Imaging's Strange Image Problem
I am writing this the evening before traveling to AIIM in Philadelphia. AIIM—the Association for Information and Image Management—along with its co-located OnDemand show—is now North America's largest information technology show, by far. COMDEX is gone. The play in IT is all about content, document and image management.
So how come the marketplace still treats imaging like a hillbilly cousin? Imaging—or "intelligent data capture" as the marketing spinsters would have you call it—has been the "next big thing" for more than a decade. I know; I started Imaging Magazine in 1991.
Back in that day, we talked about the nagging mechanical and technical limitations that imaging needed to overcome before hitting the mainstream. No lie...we worried about things like how white-out would reflect in a scanner and how to get dust off a roller. "Best Practices in Removing Staples," for cryin' out loud.
That's because imaging was all about volume. Big piles of paper that needed to become big digital files on a big optical disc. As fast as possible.
That was then and this is now. Imaging/data capture has broken free from its mundane past to take a seat at the business-performance table. Now...all we have to do is get someone to notice.
"This isn't about scanning dead documents," says Mark Seamans senior vice president of research and development at Verity. "There's a compelling business value. If I can get paper-based information online as step one, this not only creates a lower-cost way to do something, but I can also gain a competitive advantage. I can do something in three hours that it takes my competitor five days. Many times the first step is a transaction that involves paper. But that's not the ending point."
"Imaging isn't the sexiest thing to talk about," adds Bruce Milne, senior director of product marketing for Vignette. "A lot of problems have been solved, but successful customers aren't really vocal about it. It's a stealth technology; everybody's deploying it, but not talking about it much.
"We're seeing a convergence point," continues Milne. "Companies want to get off paper, and they're looking for the right combination of things like imaging and portals and workflow. Imaging can be part of a bigger customer-care initiative and a less expensive transaction model at the same time."
Arriving at that conclusion has been a challenge for the imaging market since day one. "Imaging has been, for as long as we all can remember, a slowly growing, slowly simmering market," says Bill Priemer, vice president of sales and marketing for Hyland Software. "It never really took off, like CRM or ERP or Web content management. It never had a period of explosion, either in terms of customer adoption or widespread awareness. That's because these discrete technologies—imaging, report management, OCR, etc.—have never been pulled together and sold as business solutions. The IT organizations are hip to it, but the business side—which incidentally has to pay for it—hasn't been aware of the way these technology pieces should work together. CRM means something to the marketing side of the house. Sales force automation means something to the sales organization. There is plenty of technology behind those initiatives, too, but someone was able to articulate what the technology means to the businesspeople. In our space, we've struggled to do that," says Priemer.
"It's hard to believe, but people still come up to us at trade shows and conferences, and we show them how they can start a browser-based application, ‘put the paper in the scanner and hit submit' and the relevant data is lifted from the form and populates a database application and drives a business transaction. People still say: ‘Wow! You can do that? Show me that again to make sure I heard you right.'"
Back to School
What accounts for this apparent disconnect between the solutions that are available and the apparent reluctance to embrace them? Part of the blame lies in the educational outlets that the imaging business should have been able to rely on. We have in large part failed. Or, better said, abandoned. I should know; I have been as guilty as—maybe guiltier than—most in this regard. I advocated a broader view of document management as a central information technology theme way back when KMWorld started. The imaging marketplace, for me anyway, seemed to be concerned mainly with scan and store, making it a mostly custodial task that didn't fit my impression of what business-performance improvement is all about. When the consolidation and commoditization of imaging software and hardware tools began, it looked pretty bleak. When scanners started to show up at Wal-Mart, it was all over.
Then a funny thing happened. Data capture from paper—especially forms—began to show renewed signs of life. "In everywhere from furniture stores to the Mayo Clinic, paper is being used as a nearly real-time interface," says Seamans. "It's an untraditional view. Paper is usually thought of as a slow and cumbersome process. Not anymore. It's no longer ‘input to archive.' It's now ‘input to process.'"
Seamans continues: "An example: Car dealers have learned that customers who get as far as looking at possible payment plans are more likely to close on the deal. So the leasing companies put a simple form into the dealer's hands, right there on the lot. The turnover in that job is very high, but the training requirements to fill out a form are very low. So they take 10 or 12 pieces of information, and fax the form into a processing system and get answers back in three or four minutes. "Forms are the way that businesses connect people with business processes...that's what a form is," says Seamans. "So we need to do more than just automating the first step of scanning and lifting data off a form. We need to help with the next step: where that data needs to go as part of that business process—applying for a car lease, for example. That takes our involvement beyond just the data-entry part, but also the process that it supplies and enables."
"The intelligence built into imaging systems today is beyond what people generally know about the imaging world," Milne asserts. "It's time to refresh people's understanding of what's possible. Bar code and OCR/ICR technology can allow a workflow with disassociated people using a document to ‘meet up' at a later date, almost completely automatically. You use a cover sheet with a bar code on it, for example, fax it to a clinic, and tell them to mark up the record and fax it back using the same cover sheet. The bar code allows the document to rendezvous with the patient record. I don't think people realize this is possible today." "The invention of unstructured forms processing has opened the spectrum to a larger marketplace," says Sam Schrage, vice president of operations and international sales for AnyDoc Software, Inc. "Because we can process ‘any doc' (no pun intended, I'm assured) the door is open for solving vertical business problems at a totally different level in the organization."
Charles Jackson, founder, president and CEO of AnyDoc, also thinks that capture is being viewed in a new light...by a few. "Fifteen percent of the companies that should be automating their capture are actually doing something about it. The other 85% don't know it exists. They are amazed that we can do the things we can with forms. And when you show them how unstructured forms can be automated, it blows their minds. They just didn't know that a machine could look at an invoice and figure out all the pieces."
Bruce Milne agrees that certain portions of the market could use a good dope slap when it comes to getting the picture on imaging. "Large-scale ingestion, indexing, storage management and retrieval of electronic assets—such as "Check 21"* processing, tax rule processing, claims, underwriting, loan origination—are all business processes that are heavily dependent on paper and could do with some modernization, especially in the light of offshoring and outsourcing, where you want to make that piece of paper portable to any desktop, any time of day.
"The late adopters are just starting to get the picture. The banks have been doing imaging for quite a while, and they've been doing customer self service for a long time, too. It's the laggard industries that we're starting to see now. Insurance? Try to do a beneficiary change on-line. Or look up your cash value on a whole life policy. Or the status of a claim. It's just not available," says Milne.
"The people I talk to are concerned about things like cycle times, and the revenue cycle, and compliance with HIPAA on the healthcare side. If you say ‘imaging' to them, they know what you're talking about, but they're unclear on the applicability of the technology to their business problems," Milne continues. "But the maturity is starting to appear; it's beginning to dawn on these people (who run specific business units) how imaging technology can help them."
Getting the Word Out
Sam Schrage parses the argument into two neat bundles: "You have to look at it as ‘document processing' versus ‘document management.' On the document management side there are still long sales cycles. That's because doc management requires a workflow, and that requires a professional services component, and THAT extends the sales cycle and drives up costs.
"Document processing, on the other hand, is predictable. You can determine exactly the level of processing you can expect, the percentage of accuracy, how many operators it will take, how many servers, how many scanners...there are also many vendors out there who can do the same thing, so there are very few differentiators. That's why software developers, like us, have moved into other areas, such as unstructured forms processing, as a way to differentiate ourselves from the competition."
People in the community—the people who go to AIIM regularly, for instance—get it, but the general outreach has been underwhelming.
"The dollars flow to where the volume and the greatest amount of pain is," says Chuck Jackson. "Right now, the pain is in accounts payable and invoices. On the healthcare side, it's in Explanation of Benefits (EOBs)." Jackson points out an unexpected trend: "The emergence of distributed capture as a viable process has actually allowed centralized capture to come back—you can now decide which processes are best departmentally done and which are not. You don't have to give up your documents—for instance, putting them in pouches and sending them off somewhere for processing—which is very important for some applications. We can still process them in one location, and benefit from the economies of scale, without physically handling the paper at the centralized capture site."
The concept of image-derivative business processes is, in fact, central to today's generation of data capture vendors. "We don't sell any imaging these days without a workflow," says Hyland's Priemer. "The core for us is extracting information from a document, using that data along with some decisions the user makes, to route those documents. The integration of imaging and workflow is key. Layer on top of that the Web, so you can push those interactions or workflows outside the walls of your building, and you enter the realm of solutions—high-end Treasury management, for instance. We compare metadata extracted from the document and match it to databases, and use any discrepancy to trigger a workflow through the Web. In the old days, a discrepancy would have meant someone goes to a fax machine and then waits for another person to work the file.
"In addition to workflow and Web distribution has been the integration of image-derived data with line-of-business applications. Those are the three things that have allowed the broadening of traditional applications into today's mainstream position," Priemer says.
Is there any technical barrier that hasn't been solved? "The speed and quality of capturing the image itself has been there for a while," adds Priemer. "Where we've had to concentrate our efforts is on the efficiency and accuracy of the indexing process. That's always been painful, time-consuming and expensive. Bar codes and other metadata tags have really streamlined the process; now a workflow can begin before the document goes to ‘central scanning' even before it's arrived, and we know the data from the document will be accurate throughout the process when it catches up and is matched to the transaction."
Getting it Together
"Here's an interesting trend," notes Vignette's Milne. "In the past 18 months or so, no RFP we've received has been for a single product category. They're looking for a combination of basic blocking-and-tackling technologies that solve business problems. That's why the term ‘ECM' (enterprise content management) is so popular in the vernacular. ECM, basically, munges together several technologies into one solution. The perception is that this addresses customers' business problems in all the many ways necessary."
Mark Seamans confirms the "business-first" trend among solutions providers. "We've (Verity) begun to verticalize our business, and gear toward a solution-sale model. For example, we reorganized and segmented into insurance, healthcare, higher education, back office operations, etc. That approach works, and has had as much to do with successes in this industry as product development has. Before with imaging, it was all about storage—‘Get it off my desk.' Now it's about information management and retrieval. Now instead of a room full of paper, we have a room full of storage devices. So retrieval is key, but retrieval must be presented in the context of a business process. We're looking for a way to take information from the paper process and synch it, validate it, add to it the pieces in a business process that were missing before.
What that means, of course, is an integration of technology categories that weren't really integrated before. There are now a lot of consortiums of odd bedfellows." It's nice to think that buyers of imaging products are doing so strictly to make their businesses run better. But there are certainly other forces at work, are there not?
"Lowering costs have also played a role," says Priemer. "It used to be the case that only the largest lenders or insurers could afford these solutions. Now with cost reductions, the same solutions are available in the mid-market, too."
Which begins to explain the somewhat unremarkable adoption path for imaging. For a long time, imaging was perceived as expensive and complicated. And it was. When it's a complex problem, smaller companies have a harder time devoting people to the projects. The users had to not only assign people, but dedicate them—from selection to design to implementation.
AnyDoc's Chuck Jackson agrees that the perception among the user community no longer matches the reality. "Take OCR accuracy as an example. The accuracy is so much better in the past five years that it qualifies as the single technology component that has propelled imaging to where it is today. Back when we first developed software for forms processing, the first system took 45 seconds to process one form. Now that same form takes less than a third of a second.
"When we started developing software for unstructured forms five years ago, OCR was one of the Achilles' heels. OCR was good, but not really good enough. But we bet that it would get better. The bet played out well." Sam Schrage thinks this kind of information is not getting to the right people at the right time, to paraphrase a KM mantra. "The imaging community has stopped the education process. We haven't gone out to the midmarket to teach them the benefits of this technology. Everyone goes after the big accounts, where the big money is. That's because this technology is channel distributed, by and large. The channel guys go after the big accounts, not the medium-sized ones.
"We had someone tell us ‘OCR? That doesn't work.' That was a quote from an attendee at a conference two weeks ago. So here we are, in 2005, and we still have people who believe OCR doesn't work." (For the record: it does.)
Why is that? "Every two to three years, there's a wave of new guys entering the market. They don't have the experience that a lot of vendors in this marketplace, who've been around a long time, have, and they implement solutions that don't work and leave a bad taste in the customer's mouth. So the customers just close it down and wait two or three years before they're willing to look at it again."
Education is apparently lacking, and maybe this White Paper can help turn the tide, if only a little. It'll take a lot more. "There are still no industry behemoths evangelizing this technology," says Priemer. "The largest companies in our space are still only $300 million or $400 million companies. The really big companies—the HPs, the Oracles, the Microsofts—have awfully big megaphones. Getting the world to know about our industry is a pretty tall order for the companies that currently are in it."
What you're reading now probably does NOT qualify as a "big megaphone." But it is full of the kind of eye-opening information that Sam and Chuck and Bill and Bruce and Mark all agree is critical to the acceptance and proper application of imaging and data capture technologies where they can do the most good. Here's hoping it's loud enough to start the process. Happy reading!
Andy Moore is a 25-year publishing professional, editor and writer who concentrates on business process improvement through document and content management. As a publication editor, Moore most recently was editor-in-chief and co-publisher of KMWorld Magazine. He is now publisher of KMWorld Magazine and its related online publications. Moore acts as chair for the "KMWorld Best Practices White Papers", overseeing editorial content, conducting market research and writing the opening essays for each of the white papers in the series.
He has been fortunate enough to cover emerging areas of applied technology for much of his career, ranging from telecom and networking through to information management. In this role, he has been pleased to witness first-hand the decade's most significant business and organizational revolution: the drive to leverage organizational knowledge assets (documents, records, information and object repositories) to improve performance and improve lives.
Moore is based in Camden, Maine, and can be reached at andy_moore@verizon.net.