Cover Story:

The Key to Enterprise Resource Management:

The Document Supply Chain

by David Yockelson

Watts Wacker of the Stanford Research Institute as noted that " the value of information about a transaction has exceeded the value of the transaction itself." This quote succinctly proclaims the evolved value of information management as a key infrastructure capability for companies intent on succeeding in the next millennium. Across vertical markets, organizations have begun to consider ways in which the management of information, documents, and processes can become part of an enterprise wide technical architecture (EWTA), providing the foundation for customer management, knowledge management, and innovation in markets and products ultimately for competitive differentiation and success. Since the core component of information transacted either through commerce or collaboration, regardless of medium, continues to be the document (in various forms of product specifications, EDI transaction sets, invoices, customer correspondence, etc.), companies must build appropriate mechanisms into their strategic architectures to manufacture, inventory, and manage logistics for documents. We have designated this set of internally and externally driven mechanisms the document supply chain (DSC).

Unfortunately, we do not believe that technology platform vendors (IBM, Microsoft, HP, Oracle, etc.) will provide information management (including compound document management, image capture/management, output management, and workflow) facilities adequate for most corporate requirements. Basic services (document access and version control, routing, capture, search) will be readily available at an additional per seat price (above a "standard" desktop) of <$20 by YE99, but robust facilities (document assembly, process creation/management, intelligent data/object replication, print stream management, etc.) will be available few platform vendors prior to YE99 (without acquisition). Best-of-breed DSC vendors will continue to own the market through 2001.

What Comprises the Document Supply Chain?
As in a product/service oriented supply chain, there a number of core components manufacturing, inventory, and logistics (or distribution) that comprise a DSC, and each of these has various constituent facilities or components that are document-focused:

Manufacturing
Creation/Authoring
Capture
Assembly
Identification
Inventory
Management
Retention
Catalog
Search/Request
Logistics
Logical/intelligent distribution
Physical distribution

As in a traditional supply chain, distributors, sourcers, integrators, etc. are key to successful DSC implementation and ongoing support, either in terms of delivering the appropriate capabilities for "insourcing" or providing outsourced services (i.e., clearinghouse, secure delivery, warehousing, etc.).

Provider Positioning in the DSC
Last year, industry pundits lauded FileNET's new management and product acquisition/integration strategy as being "IBM-like." However, we believe that a model echoing IBM's former IT dominance (e.g., being the sole provider of every technology facility as well as the progenitor of standards) is inappropriate in an age of virtual corporations, web oriented standards and connectivity, increasingly shortened development cycles and dramatically fast technology innovation. Indeed, an SAP model might be more appropriate, focused on providing infrastructure-level facilities that deliver business value while creating a secondary market of companion technologies and third party integrators/sourcers. While customers would certainly prefer to have the broadest set of facilities (products and services) delivered easily and economically under one vendor umbrella, we do not believe that any one vendor will satisfy all DSC requirements, nor can one vendor support all varieties of delivery and integration across DSC components or organizational variants. There will be both "supermarket" and "boutique" providers of products and services, there will be partnership and acquisition, but as the demands of the DSC are transformed through evolving customer and competitive imperatives, all must not only co-exist but strive for connectivity with respect to both organizational and technical DSC constituents.

We believe that vendors we had formerly categorized as "strategic" are leading the DSC charge, although all are missing components that could be critical for companies with severe requirements in particular parts of the chain (e.g., database publishing/output for customer communications; infrastructure-level, transactional, process/state management). For example, FileNET, IBM, PC DOCS, Open Text, Documentum, and Altris all possess a set of function containing at least one best-of-breed (or close) element, but all require additional help for records management, multimedia publishing, etc. While we expect the "strategic vendor" approach to continue, we also believe that certain best-of-breed suppliers/technologies will have a leg up on infrastructure demand/readiness. In particular, the compound document management and workflow areas are currently in greatest demand relative to forming the basis for corporate DSC initiatives. All of these should start with compound document management and workflow suppliers.

Putting Links in the DSC
Though we find an ever-increasing number of companies buying these facilities as infrastructure (rather than independent applications), selecting the right "future proof" DSC components is quite difficult. Over the next 2-3 years, amid acquisitions (focused on uniting CDM with imaging, CDM with web content management, workflow with middleware, etc.) that will strengthen various DSC technologies, we believe that users will begin standardizing warehouse and logistics capabilities (and vendors) while maintaining a varied approach to manufacturing. A key task will be to understand which of the major DSC areas are strategic and which are tactical (see Figure 1). Post 2000, flexible DSC architectures (likely XML oriented) will enable plug and play of best-of-breed components when necessary.

Manufacturing
The manufacturing segment of the DSC is comprised of mechanisms and products that enable users to create (or author), capture, convert, or automatically generate documents (or in web parlance, "content"). Clearly, the major manufacturing platform today is Microsoft Word, although there are still legions of Frame, Interleaf, Quark, SGML, and other authors around. From a document image standpoint, although there are many image capture applications (available as a part of a comprehensive imaging environment or attached to other platforms such as document management, ERM applications, etc.), TIFF (tagged image file format) is the most widely accepted standard (and will continue to be so for the foreseeable future, despite the rise of JPEG for photographs). For conversion, there is no standard facility other than the viewing facilities found in document managers, browsers, operating systems, groupweb products, etc. (mostly Inso's Outside In or Verity's KeyView), although we believe that conversion to Adobe Acrobat PDF (portable document format) should be mandatory for any document warehousing/logistics venture (and is also viable for conversion from TIFF via Acrobat Capture, particularly for low to medium image volumes). Futures across this space point to a requirement of flexibility; that is, we believe that it is mandatory to enable authors to create in the platform(s) of their choice while providing a warehouse and logistics infrastructure that can accept, convert (or render), and deliver document content to any user on any platform.

Originally, SGML was thought to be THE platform for this panacea of publishing. However, difficulties and expense associated with training (tagging, DTD development, management, etc.) and conversion focused SGML mostly on niches specified by the government (through early CALS initiatives) such as aerospace (ATA 1000 regulations et al) and manufacturing. Complex publishers (both commercial and corporate) also have experience with SGML. Ultimately, XML (extensible markup language) will replace SGML as a major repurposing and multi-platform publishing platform, but it will not be a mass market phenomenon until late 1999 (e.g., MSFT's "save as XML" in future editions of Word, increased presence/use of the OFX format in banking). Nearer term (YE98), Interleaf (with Microstar) will focus on XML with "BladeRunner," an XML collection and publishing engine that can accept, convert, and assemble most document formats to XML for management and/or publishing. Xyvision, Texcel, and Chrystal, traditional SGML-oriented authoring and management vendors (with ArborText and Microstar), have also begun to add extensive XML warehousing and publishing capabilities.

Warehousing
The Document Management Alliance's (DMA) attempt at constructing a standard for linking to, from, and among different compound document management (CDM), imaging, etc. systems, is still incomplete (no compliance rules, poor end user and major vendor acceptance). Thus, we encourage users that are considering products to weigh standardization on a single vendor's product/product set vs. multiple tactical (or best-of-breed toolkit) purchases. For example, companies considering volume purchases of Eastman Software's work management for Exchange (WMX) products (WFX for workflow; DMX for document management) must judge function and price (roughly $200/seat combined in volume) against more robust vendor products/platforms that can solve the sophisticated AND simple problems (e.g., Documentum, Open Text). Tough negotiation will likely result in a roughly equivalent price, and while higher end products do not yet integrate as cleanly with Exchange, they can prevent multiple CDM decisions/purchases over time for other business functions. Still, companies with corporate-wide Documentum licenses might find tactical opportunities for DMX for with less sophisticated customization and warehouse requirements (and for which deployment costs must be minimal). Moreover, FileNET's position of strength is as a "strategic vendor" that can provide multiple document manufacturing, warehouse and logistics functions (and its Panagon integration) will enable it to maintain a leadership position in the DSC market through 2001, although it will be increasingly challenged by:
    Documentum best-of-breed CDM infrastructure and increased DSC expansion by 1999
    Open Text collaborative knowledge management environment today, multiple DSC services by 1999 (Documentum and Open Text will deliver improved imaging facilities by mid 1998; both integrate with Acrobat Capture today via Cornerstone)
    IBM EDM Services division that will match its products with Documentum (4Q98) and FileNET (mid 1999).
    PC DOCS more from its ability to sell multiple products into the extensive PC DOCS Group installed base than from product superiority; PC DOCS must improve imaging and workflow (2H98) facilities despite numerous partners.
Optika, USI, BancTec, Eastman Software, and a variety of others will also contend for DSC dominance, either through technology differentiation (e.g., Optika with electronic commerce tie-ins, USI with cross-application process management) or by acquiring/partnering for additional DSC functions.

Logistics
We note two types of logistics (or distribution) within the DSC. First is logical or intelligent distribution, or the ability to reference, organize, route, or provide access to warehoused documents without physically (necessarily) moving the document itself. Rather, logical links such as catalogs, pointers, etc. are used to provide users connectivity to documents. Within this collective are workflow, search/retrieval, agentry, and document organization facilities. Of these, workflow is closest to becoming an infrastructure element, both at the low end (collaborative) and the high end (closely related to middleware). Search/retrieval is next, although many companies still balk at high product cost. We believe that this will change over the next 9-12 months as volume pricing (i.e., implementations of thousands of seats) approaches $50/seat.

Physical distribution of documents is a growing market. Just as companies demand certain quality of service, trust, etc. for EDI transaction set documents, they are beginning to demand secure and certain transportation of documents containing richer content (e.g., CAD, Word, PDF). Numerous vendors have begun to address this, although most are small companies (Tumbleweed, NetDox). eParcel, spun out of Mitsubishi, provides not only an IP and server-based approach but has also focused on network management and delivery of Internet content (e.g., advertisements/marketing information). This market will grow strongly through 2000 as Value Added Networks (VANs), Independent Service Providers (ISPs), companies, governments, and integrators begin to include these facilities to provide the electronic equivalent of UPS and/or FedEX (indeed, Tumbleweed has already established a relationship with UPS).

Applying DSC Principles to Provider Evaluation
In 1997, META Group found that organizations requiring a range of information and process management services had two basic options: select from various best-of-breed vendors and products and integrate these as needed to provide solutions; or take a "strategic vendor" approach that offered stronger negotiation and integration opportunities but could sacrifice some best-of-breed capabilities. We found most G2000 (Gobal 2000) organizations begin to focus on the strategic vendor approach, but also found certain exceptions. For example, workflow and compound document management (CDM) facilities began to be demanded at an infrastructure level, driving some companies to focus on best-of-breed for these technologies but at a higher level for others (e.g., document imaging, output management, web content management, search/retrieval). But companies focused on knowledge management might start with search/retrieval, groupweb, and intranet content management facilities while sacrificing best-of-breed or even delaying product/technology selection in other areas. In 1998, we believe that organizations must take the more holistic DSC approach, determining requirements across multiple lines of business (thus reaching outside the company as well as internally) for each of the aforementioned constituent areas. In turn, end users must consider both strategic and tactical vendors relative to offerings across the DSC as well as discerning various suppliers' abilities to provide function and services in a "DSC-pluggable" capacity (and in so doing, recognizing the growing availability of standards ODMA, DMA, WebDAV, XML, jFLOW, SWAP, etc. in the DSC arena).


BIO -- David Yockelson is Vice President and Director of the Advanced
Information Management Strategies service (AIMS) at META Group. David
focuses on issues such as electronic commerce, document supply chain,
knowledge management, web content creation/management, and IT/business
organizational issues. David can be reached at 203-973-6700 or
david.yockelson@metagroup.com.