the customer’s account. Data that the OCR system deems correct then automatically populates corresponding fields in the PSCU database, eliminating the need for manual data entry. Meanwhile, questionable data captured by OCR for AnyDoc is presented to a human operator for verification to ensure data accuracy.

The OCR solution also helped eliminate much of the manual document preparation work with which the case management department was burdened. “Our OCR solution can automatically identify document types by extracting and identifying keywords from the first page of the dispute form,” says Howden. “We no longer need to sort documents or attach cover sheets or separator sheets to each batch of dispute documents. Plus, our OCR software comes equipped with an automated image rotation feature that keeps us from having to ensure that each document is facing the same direction before inserting it into the scanner.”

Each of these attributes of OCR has saved PCSU time — and in the business world, time is money. Prior to implementing its OCR solution, the case management department consisted of 78 full-time employees. Now, the group is producing the same amount of work with only 63 employees. Furthermore, the staff of five that was once necessary to perform document preparation tasks has been reduced to two thanks to the OCR solution. The other three employees have since been trained to perform more productive duties elsewhere in PSCU. Since PSCU hasn’t had to replace lost headcount due to the automation benefits its OCR solution provides, the company now saves approximately $125,000 annually (when compared to the previous system) in labor costs alone.

The case management department has further streamlined the dispute resolution process by adding AnyDoc Software’s FAXit module to the system. FAXit allows data from faxed dispute documents to be captured without a physical printout of the fax ever being made. The electronic fax transmission is simply fed directly into the OCR engine, where critical data is then extracted.

USE OCR TO ELIMINATE POINTS OF HUMAN FAILURE

The success the case management department experienced with OCR has prompted other departments within PSCU to adopt the technology. For example, the company’s accounts payable (AP) department is currently implementing the AnyDoc INVOICE module to automate invoice processing (see sidebar on page 18). 

“The main reason we’re implementing OCR technology is to reduce the chances for manual data-entry errors,” says Jeremy Starling, director of accounting systems for PSCU. “For example, miskeying an invoice number is a common human error that potentially can have very costly consequences. If an invoice number is miskeyed, it is likely that an AP representative will issue a duplicate payment for that invoice, which can cost a company thousands of dollars. OCR eliminates the human element and improves overall data integrity.”

In its initial implementation of OCR software, PSCU’s AP department will scan each invoice on the back end, after the paper document has gone through a paper workflow and approval process. After scanning, each image will go through an OCR process that will extract vital invoice data that will be used to create an image index. This index will initially be used to quickly retrieve invoice images internally. However, future plans are to try and make these images and the invoice data available to external vendors via an Internet portal, so these parties can visit a PSCU Web page and inquire about the status of specific invoices.

The AP department also has plans to eventually scan invoices and launch the OCR process on the front end as invoices are initially received. This will allow PSCU to use captured data to automate electronic workflow and approval processes and further enhance productivity and accelerate turnaround times.