ECM laggards beware…
Core vs. Context
In his book Living on the Fault Line, Geoffrey Moore argues that to maintain competitive advantage, your company must focus the bulk of its attention on its core. Core activities are the services and products that differentiate your offerings in the eyes of the customer. At the same time, you must still effectively manage your context activities-those processes that are key to sustaining your business but aren't competitive differentiators-through outsourcing, partnering or other means. Unfortunately, context activities get in the way of focusing on core advantage and competitive advantage erodes.
The more time and focus you apply to core activities in product creation or service that attract new customers and keep customers coming back, the greater your competitive advantage. The greater your competitive advantage (the Competitive Advantage Gap or GAP) and the longer you can keep this advantage (the Competitive Advantage Period or CAP), the greater the market values your company. This concept is so powerful, that many financial analysts now use CAP x GAP analysis to determine the market cap of high tech companies.
To determine what is core, Moore suggests making a list of the things that bring customers back to you. If you are uncertain, ask your customers what they like about your products or services. Gather information from customer-interfacing employees. Once you determine what is core, Moore advises to put the bulk of your available resources on it. The result of doing this can result in cost saving, but more importantly, it restores your focus in the business, enhances your position and perception in the market, and increases market capitalization.
The Shape of the ECM Market
The term “Enterprise Content Management” did not even exist prior to 1998. However, since 2001, all the major players within this now recognized market space started to identify themselves as ECM leaders. What followed was a period of consolidation where the larger companies started acquiring missing pieces of an overall architecture. This architecture was presaged a bit by analysts such as David Yockelson of Meta Group. However, this overall architecture was defined more by which competitor acquired what piece of technology. Within a period of two to three years, what had been the convergence of simply document management and web content management now included image management, records management, digital rights management, digital asset management and archiving among other functions. This set of functionality has now stabilized and the acquisitions of new functionality have slowed.
This process has commoditized the market. Competitive differentiation is now based upon size rather than functionality. In this consolidation and commoditization period, three clear winners have emerged who have grown market share faster and have accumulated more comprehensive stacks of ECM capability. These are IBM, EMC-Documentum and Microsoft. Both IBM and EMC believe they are relatively close to each other in market size and growth rate of 30% per annum. Analysts now believe that the Documentum component of EMC is now worth $2.5B of EMC’s $27B market cap. As of 2005, Microsoft had sold over 6 Million Sharepoint licenses. The performance of all other ECM has remained flat at best.
With the commoditization of the market, there really is little to differentiate the various players. All platforms have basic library services, metadata management, search and retrieval, workflow, portal integration, and development kits. In addition, common applications are document management, web content management, image management, forms management, and records management. To distinguish an ECM vendor in the future requires new applications or the enhancement of existing vendors. At this point, no vendor has a strategic advantage with its repository. The fact that several vendors have more than one repository or underperforming has actually been a strategic disadvantage. The underlying repository has not been the reason customers have purchased an ECM system, but it has been a reason why customers have rejected an ECM system.
This is not to say that compelling problems that need to be solved in the ECM market. ECM remains hard to use despite the fact that it must be used by some of the least sophisticated computer users. Cost and difficulty of implementation of ECM still remains very high. Very few large corporations have actually been able to implement an Enterprise-wide content infrastructure since distributed and federated solutions are generally hard to set-up, difficult to manage and poor performing. Performance remains inadequate to provide real-time access, analysis and preferences. Automation of classification, data entry and administration still seem very far away despite being discussed for years.
There have been many attempts to standardize the Content Management technology space -- ODMA, DMA, JCR, iECM. However, it is in the context of the market consolidation and commoditization that major players are participating more in the standards process. Starting with JSR-170, despite the long time required to develop that specification, there are now moves by the major vendors to actually implement standards. IBM has been working on their implementation for over a year and intends to make major announcements in the later part of this year. Documentum plans on providing an implementation in their next release, as does FileNet. Oracle has made JSR-170 the foundation to combine all their content technologies, including content access from Oracle Applications.
Some of the interest in the real sense that repositories can now be fungible and the larger players can displace some of the smaller vendors in their largest customers. However, much of the pressure comes from the large customers themselves. As these companies implement Service-Oriented Architectures to consolidate their other systems, they demand that they be able to do so with their content repositories. They don’t want to have 25 different types of repositories any more than the large vendors do. AIIM’s iECM has been formed by the largest vendors -- EMC, IBM, FileNet, and Microsoft -- along with some of their largest customers to define a Service-Oriented Architecture and Web Services interfaces. The iECM working groups have identified the scope for this specification and the first draft of the web services could be available by the end of the year. The timetable for this is driven by JSR-283, which expects to have version 2 of JCR available by the end of this year.
The scene is set for real standardization in content management and commoditization to the point of real replacement and swap out of existing systems. The environment is very similar to that found when the database market first standardized through SQL-89 and SQL-92. This was basis upon which the smaller players in DBMS disappeared and the new era of client-server and ultimately the web appeared. The changes in the ECM market can be just as profound.
Power of an Open Source Community
It is in the context of these changes in a market that in order to play, you must change the rules. That is what Alfresco has done by introducing the open source model. Open source levels a playing field that is dominated by vendors based upon their size. The most successful example is Linux which is now, by far, the most dominant form of Unix. JBoss has been able to take the top spot in terms of number of installs for all application servers. MySQL is proliferating faster than any other database to challenge old market leaders such as Sybase and to make life uncomfortable for Oracle and Microsoft.
Open source is able to compete three ways. The first is the breadth of distribution that comes from being able to download for free and try a product before paying for it. Open source is therefore able to go farther and broader than even Microsoft to places that commercial software has not been to before, especially Enterprise Content Management. It allows organizations that are either in a non-critical part of an enterprise and have no evaluation budget or in an enterprise that is far from any sales offices to try the software.
The second competitive factor is that the community helps push outwards to others and to support itself in the process. This word of mouth requires little or no marketing spend to move others to try the system and that has certainly been the case with Alfresco. People are more likely to write blogs or even standard articles on open source than they are on commercial software.
The third way the open source competes is through much faster innovation. Because the source code is open there is more cooperation and input on the development of new capabilities. Users feel freer to make suggestions all the way down to the code level. In addition, there can be an open dialog on how best to approach the development of software. There is also more cooperation between different open source groups on the development of technologies. For instance, Alfresco works very closely with JBoss, Lucene and other open source communities in fixes and new components. In addition, the level of testing is much higher and testers self-selected rather than recruited which leads to much faster maturation of software. Finally, other members of the community contribute functionality that they have found useful and think others might as well. As an example, a large pharmaceutical company is developing Alfresco’s Wiki capability. From the Alfresco experience, the development of new ECM functionality is anywhere between two to four times faster than in the old, commercial (Documentum) model.
Next Generation Repository
Facing an aging, commoditizing Enterprise Content Management market, Alfresco has used open source to provide a new approach to ECM. This open source platform accelerates the development of an ECM solution and can ultimately outdistance the ECM laggards. I have no fear for our future.
With the benefit of deep experience and a blank slate, Alfresco has used best of breed open source components to construct very fast, extensible and standards-based repository. Using open source components that have matured very rapidly, we have avoided building many of the pieces that other repositories had to build for themselves. Some of these used by the repository include:
- Spring - A framework that provides the wiring of the repository and the tools to extend capabilities without rebuilding the repository (Aspect-Oriented Programming)
- Hibernate - An object-relational mapping tool that stores content metadata in database and handles all the idiosyncrasies of each SQL dialect
- Lucene - An internet-scale full-text and general purpose information retrieval engine that supports federated search, taxonomic, XML and full-text search
- EHCache - Distributed intelligent caching of content and metadata in a loosely coupled environment
- jBPM - A full featured enterprise production workflow and business process engine that includes BPEL4WS support
- Chiba - A complete Xforms interface that can be used for the configuration and management of the repository
- Open Office - Provides a server-based and Linux-compatible transformation of MS Office based content
- ImageMagic - Supports transformation and watermarking of images
Using these core open source technologies, we have created a repository that is more flexible, scalable and faster than previous generations of content repositories. Some of the characteristics of the Alfresco repository include:
- Content Aspects - These pluggable components can be added at runtime and allow multiple metadata types and classifications, capabilities such as versioning, auditing or lifecycle or add new methods such as special permission control or a new publishing capability
- Event-based Rules Engine - Based upon events in folders and on content objects, process actions such as execute workflows, transform content, modify metadata, classify objects, or extract metadata.
- Templating Language - A very simple scripting language that provides the mark-up and transformation of content, the embedding of metadata in content and the construction of compound content from a virtual document or structured folders.
- Service-Oriented Architecture - Designed from the beginning to fit in SOA architecture, Web Services provides a coarse-grained interface that allows the repository to interface with languages such as PHP and .NET and to integrate to other systems such as CRM, ERP and BPEL process engines.
- Security - The repository provides a state of the art Access Control List and Role-based Access Control mechanism that evaluates efficiently at runtime to filter query results using a distributed caching mechanism.
- Workflow and Business Process Management - Using the JBoss jBPM engine provides one of the most complete business process capabilities available in an Enterprise Content Management suite. This engine supports the BPEL 1.1 standard and will soon support 2.0 of the specification. In addition, due to its use of Java reflection, workflow scripts can directly access repository metadata and process information as well as run queries directly.
- Taxonomy Support - The Lucene engine has been configured to index content, metadata and hierarchical classifications. These classifications can be loaded using an XML format and can be multi-dimensional. This allows searching within the context of more than one classification dimension providing a slice ‘n’ dice capability and narrowing of searches using metadata.
- Standards Support - The repository supports Level 2 of the JSR-170 specification with the fastest implementation of the standard available on the market. It is the only implementation currently that is thread-safe, transaction-safe, clusterable, and work in multi-processor environment. In addition, support for web services provides support of other languages and BPEL business processes.
- Clustering - Due to the lightweight object structure and the stateless service interface, the repository is easily clustered using a distributed cache in a loosely coupled environment and allows the repository to make use of spare computing capacity.
- Replication - Taking advantage of the native replication capability of the underlying database, the repository can handle very fast replication of metadata and supports the efficient transfer of content from repository to another. Bi-directional replication is under development
- High-Availability - Designed as one of the few open source systems to run in a 24x7 environment, the repository uses replication to provide automatic fail-over of the repository server in case of system failure. This can work using synchronous replication for low latent delay in same data center or asynchronous replication for disaster recovery.
- Import / Export - The repository supports a simple XML-based metadata and standard .zip format or a JSR-170-based export format to provide import or export to or from same repositories or other repositories such as Sharepoint or JSR-170 compatible repositories.
My long and tedious list is intentional, because we were able to create this list in an incredibly short period of time. My intent is to impress upon you what can be accomplished with open source components and open source methodology in a period of only one year.
ECM vendors must determine what is core to their business - applications or repository. If they decide that it is a repository, then they must fight the largest vendors, now IBM, EMC and Microsoft on the basis of political structure and account control. Alternatively, open source provides an alternative on which closed source can compete. With Alfresco, we are confident in the future with an innovative environment of open source. Can the older players continue to be complacent unless they are the largest?