Document Image Processing: The Big Dippers

Document image processing is not for the reseller after a get-rich-quick scheme, but if you are willing to invest some time, it can be a very lucky DIP, says Sean Hallahan

Document image processing (DIP) is one of those deceptively attractive IT solutions that the industry has thrown up over the years. And so it is ? if the customer is prepared to spend a large amount of money and effort to install a DIP system.

DIP, which first emerged in the 1980s, is a way of filing, storing and retrieving paper-based documents electronically. In its original form, DIP was only available on a mainframe and at a premium price affordable only by large corporations. But as PC hardware and PC-based operating systems became more powerful, DIP systems became more widely available. DIP features began to be incorporated into products such as Lotus Notes, Perfect Office and Microsoft?s offerings. DIP spawned both workflow and groupware software which gradually diverged from their parent but are now coalescing.

Although many software developers now include DIP or document management features in their standard packages, it is still a specialised field as far as resellers are concerned. To implement a full DIP system requires not only an understanding of the hardware and software characteristics of the products, but an understanding of the way in which the customer?s business works. Installation of a DIP system usually requires a considerable degree of consultancy so the vendor knows which system is most appropriate for the user.

DIP is a part of what is more generically termed document management systems. UK consultancy Cambridge Market Intelligence (CMI) offers a definition of document management systems in its Document Management Report.

?The main tasks of document management systems are to import, create, handle, process, maintain and distribute documents in electronic form to provide improved working efficiency and support for people who need or have to rely on such documents in their jobs. In order to fulfil these tasks these document management systems have to resort to new techniques and technologies. The most important functions of document systems are document entry and storage, document filing and retrieving, and document distribution,? says the report.

There are three market leaders in the DIP business: IBM, which has products running on a variety of platforms, from PCs to mainframes; Filenet, a company specialising in document management systems which runs on Sun, Hewlett Packard, IBM and Windows NT systems; and Wang, a leader in dedicated wordprocessing systems in the 1970s which failed to recognise that its products were about to be made obsolete by the birth of the PC.

Wang turned its attention to image processing and workflow systems with the help of a $25 million investment from IBM in 1991. It has just sold its software business to Eastman Kodak ? the division was renamed Eastman Software to allow Wang to concentrate its efforts on systems integrations and services.

IBM has a number of products which fall into the broad category of document management and DIP which it has brought together as the Enterprise Document Management (EDM) suite. According to Paul Chiswick, IBM channels manager for the EDM suite, there are two sorts of users.

Production users are those for whom document management and image processing is essential to the day-to-day running of their business. Ad hoc users only use the technology occasionally.

Increasingly, like other companies, IBM is embedding document management features into other products, such as Lotus Domino. Chiswick agrees that the installation of a DIP system is not an easy process and that customers need the benefit of consultancy. ?As soon as you put in an imaging system you change the way people work,? he says.

In 1994 there were over 80 European suppliers of document management systems, which comprised DIP, groupware and workflow applications. But since then, according to Colin Barnes, a director of Eastman Software, there have been a number of mergers and consolidations that have reduced the number of suppliers.

DIP products come in all shapes and sizes. At the top end, a DIP system from a company like Filenet can cost from $2,000 to $4,000 a seat, but there are cheaper options and some products are being bundled with operating systems.

Many of the companies in the business have agreements with operating systems suppliers. Wang has an agreement with Microsoft; Filenet has one with Novell.

Many of the companies in the DIP field focus on vertical markets. Documentum, one of the most successful vendors, concentrates its efforts on the pharmaceutical area, while Filenet, specialises in the financial sector. Both have made strenuous efforts to shift themselves away from dependency on their original vertical market segments with a degree of success.

Documentum was set up in 1990 as a subsidiary of Xerox. Xerox remains one of Documentum?s largest customers as well as being a major shareholder, according to a report published last year by the consultancy Ovum, entitled Ovum Evaluates Document Management. Xerox also acts as a reseller of the Documentum software suite. In 1994, Glaxo Wellcome accounted for 34 per cent of Documentum?s $8.9 million software licence revenue.

?Documentum?s marketing strategy has been to take its very general purpose product and use it to go after high-value business-critical processes in vertical markets. It uses a bowling alley approach, targeting a specific functional area within a vertical industry and then using this kingpin stronghold to knock over succeeding layers of functional areas in the same vertical market,? says the report.

DIP and document management systems in general are not off-the-shelf products, although some of the components which make up the system, such as scanners and barcode readers, are. Typically, a high-end system from companies such as Filenet and Documentum requires a considerable degree of consultancy and can take up to two years to install.

According to Barnes, there is an opportunity for resellers to get into the DIP market, but they have to be prepared to invest time and money in training and to wait for payment.

?Most dealers are hardware oriented, although some are trying to get into the value-added business. But I have not yet found one that fits the profile we require,? says Barnes. He says that as profit margins on hardware fall, some dealers may consider it worthwhile to get into the DIP market.

?Hardware volumes will peak and flatten out. With the erosion of margins, some dealers may find a niche market in document management,? he says.

Barnes? view is shared by Simon Stammers, marketing manager at Kodak, which supplies a range of scanners.

Stammers regards the field as too specialised for most resellers, although he believes that there may be a market for Vars that are prepared to specialise. Certainly there is no reason why a reseller could not supply some of the hardware components that go to make up the systems such as storage devices, low-end scanners and the processors themselves.

?The way things are changing dealers may well want to get into selling CDs and jukeboxes into the document management market,? Stammers says.

Chiswick claims that the companies that want to take and resell IBM?s imaging products are the high-end software developers and resellers such as JBA and Baan rather than the average dealer.

One of the problems of DIP systems is that the input devices that are used, usually a scanner, are not 100 per cent accurate. Most companies claim accuracy of about 98 per cent, but this still leaves a considerable margin for error on a 2,000-word Ascii A4 page. Most systems use optical character recognition (OCR) scanners, but these are expensive. Kodak?s latest range of scanners starts at between #21,000 and #25,000 for a model which will scan at 80 pages per minute. A 120ppm system costs between #35,000 and #39,000 and a top of the range 160ppm machine #57,000 to #65,000.

Applications tend to vary from customer to customer and by industry sector. National and international laws and regulations require some organisations to archive documents for many years. Construction and engineering companies have to keep plans and documents for long periods for archiving purposes. Insurance companies and financial institutions are heavy users of DIP systems, but their documents are accessed more regularly for claims processing.

OCR scanners are notably better at reading documents which are typewritten or produced by computer, which in the engineering and construction industry is likely to be the case. In the insurance sector, however, many of the forms processed will be handwritten in cursive script in which an alternative scanning technique is used called intelligent character recognition (ICR). Some companies have taken to barcoding documents or scanning only on a single field.

Despite these problems, document management is a growing field. A number of technological developments out- side of the area are likely to give a boost to imaging systems. Documentum, for instance, sees the Web as providing a boost to document management systems. It has introduced Rightsite, a ?family of products that satisfies the complex challenges of capturing, managing and assembling the frequently changing information on corporate Web sites?.

Chiswick believes that the network computer (NC) will provide an ideal terminal for document management tasks. It is IBM?s view that the NC will initially be used to replace the mainframe and mid-range terminals rather than PCs.

Chiswick, like most of the DIP suppliers, does not believe that most dealers will make a killing out of selling document management solutions. But he does see an opportunity for dealers in selling NCs as customers begin to use them as image processing terminals.

Some inexpensive PC-based standalone products ? in the region of #200 ? are available, according to Martin Waldron, a senior partner at consultancy Inform Systems. He believes there will be a huge surge in PC-based systems, but these are likely to be limited in scope and probably only used at workgroup level. Waldron also predicts that the internet and intranet will give a boost to DIP.

It is unlikely that DIP systems will ever become a mass market product for the channel. The various hardware components that make up the technology, the PCs and workstation clients, the servers, optical disks, scanners and other peripheral devices may provide some profit for the dealers, but the full-blown document management software will continue to be sold and installed by large software houses and consultancies.

Selling document management systems is not for the faint-hearted. Any reseller looking for a quick sale and quick return on investment would do better to stay out of the field. But for those willing to invest in staff and time, consultancy and training will be where the profit is made.