Some have predicted e-business will be the "Killer App" of the Internet. Whether or not this is the case, there is great potential for organizations doing business online (in both the mass market to consumers and between each other in industry) to take it far beyond that which we know as e-business today. This paper describes the state of the art in e-business that is yet to be fully realized, but then goes further to review radical ways in which the technology of tomorrow will allow whole supply chains to cooperate electronically to meet customers' needs. In addition, it suggests that if such a future scenario can be achieved, it will reduce operational costs and foster closer and more productive relationships between providers of products and services and their customers. Indeed it may even be a stimulus to re-engineer organizations themselves.
As more and more companies trade by e-commerce, they become part of online communities made up of their own departments and those of their suppliers and customers. The secure infrastructure in which this e-commerce takes place is becoming known as a community of interest network (CoIN). These are typically created from Internet protocol (IP), virtual private network (VPN), or extranet technology. Once such a CoIN is established, its members can easily use the secure online environment to publish and share information and communicate with each other. This information can be the business knowledge essential to enable commercial revenue generation.
Directory enabled networks (DEN) allow management control based upon parameters and profiles. These are contained within a specialized (optimized for reading) database known as a directory. The directory contents are typically queried using a streamlined open standard such as the lightweight directory access protocol (LDAP). Directories offer much higher read performance for such queries over a traditional (e.g., relational) database. They may be scaled, distributed over networks, and replicated for resilience or performance.
Combining CoINs with DEN offers a scalable approach to coordinated autonomous management along a supply chain, from suppliers through service providers to customers. In this world, each party in the chain (including end users and customers) is empowered to take responsibility for the parts of transactions and business processes which they naturally own, or which they can better control for themselves. This is referred to here as dispersal. Allocating (or dispersing) these responsibilities optimally, and devolving the management of them within each organization, leads to increased efficiency, reduced costs, and closer, more productive supplier-provider and provider-customer relationships.
Modern workflow technology allows processes to be defined and executed such that human and automated steps can be efficiently combined to deliver a result. Roles of people can be referenced such that delegation or automatic resource load balancing can be achieved. Combining workflow technology with directory-enabled CoINs provides a means to enable end-to-end automated business processes along the length of entire supply chains. The people and the roles and responsibilities they fulfill are examples of profiles held in the directory. The distributed nature of the directory over an IP-based CoIN allows the management of such profiles to be distributed too.
This scenario, although not yet common today, is the assumed state of the art in e-business and the foundation for the remainder of this paper. It is illustrated in figure 1. The next section compares the current situation with this scenario. The remainder will concentrate on how this situation could be taken even further, with a specific look at the need to cope with global diversity in communities. A section describing practical experience of work done to test aspects of the research ideas is also included.
With the explosion in numbers of organizations having an Internet presence, it is easy to assume that e-business has already arrived. However, when analyzed, e-business currently means little more than repeated e-commerce transactions. While it is true that customers are increasingly able to interact with back-office systems of commercial enterprises to identify required products and services, place orders, and make payments over the Internet, there is little evidence of community, transferred responsibilities, or end-to-end process automation.
True "Communities of Interest" barely exist for general e-business. The scant evidence of communities ranges from restricted-access websites for "registered" members to the personalization of the content of a website based on a user's previous visit or registered profile. Even the registration is typically limited to one organization and has to be repeated with each party, even if a service offering involves more than one party. Registration is more often simply used as a marketing tool, tracking usage and visits of the website by actual or potential customers. There is little notion of belonging to a community and hence no obvious real benefit from the customer's point of view. The provider-customer relationship is asymmetric in this respect. Customers are often left wondering exactly what is in it for them.
When a transaction takes place in this current scenario, it is often piecemeal in the sense that the online experience for the customer probably takes no account of previous or potential future transactions. Current transactions are typically isolated with respect to other services and aftercare. (For example, an order is submitted online, but any after-sales care has to be done by another means.) From the commercial organization's viewpoint, the infrastructure it uses to communicate with customers is often different from what it uses for online processes. Either way, the notion of total e-business is absent.
There is little if any evidence in current e-business of traditional responsibilities being challenged and re-assigned. Where customers are "allowed" to be empowered by providers, typically responsibility is still withheld by the provider. This is particularly true in terms of information, which is still very give-and-take-oriented with copies of data being made every time it is communicated electronically, as opposed to true sharing of a single data source. Note that there is little true dispersal (as already defined) going on here; each party then takes ownership of its copy. It is especially apparent with data which is obviously best controlled by the customer (e.g., customer contact details). Generally, the traditional approach of the provider taking responsibility for everything is apparent. Coupled with this "desire" to keep all records in-house is the lack of trust that any other approach would be reliable or dependable enough. This is unsurprising given the lack of community building already highlighted above. The latter supports the former.
Current business processes have, in most cases, simply been transcribed to an online equivalent. Customers are given limited access to the systems of the company they do business with. The company defines, controls, and therefore takes responsibility for the processes that customers can interact with and there is little if any scope for tailoring of the customer's part in the process to match their way of working. Process automation is therefore centralized and owned by one party. It is neither configurable, nor end-to-end, and is usually restricted to within one organization in the supply chain. Just as with the community and responsibility aspects, there is a lack of symmetry with the process automation. It is invariably dominated by the control of one organization, usually the one that originated the electronic approach.
This lack of symmetry about community status, responsibility for information, or business process automation is also apparent to varying extents in some of the best-quoted examples of e-business today, such as ANX , Cisco  and Marshall Industries .
Having analyzed traditional approaches to e-business and current examples of online business process automation, the remainder of this paper discusses ideas to extend e-business beyond tomorrow. In particular, as well as building on the aspects of community, responsibility transfer, and whole-supply-chain automation, it incorporates a number of specific advanced principles:
The following sections now look at each of these in more detail. Each section ends with a brief reference to the technology trends that will support these principles.
The communities of interest of the most advanced current e-businesses are based solely on the core supply-chain parties. There is clearly scope to extend this to more indirect parties (such as auditor and government organizations) to enable total e-business. This is widening of the community. Just as practically all types of organization (even the most official public bodies) have now embraced the Internet, the widening of communities of interest will be their natural progression.
So who are the potential indirect members? Government offices provide one source. These might include various industry regulators, tax collection departments in the treasury, and many other trade and industry departments. This will truly be online government. Other indirect parties would include legal bodies such as commercial auditors, as well as strategy, marketing, and other business consultancy enterprises. A number of the issues that face e-commerce and e-business today are related to the traditional legal and, to a lesser extent government-based, cultures that exist and their inertia to embrace the information age at the same rate as the organizations that rely on them. Clearly the culture shift in some parties in these widened communities will be very significant and this will be a long-term development.
The value of widening is apparent, however, especially to those who are already at the forefront of supply-chain e-business. Most organizations that have exploited intranets internally have acknowledged large savings in business process efficiency. Those that have linked their intranets to suppliers and customers (forming extranets) report much higher savings still. Organizations that build a true community and widen it to encompass all parties with whom they interact to do business can expect to maximize their efficiency.
All types of organizations can benefit from becoming a party within an extended CoIN e-business supply chain. It offers possibilities for real changes in the way business generally operates by facilitating real-time tax collection and financial auditing for shareholders.
So how might the widened communities be formed? Because of the culture changes involved for all parties, as well as the trust that it will be necessary to build, membership for indirect parties may be staged. A form of dynamic CoIN membership might be a prelude to later permanent membership. Dynamic membership for short periods on the basis of need (e.g., for traditional commercial audit schedules) might be most appropriate from an infrastructural viewpoint when an organization needs to work within many different communities but just for short, well-defined periods of time.
Communities can be extended in other ways as well as width. The depth of communities can also be increased over what has so far been achieved. Depth has to be based on trust, and so, like widening, cannot be expected to occur quickly. The sense of community between parties will likely deepen according to a basic phased template over time. Initially, each party's basic credentials will admit it to the CoIN and determine the level of interworking. As this basic relationship develops over time, an e-business bonding effect will result in its maturity. The parties will form a history of federated working leading to true trust-based e-business.
As depth of community increases, CoIN membership becomes a major value-adding exercise and eventually an essential strategic advantage for all parties. (This is analogous to the way that few organizations wish to be seen without a World Wide Web presence today.) The benefit of CoIN membership is made obvious to end-customers. In some circumstances, one can imagine smart-cards being used to signify membership in a similar way to how possession of a cash-card today allows you to get cash from any of many machines around the world (i.e., denotes worthiness to get access to money).
This increased "depth" has implications for the formation of alliances, partnerships, and mergers. Within the dynamics of specialist supply chains, it may well simply result in much more cohesive business relationships. The relative business priorities within each organization become more explicit as parties work closer and closer together. It is then much easier to optimize these business processes and responsibilities to the direct benefit of the end-customer.
The extension of IP-based CoINs is supported by security technologies and the ubiquity of IP networks. Newer security technology increases confidence and helps build trust. The ongoing development and scalability of IP protocols will also assist with CoIN extension. This includes not only core protocols such as IP version 6 and beyond, but also new supporting protocols for mobile IP, multicast, etc.
The extension of communities described in the previous section should not just be allowed to happen according to traditional evolution. To achieve the optimal CoIN it should be designed with symmetry in mind. This requires collaborative design between organizations, and in itself, both requires and helps build the trust mentioned in the previous section. The cooperation needs to happen at all levels and with all aspects. From a financial viewpoint, for example, shared investment should mean shared benefits.
By symmetry, we mean that all parties in the CoIN have a relative share in its design, particularly from the responsibility aspect. Symmetry applies to responsibility for information and process, as well as for provision of infrastructure. In the case of the latter, it means a move from shared hub systems to segregated direct access to each other's systems. The cost/benefit ratio is much better when existing systems and data form a symmetrical, integrated CoIN infrastructure. However, it is symmetry in the design of the processes and responsibility for information which is key to realizing an optimized CoIN.
The key to achieving an "honest" symmetry is to identify and optimize who owns what. The analysis of responsibilities from the top level down to detailed processes should be made explicit. It has been shown that this analysis is key to organizing longer-term business strategy . The parties within the CoIN must be prepared to relinquish responsibilities that they have traditionally held, and design into the CoIN the relevant access to other parties' systems who will accept (or more often, reclaim) ownership of the data or process. This may involve radical changes of thinking such as customers taking responsibility for billing or other typically supplier-owned aspects.
Achieving true symmetry of operation requires a risk-taking set of community members. It involves an element of organizational re-engineering as well as significant cultural change. This makes symmetry very difficult but desirable to achieve. However, it results in true communities being formed based on equality. The aim of true CoIN symmetry is to eliminate duplication of process and data, and to build a community within which all parties feel that they have a unique identified role in the supply chain. Each community member must have a relatively equal stake helping to build a trusted business environment.
The achievement of a symmetrical CoIN is not, however, simply a technical design matter. It is also a strategic and highly political issue. The careful technical design must not be compromised by the desires of any party to "use" the CoIN approach to dominate (e.g., using it as a marketing ploy to show the industry how it has "led the way"). This approach demands that the dynamics of the industry or commercial sector it is applied to are consistent with a collaborative and cooperative method of operation. If this is impossible, then it may be better to look at approaches other than that described here.
Security technology for IP networks and applications will assist in achieving symmetry as confidence and trust are built up in the context of corporate IP networks.
Given the network element of the CoIN approach, there is an absolute need to avoid centralized management. There is little point in dispersing responsibility for information and processes out to various parties if the organization and management of those responsibilities itself creates centralized bottlenecks. (This is not meant to preclude centralizing management responsibilities within an organization for efficiency, but rather to prevent the notion of centralization across the whole CoIN.) The fundamental idea is to move true responsibilities as close to the optimum parts of each organization as possible.
Directory-enabled network management of the enterprise local area network (LAN) is offering centralized management and control of desktops, "relieving" users from individual self-configuration of their personal computers (PCs). An example of this is the Microsoft Active Directory (MAD) , being provided within Windows 2000. However, despite some similarities between this use of a directory and directories used to enable CoIN for e-business, it is important to understand the main difference. Within an e-business CoIN, the community is typically very large and spans many enterprises. The CoIN, by its secure connectivity infrastructure, provides a means to distribute access to a single logical directory across all the member organizations. (Of course in practice, the directory may be implemented as a number of synchronized distributed directories.) In the case of the MAD and enterprise desktop management, the benefit for most organizations comes from the centralized control of software loaded on networked machines by one administrator in one place. By contrast, in the case of the directory-enabled CoIN for e-business, the benefit to all CoIN members comes from the distributed access by many people to a single, logical, and common standards-based directory across many enterprises.
Within the common directory, each enterprise can "self-manage" its own organizational profiles. This facilitates the inter-organizational management autonomy which is essential. The distributed parts of the directory in each enterprise enable the process automation and information sharing that are fundamental CoIN properties. Further, any organization may use the directory to ensure that policy set by high-level management is implemented by lower-level management and followed by all members of their organization.
Policy could drive many aspects:
Software agents, for example, could be used to negotiate around policy conflicts among cooperating organizations within the CoIN. Adaptive management techniques may also be useful to cope with changing circumstances. In this way, future directory-enabled CoINs will support autonomy mixed with cooperation within and between organizations in entire supply chains. Key to achieving this will be the continued development of distributed directory technology. This development will include addressing issues such as scalability, linkage to directory-enabled applications, and security.
As discussed already, the true sharing of data (having a single source under the responsibility of the party who really owns it) is a basic fundamental of the CoIN approach. But replacing the "give and take of information" with "single truth data" will not be sufficient in tomorrow's e-business world. Instead of simply granting access to a single source of information, semantic labels are attached to the sources. Dictionaries can then be built of standard labels. This is already happening to some degree by initiatives such as the RosettaNet  collaboration of organizations.
The meaningful information (or knowledge) exchanged between organizations facilitates "total e-business" as data can be exchanged between business processes and typically organizational departments within a company. It also supports end-to-end process automation (given distributed workflow engines) as data can be meaningfully interchanged between organizations along the entire supply chain.
In business, especially with the arrival of the information age, knowledge is increasingly becoming synonymous with power. It is therefore possible to imagine a value associated with knowledge and the sharing of it with other parties. Preserving the semantics of data between organizations is absolutely key to this value. But as we know to our cost, storing and managing data can be very expensive too. The need to carefully analyze and assign ownership to basic data is essential. The use that an organization makes of this data to derive valuable knowledge for its business then becomes an explicit strategic issue. Indeed, once the transition to the type of future e-business described here has taken place, there is undoubtedly potential to re-engineer organizational structures to suit this new operational regime.
The extended markup language (XML) family of technologies supports this semantic labeling of data and conversion between business language used by different organizations. There is also support within this family of technologies to customize the appearance and presentation of the semantically tagged data. This provides for customization (e.g., automatic abstraction) based on content.
Extended communities for e-business may span a global reach. There is therefore an inherent need to ensure that the inevitable rich diversity that occurs globally can be accommodated. It is insufficient to be "connected" by a ubiquitous IP network if the community you belong to neither understands you, nor you it. There are a number of areas of diversity to contend with, including:
Each of these areas deserves a paper in its own right. However, the technology described so far to facilitate these extended communities can also be used to help cope with global diversity. Examples of future applications of such technology include:
International bodies could take responsibility for the global standards that would be needed together with this technology.
BT has significant experience on which to base this research work. It has one of the largest intranets in Europe and is a global supplier of IP networks to other organizations. Extranets have been implemented in a number of areas, such as between the marketing group and contracted advertising companies to share and communicate campaign media. BT also supplies IP virtual private networks (VPNs) to global companies and is an established customer of other organizations that lead the way in e-business (e.g., Cisco).
One of the first outputs from the research described in this paper was the application of principles such as dispersed responsibilities, single truth data, and simple process automation within a CoIN to the wholesale telecommunications business (the "Carrier Services" market of traffic interconnect) . An industry trial was run over nine months involving five telecommunications operator organizations, in what is one of the most deregulated and competitive telecommunications markets, the United Kingdom. The trial covered organizational and cultural aspects as well as technology. The results showed that inter-organization communication was streamlined and costs of operating processes for all CoIN member organizations were reduced.
Current lab experiments are applying a directory-enabled CoIN combined with inter-organizational workflow to a new mobile communications network. The CoIN members here include Service Providers, Network Providers, End Users, and Contract Suppliers/Service Providers. Dispersed responsibilities are allowing End Users to carry out some service management and network management operations directly and the business processes operating the service have been designed in a radically different way from traditional approaches. User organizations may also set policies for service provision within their organization, which is then activated and enforced by the network and service management systems.
In addition to this practical application of the research principles, some related work has been modeling  how businesses might be organized in radically different ways to suit the future scenarios of global supply-chain e-business described here. This modeling has been set in the context of some PEST analysis .
The natural consequence of much of the extended e-business approach of the future described here is the whole re-engineering of organizations from the top down, to reflect new cultures and arrangements for e-business. Although as yet unproven, those who take the risk and move to this way of doing business are likely to radically reduce their costs, harness the trust of suppliers and customers for reliable repeat business, and improve their levels of service.
Indeed the flexibility of organization needed to cope with this type of e-business (as well as the configurability of the supporting systems and processes) will also permit those who embrace it to move quickly to take advantage of new businesses and markets as opportunities arise.