Arthur Andersen Consulting (AA) is a global consulting firm organized in four main divisions that offer tax and legal, business consulting, financial and auditing services to medium and large size target clients. This services are provided by nearly 120.000 professional consultants that operate across 74 countries, and are organized in a complex matrix given by the intersection of several target industries (e.g. telecommunications, automotive) and service categories. Country specificity, industries and service categories are the basis of a work content that completely relies on the use of knowledge as the primary resource of value. That is to say, a consultant's work is primarily given by the ability to apply business knowledge to solve a client's business issues or problems. AA consultants are thus facing a particularly and increasingly complex, dynamic and differentiated environment. As a result, knowledge used by consultants becomes increasingly complex, dynamic and differentiated to effectively respond to emerging business needs.
The information and communication technology (ICT) revolution and the related explosion of business complexity has influenced the knowledge needs of consultancy and, as a consequence, the nature of a consultant's work. The revolution raised new issues and questions:
The opportunity of ICT to store and communicate information at a low cost represented the natural path to solve this knowledge managing issues. AA gave different concrete answers in terms of managing knowledge within a technology enabled environment. Briefly we describe some main approaches, which we interpret from an evolutionary perspective.
In a first approach, named the Knowledge Base (KB) approach, knowledge is seen as a set of formal representations of given knowledge domains. The latter are given by industries and service categories while the former by documents such as methodologies, presentations, articles and engagements descriptions. Knowledge is collected through a centralized process of gathering, organizing and formalizing context specific experiences that through generalization can be used in other contexts. Formal repositories of knowledge are called Knowledge Bases, initially distributed to consultants on a CD support and later made directly accessible through a corporate Intranet. Specialized organizational entities played the role of knowledge producers, while consultants played the role of knowledge consumers.
A second approach, the AAOnLine approach, arose from some serious limitations displayed by the KB approach. On the one hand, the attempt to produce generally valid ``knowledge packages'' generated oversimplified contents that were practically useless, except as high level descriptions of consultants' knowledge domains. From this point of view, knowledge consumers didn't consume centrally packaged knowledge. On the other hand, it became evident how consultants' daily experiences and tacit knowledges weren't completely packageable. Rather then collecting and formalizing these experiences, AA decided to make them accessible at a source level. The basic idea was to give consultants a global discussion forum were each consultant could discuss, ask and give information about any topic related to work. This global forum, named AAOnLine, was organized in communities of interest, each community being a virtual space where consultants with a common interest could collaborate and share their individual knowledge. Each community was organized by a Knowledge Manager who had the role of supporting and enabling the interaction within the community.
However, the AAOnLine approach displayed some limitations as well. After a first period of enthusiasm, the number of participants and contributions begun to decrease. This reduced the quality of available contents, which in turn discouraged participation and contribution. A possible explanation of this partial failure is that AAOnLine was a virtual space where people who speak different languages (not just intended as country specific) could met without having enough technological and conceptual tools to understand each other. Context specificity, in the terms of linguistic and semantic differentiation, was represented by the specificity of individual contributions, each contribution being characterized by a particular use of language. Nonetheless, this diversity remained implicit and unexpressed, so that accessibility to individuals contributions didn't correspond to accessibility to individuals' interpretive perspectives. In other words, within AAOnLine consultants were asked to read other consultants' contributions without having the ``linguistic key'' to interpret the intended meaning of used terms. Within a physical environment, this opportunity is ensured by the dialectic process of meaning negotiation that is hardly reproducible within an asynchronous communication environment like AAOnLine. As a result AAOnLine, rather then solving the problem of linguistic and semantic heterogeneity, became the expression of its apparent irreducibility.
The KB approach made evident the need to consider knowledge as a tangible resource, organized and represented in a way that fosters its accessibility and, as a consequence, the value of its replication. The limit was an oversimplified view of the process of knowledge production that underestimates - or perhaps ignores - the problem of context specificity. On the other hand, the AAOnLine approach made manifest the need to consider contexts as a constitutive dimension of knowledge rather then a ``noise'' that interferes and negatively influences knowledge production. The limit was the lack of both technological and conceptual tools to sustain meaning negotiation and knowledge exchange by making contexts accessible through an appropriate representation and interaction process. As a result AAOnLine recognized the value of contexts without providing tools to transform this potential in actual value. We may say that the KB approach focused on knowledge as a product without considering the nature of its generative process, while the AAOnLine approach focused on the constitutive elements of knowledge production without providing enough tools to configure it as a reusable product.
The need to consider knowledge both as a product characterized by accessibility and replication and as a process characterized by context specificity and meaning negotiation is the current process of AA knowledge systems. Nonetheless, although the issue has now been made explicit, the solution is still to be found. Currently, AA is experimenting a ``third way'' that tries to merge the two issues, called the KnowledgeSpace approach. KnowledgeSpace is a corporate Intranet that is characterized by the following features:
AA is our case study in the development of an innovative KM system. The paper focuses on new conceptual and technological tools that can support knowledge representation within a community and knowledge integration through meaning negotiation across different communities.