Write Off Indexing Costs

Pete Stanis, technical service specialist at CUNA Mutual Group (Madison, WI), describes document imaging as his company's "field of dreams." Once it built an imaging system for its claims processing department in 1998, this $9.3 billion financial services provider had more than 30 other departments clamoring for the technology. By 2002, more than 4 million documents a year were being captured, indexed, and transmitted to ImagePlus content management software from IBM, running on an AS/390 mainframe computer. Though CUNA Mutual enjoyed the efficiency and cost-cutting advantages of document imaging and workflow, indexing was one area where it could realize additional savings.

"About 99% of the indexing was manual," says Stanis. "Scanned images were sent to key verifiers who entered data in the relevant fields." CUNA Mutual realized automated indexing would cut costs, but it didn't want a solution that would be cumbersome for users and offset any savings. The proposed solution was to generate cover pages to separate and identify batches and aid in indexing. CUNA Mutual's imaging system used Captiva Software Corp.'s (San Diego) InputAccel software for document capture and zonal OCR (optical character recognition). However, the capture software required a typed and printed cover sheet. "It's cumbersome for business people to have to get up and run to a printer for a cover sheet," comments Stanis. "We felt it would be easier for our users if we provided an internally generated form which could be filled out by hand."

Single-Vendor Solution Minimizes Development Issues
To meet the requirements for reading handwritten text and recognizing document types, CUNA Mutual opted to implement form identification and ICR (intelligent character recognition) from Captiva. With a combination of several optional modules, CUNA Mutual was able to design a system that matches incoming images to known document types and lifts the indexing data using ICR, eliminating the need for manual indexing.

"What set ActionPoint [which merged with Captiva in 2002] apart from the other capture products was the method of coding that ties all of the modules together," comments Stanis. "This allowed us to create a highly customized solution for our specific needs. For example, our claims department felt their overall workflow was strong before imaging was introduced, so they wanted a solution that would meet or exceed that process without requiring extensive retraining of the 120 employees." Stanis says three IT staffers oversee the imaging system for the 5,000+ employee company and have been able to make necessary adjustments with only an occasional need for tech support or professional services.

The tight integration of the modules and the ability to customize in a reasonable period of time prompted CUNA Mutual to choose Captiva for form identification and ICR. "Working with Captiva gives us a one-stop shop for all of our software products," says Stanis. Using a single-vendor solution reduces development issues and means there is just one number to call when problem resolution is required.

CUNA Mutual also uses the routing and workflow capabilities of InputAccel. Its claims department, for example, has 14 teams located in various states. "With InputAccel, we have one integrated document and data capture solution that makes sure the data is getting to the right people," comments Stanis.

Payback In Less Than A Year
To integrate the handwritten cover sheets into the imaging system, CUNA Mutual developers completed a three-day class offered by Captiva. Within a week, the group had developed a prototype. Over an eight-month period, various departments were added incrementally. Developing a solution specific to the documents and processes of any given department required about a week of development time. Before the rollout began in mid-2001, it was estimated that manual indexing costs were about $136,000 annually. At the end of the year, the project was at about the halfway point and cost estimates had dropped to $67,000. Within 8 months, CUNA Mutual realized a 75% to 80% reduction in manual indexing costs, and the payback period was well under a year.

In the future, Stanis foresees that his department will be taking a closer look at color scanning, multi-streaming, and increased data extraction. In the meantime, he is confident about the capabilities of CUNA Mutual's existing imaging solution. "From a technical standpoint, this application is rock solid," he says. "Nothing is more satisfying than walking into a meeting to discuss new requirements and knowing you can handle whatever the group needs and steer them to the best possible solution."