Predictions for the Next Ten Years

Could Community Source be a Better Approach for Supply Chain Applications?

By Steve Keifer
VP of Industry and Product Marketing

There are not many examples of open source in the B2B integration space. Many corporate end users and service providers are running B2B platforms on open-source software such as Apache’s web server, Sun’s MySQL database or the various distributions of the Linux operating system. But very few vendors have introduced open-source models for their communications, translator or integration broker software. Is this just an oversight by the vendor community or is there no value to be gained from open source in B2B? 

Can B2B Vendors See Beyond License Revenues?

Perhaps, the highest profile example of open source in the B2B sector was the integration software vendor SeeBeyond. SeeBeyond had one of the most competitive B2B integration broker suites in the marketplace prior to its acquisition by Sun Microsystems in March 2005. Sun’s strategy was to license SeeBeyond as an open-source product to drive adoption. However, following the acquisition, the sun appeared to set on SeeBeyond as it enjoyed considerably less visibility in the market. The future for SeeBeyond is not bright with Sun’s pending acquisition by Oracle. Having recently acquired BEA, Oracle already boasts a strong portfolio of integration products in its Fusion suite.   

The best successful example of open source in the integration market is Talend. Talend is a leading developer of integration and Extraction, Translation and Loading (ETL) software.  Talend has an open-source version of its products, which it makes available to the developer community through a GPL license. However, Talend’s own R&D team drives the product development roadmap. Revenue is generated primarily through services to support its integration products – consulting, training and technical support. Talend also has introduced an on demand version of the product, which they will host and operate on a subscription basis. 

Understanding Open Source

One of the key barriers to adoption of open source in the vendor community is that it is often associated with “free.” But the actual definition of open source does not specify that the software must be free. Instead, open source refers to a different approach to managing intellectual property rights for software. For example, Gartner defines three key principles for open source:

• The general public has access to the source code for the software.
• Licensees have unlimited rights to modify the software for their own use.
• End users and companies have rights to freely distribute the program.
  

Many open-source products have no upfront license cost or recurring maintenance fee. However, such software is hardly free when one considers the servers and IT personnel required to run the applications. Another popular misconception about open source is that it is unlicensed. Open-source software publishers enjoy the same copyright laws as “closed-source” models. The difference is that licensors of open source relinquish all of the control that traditional commercial software vendors typically maintain.

The benefits of open source are many.  By allowing the general public to enhance the source code and use it for their own applications, the software gains access to a community of “free” developers. Changes to the code might include feature enhancements, defect removal or security fixes. Proponents of open source believe that the community development model accelerates the innovation process by creating faster time to market for new features. The beliefs are supported by numerous successful case studies, including the Mozilla Firefox browser, Eclipse development environment or the Mediawiki application that runs Wikipedia.

Community Source

Community source is a type of open-source approach in which a group of companies unite to develop software that solves a common business problem. Typically, the application being developed either is not available from a commercial software vendor, or is available, but only via a cost-prohibitive licensing model. There are several different models of community source, but the most popular is called a “gated community,” in which only selected member organizations contribute to the development. The gated community concept differs from traditional open source to which the general public can contribute. The focus of community source is not to build commercial applications for resale, but rather to create useful software that the developer community can leverage for business benefit.

There are numerous examples of successful community-source models in the public sector and higher education. Examples of local government uses of community source are widespread in Europe:

• Association of Developers and Users of Open-Source Software in Administrations and Local Communities (ADULLACT) in France
• Programverket in Sweden
• CommunesPlone in Belgium
  

Another successful case study can be found in higher education with the Sakai Project. MIT, Stanford University, Indiana University and the University of Michigan started Sakai in January 2004 to develop a community-source platform for e-learning. The universities contributed 27 developers and millions of dollars to seed the project. Shortly thereafter 20 other higher education institutions joined in, accelerating the development of the many new releases introduced in the past five years.

B2B e-Marketplaces – A Closed-Source Community Model

In the past 10 years, hundreds of millions of dollars have been invested to build collaborative, community-oriented applications for the supply chain. During the dot-com era, there were 2,500 electronic marketplaces formed, each with ambitions of leveraging newfound e-commerce technologies to rid the supply chain of inefficiencies and waste. Independent marketplaces, sometimes called public marketplaces, were the first to emerge in the late 1990s. These independent marketplaces, such as VerticalNet and FreeMarkets, were founded by venture capitalists and managed by entrepreneurs.  Shortly after the introduction of the public e-marketplaces, consortia of buyers and sellers, threatened by the potential disruption to historical supply chain dynamics, rushed to build their own exchanges.  Examples included CPGmarket.com in the consumer goods segment; SupplyOn in automotive; E2open in high tech; WWRE in retail; Exostar in aerospace; and Elemica in chemicals. 

Of course, all the e-marketplaces were closed-source models with applications owned by the exchanges themselves.  However, the closed nature of the software development was not popularly viewed as a reason why the e-marketplaces failed to achieve their ambitions. Some of the challenges were simply a matter of timing. For example, during the early days of the Internet there was cultural resistance to new business models. However, other barriers such as pricing were more fundamental. The consortia and public e-marketplaces failed to develop a pricing model that appealed to customers. Some exchanges assessed transaction fees, representing a percentage (0.5% to 8.0%) of the dollar value of goods bought or sold. Others charged annual subscription fees for unlimited usage of the marketplace services. A few creative exchanges experimented with advertising models and referral fees. In the end, customers were reluctant to pay. 

In hindsight, the most successful of the e-marketplaces were the private exchanges. Major buying organizations not affiliated with any exchange (as well as many consortia investors) began to develop private marketplaces starting in 2000. Perhaps, the largest and most successful is Walmart’s RetailLink, which was actually started in the 1990s. The goals of private marketplaces were simply to improve supply chain efficiency, rather than to generate transaction fees or grow towards an IPO. Funding was provided as part of the traditional IT budget rather than through external capital markets or equity investments.

Commercial factors are among the primary reasons why most joint ventures, consortia and technology alliances in the supply chain fail to achieve their original promise. Whenever a separate legal entity is created to support the B2B e-commerce platform, there becomes a shift in focus from just solving business problems towards generating sufficient revenues to sustain operations. The “revenue distraction” occurs whether the entity is a for-profit or not. Unfortunately, many of the B2B e-marketplaces suffered from irrational exuberance about the opportunities to generate ROI for their equity holders. Management’s focus became principally-oriented towards the goal of an IPO rather than satisfying the community’s technology and business needs.

Data Pools – A Closed-Source Community Model

Another example of a community-oriented supply chain application is the Global Data Synchronization Network (GDSN) championed by the GS1 standards organization. Data synchronization enjoyed significant hype during the period of 2003-2006. Data sync standardizes the process for exchanging product data such as pricing, branding, packaging and weights between retailers and suppliers. The industry believed that data sync would generate significant ROI by reducing invoice deductions, curtailing out-of-stocks and accelerating new product introductions. GDSN has achieved greater success than the B2B e-marketplaces, but has not grown as quickly or as broadly as originally expected. 

The GDSN consists of a loosely-coupled network of data pools, which are commercial entities designed to provide data sync services to a community of companies within a specific geographic region or industry subsector. In Europe and Asia many of the data pools are GS1 member organizations from a specific country such as the UK, Australia or Korea. In the US, there several data pool models. The GS1 US organization has a subsidiary 1SYNC, which is the most broadly-used data synchronization service.  Additionally, there are numerous technology vendors which offer data sync services using their own certified data pools. A mix of both for-profit and not-for-profit business models exists in the GDSN market, but all data pools are based upon closed-source software applications.

There are numerous reasons why data synchronization has not experienced greater adoption. Some believe that the lack of focus on price synchronization, which generates the greatest ROI, was a deterrent to investment. Others blame the standards-development process for excessive scope creep and unrealistic expectations. However, few would disagree that the pricing models remain a key barrier to entry. Many retailers have been reluctant to pay subscription fees to commercial service providers for the use of data pools. Some companies believe that the pricing models offer a fair value. However, many others believe that there should be no fees for sharing information between business partners. A number of retailers have circumvented the data pools by building supplier-facing portals for data entry. These retailer portals leverage the GS1 standards, but not the commercial data pool services.

Community Source for the Supply Chain

Any application which is designed for the use of multiple different companies in a supply chain, some of which are competitors, is fraught with challenges. There will always be governance issues, neutrality concerns and conflicting views on the product roadmap. However, commercial issues such as pricing are not a necessary evil. The community-source models, which have proven success in higher education and local government, could be an alternative approach for supply chain applications. 

With community source, buyers and suppliers could join together to develop collaborative supply chain applications, but without the distractions of revenue generation and pricing. Unlike a pure open-source model, the developer community would be limited to those entities approved for participation. Each participating company could choose to allocate business analysts, software developers and testing experts to the community projects. Unlike a multi-million dollar equity investment in an exchange or a hundred-thousand dollar subscription to a data pool, the use of internal resources for a project does not require incremental budget. In fact, it is likely that the resources assigned to the community project would be otherwise deployed to solve the same problem independently by customizing ERP applications. Community source offers less risk of financial exposure as well. If a major buyer or supplier became frustrated with the direction of the community-source application, the participant could simply scale back the resource commitment. There would be no “sunk costs,” such as subscription fees or equity investments to salvage. Alternatively, a frustrated participant could choose to fork development efforts with a derivative of the source code they manage independently.

There are numerous advantages to a community-source model. By allowing a group of companies to enhance the source code and use it for their own applications, the software gains access to a community of “free” developers.  Much as with Firefox, Apache or MySQL, the open-source development model could accelerate the innovation process for B2B e-commerce by creating faster time to market for new features.   

In the next 10 years, we will see more experimentation with community-source models in the supply chain. The experimentation will be a natural follow-on effect of the open-source model becoming a mainstream approach to software licensing. While community source will not eliminate all of the challenges experienced by multi-enterprise supply chain applications, it could significantly reduce the commercial complexity and, therefore, boost adoption rates. The big question is not if, but when we will see the first industry community-source project develop.