The advisory and sales process in banking requires the availability of comprehensive information and knowledge. The gathering of this knowledge is a difficult and time-consuming task. Advisers usually have to search several internal information systems and various external news and information services to collect the needed information. Sometimes it is mere coincidence that they come across some highly relevant information, and presumably there is some amount of relevant information of which they are never aware.
Today vast amounts of knowledge are available in the Internet, in company-internal databases and elsewhere. In banking -- like in many other industries -- competitive pressure is increasing continuously. The winners in this competition will be those who manage to use the existing knowledge effectively to reach their business goals. The challenge is to identify and structure it and make it available for use in business processes.
In cooperation with several banks, the Institute for Information Management of the University of St. Gallen has developed an intranet-based solution for supplying the advisory and sales processes with knowledge. Information about the customers, products, and competitors from internal data sources as well as external information like political and business news, stock quotes, market surveys, and so on are provided to the adviser in a process-oriented, structured way. The adviser has the necessary knowledge at his or her fingertips and can concentrate on the customers' needs.
In the context of this project we have developed a methodology for distribution of intranet-based knowledge. The methodology focuses on the conceptual aspects of knowledge analysis, knowledge representation, and intranet design. The methodology does not cover software engineering aspects or the deployment of a basic intranet infrastructure. The methodology comprises four essential steps that are necessary for establishing an intranet to support a given business process with corporate knowledge.
This paper describes the methodology in detail and shows its application for supporting the advisory and sales processes in banking. Each step of the methodology is explained and illustrated using "real world" examples from the project. The project results, risks, and benefits and the critical steps in the project are discussed. The development of the methodology in cooperation with banks enabled us to continuously validate each step and ensure the suitability of the methodology for use in practice.
The advisory and sales process in banking is very demanding and requires highly qualified personnel. To be able to carry out this process to their customers' satisfaction, the employees need to be well-informed about products, customers, political and economical situation, financial markets, and internal regulations. Most of this knowledge is available in company-internal information systems, on the Internet, or in the heads of some colleagues. However, advisers usually spend a lot of time looking for relevant information to answer a specific question or to carry out a certain task.
This paper describes a systematic approach for making knowledge available to employees in a process-oriented way. An intranet-based solution for supporting the advisory and sales process is used to illustrate the individual steps of the methodology.
In the context of this paper, the term knowledge is used as a generic term and includes all types of information found, for example in databases, in documents, or in the form of people's individual knowledge in their heads. The terms information and knowledge are used interchangeably. The concepts described in this paper are based on the view of knowledge management as explained in .
Knowledge management is not fundamentally new. However, recent developments in information technology have removed implementation barriers. Particularly Web technology -- in the form of intranets -- allows the integration of different media types and makes knowledge globally and concurrently accessible.
There are various methods and concepts available that deal with the management and distribution of explicit knowledge using intranets (e.g. , , , , , ). The management of implicit knowledge is discussed in classic knowledge management literature like , , , . Most of these sources focus on the creation of new knowledge, on the generation of corporate-wide knowledge networks, expert maps, yellow pages, and so forth. However, the concepts described do not focus on using implicit knowledge in business processes.  discusses knowledge management as an enabler for business process change, but focuses on organizational learning.
Existing literature leave the following questions unanswered:
What is novel in our approach is the clear focus on a specific business process that is to be supported with knowledge. It is the utilization of knowledge in the business process, which creates value for the enterprise. Therefore, the business process determines what knowledge is required.
This work has been conducted within the competence center "Business Knowledge Management" as part of the research program "Business Engineering HSG."
The research method used for this work is action research . This means that the theoretical results have been applied in projects at partner companies. The observations made in these projects have then been used to further improve the theoretical results. This research method ensures that all results have been verified in practice and are therefore suitable to solve practical problems.
The methodology provides a systematic approach to support business processes effectively with knowledge. It is focused on organizational aspects, namely on (1) the analysis of business processes and knowledge sources and (2) the design of support processes and knowledge structures. Methodological aspects concerning the implementation are not addressed as there are many suitable software engineering methods available (e.g., ).
Intranets are especially suitable for implementation. Therefore, several aspects of the methodology -- especially the part dealing with content management -- assume that the solution is implemented as an intranet. A systematic approach for the design of intranets is described in , .
The method is based on the principles of Method Engineering . The advantages of a systematic approach follow:
Figure 1. The methodology consists of four essential steps.
The methodology describes four essential steps that are necessary for designing an intranet to support a given business process with knowledge as shown in figure 1. In the following description of each step it is assumed that a business process has been selected to be supported.
Business process analysis and design has been discussed exhaustively in the last decade. Many methods for business process redesign have been created that claim to be applicable to any process. However, in practice the use of these methods was usually restricted to transaction-oriented processes, which are highly structured and standardized. In most companies there are many so-called knowledge-oriented processes that cannot be described by a simple sequence of tasks. These processes include, for example, research, development, or sales processes. While "classic" BPR approaches in general focus on making use of transactional data, other types of information like documents, books, pictures, or Web sites are often left aside.
The procedure described here is suitable for knowledge-oriented processes. The first step is to analyze the business process for required knowledge input and for generated knowledge output. The knowledge requirements consist of all objects listed either as input or output. For the process analysis, it is irrelevant whether the knowledge is available in documents, in databases, from external information providers, or elsewhere. For each knowledge input or output, the exact contents and the way the knowledge is used in the corresponding task have to be described in detail.
There are several ways to identify the knowledge requirements. Existing process documentation can be studied, employees involved in the process can be interviewed, and existing information systems can be examined. The process analysis must aim at finding knowledge requirements that help to improve the process in the future. It is not sufficient to list the information currently used in the process; the challenge is rather to identify those additional knowledge requirements that can be helpful later on.
To illustrate the methodology, a generalized advisory and sales process as shown in figure 2 is used throughout the paper. The process has been derived from real processes in retail and private banking. This process is a typical knowledge-oriented process. The tasks are not necessarily carried out in this order, and depending on the individual situation, several tasks can even be completely omitted. For example, it is not necessary to negotiate the conditions for standardized products with fixed prices.
Figure 2. Knowledge input and output for the advisory and sales process.
Each task is connected to the corresponding knowledge input and knowledge output. For reasons of clarity, figure 2 shows only some exemplary knowledge requirements. For example, to prepare for a customer contact, the adviser needs access to a customer profile (name, address, currently used products, an so on), and he or she needs to be informed about the times of and reasons for the last contacts with this customer. When the adviser has learned new facts about the customer or has sold a product to him or her, the adviser generates knowledge output when updating the customer profile and the contact history. This knowledge is then available as knowledge input to the same and to other processes later on.
In the second step, the determined knowledge requirements are matched with the existing knowledge sources. All existing knowledge sources are analyzed for contents that meet some of the specified knowledge requirements. Each relevant knowledge source is listed, described in content and structure, and categorized according to its characteristics. Each knowledge source can be assigned to one of the following types:
Each determined knowledge requirement must be supplied with at least one knowledge source. It is possible to assign sources of different categories to one requirement. For example, product information about investment funds can consist of documents describing the structure and investment policy of the fund plus the actual development of the fund price from an external source in the Internet. If a knowledge requirement cannot be satisfied using the existing knowledge sources, a new knowledge source has to be specified. To minimize maintenance costs, it is desirable to look for external sources that satisfy the requirement.
Especially challenging is the description of the expertise-type knowledge sources as they deal with so-called implicit knowledge, namely, knowledge that is only available in the heads of employees. Frequently, experienced employees are not aware of the valuable knowledge they have because they take it for granted. On the other hand, novices would be able to do their work much better if they knew about the experts' knowledge. This dilemma imposes the necessity of developing a comprehensive interview concept that considers a representative number of employees of different levels of experience.
Figure 3. Knowledge sources are matched with knowledge requirements.
Figure 3 lists the relevant knowledge sources for the example process. Again for reasons of clarity, only several exemplary sources are included. For example, customer profiles and contact histories can be found in the customer information system, which is a relational database. A document management system contains sales presentations and instructions for the sales employees. Product knowledge is stored in a product information system, but the advisers themselves possess extensive implicit product know-how. Finally, to get the latest financial news, an external Web site is accessed.
Once the knowledge requirements and knowledge sources have been analyzed, a knowledge structure has to be developed. The procedure for knowledge organization described here favors an implementation as an intranet. This allows building a navigation structure on top of the existing knowledge sources. Thus, most sources can be reused without extensive modifications. The total cost for the project can be reduced this way.
However it is obviously not possible to make implicit knowledge directly available in an intranet. There are several possibilities for the organization of implicit knowledge:
A navigation structure can be built in various ways. A combination between a process-oriented and a subject-oriented navigation structure has proven successful. The process-oriented navigation provides links to the relevant knowledge sources for each task of the process. The employee is guided through the process and has the needed information at his or her fingertips at any time. However, this navigation structure fails when the employee needs to access knowledge that is not directly related to the task he or she is currently carrying out. Therefore, a subject-oriented navigation structure has been created that provides access to the same knowledge sources. To complete the navigation structure, an index and a search engine are provided to the user. The index enables the user to find contents by keyword; the search engine provides a full-text search.
Figure 4. Process oriented and subject oriented navigation structures.
Figure 4 shows parts of the navigation structure that have been designed to support the example process. The process-oriented navigation on the left-hand side shows three exemplary tasks of the process. For each task, several links are listed that point to the corresponding knowledge sources. The totality of these links satisfies the knowledge requirements as determined earlier. The subject-oriented navigation on the right-hand side includes various topics. The News-links lead to company-internal and to external news, the Customers and Products sections give access to the corresponding information systems, the Processes section provides information about the way the processes have to be carried out, the Experts section enables the users to access the expert directories and maps, and the External Links section provides a systematic collection of selected relevant external Web links. An index is included that gives access to the information by a specified set of keywords, and a search engine allows the user to do a full-text search.
Support processes have to be defined for creating, distributing, maintaining, and removing the contents of knowledge sources. Responsibilities have to be defined for carrying out these processes. The concept for content management is based on the framework for intranet management, which was developed earlier by the authors . In the following paragraphs, only the changes and additions to this framework are described.
Content management is necessary for all three of the types: data records, documents, and expertise. While data records and documents can be managed as described in , expertise has to be available explicitly to be managed this way. The content management processes, especially the information maintenance process, depend on the way the expertise has been codified. For implicit knowledge that is converted to explicit knowledge, the knowledge owner is at the same time the author of the document. Regular verification and update of the codified contents by the author is necessary.
For expert directories and maps, there is on the one hand an author, who generates and coordinates the whole document. On the other hand, all experts have to provide information about their knowledge in the form of knowledge profiles to the author. In this case, the maintenance process deals with keeping the whole document up-to-date (e.g., when experts leave or when new experts arrive) and keeping the profiles of each expert up-to-date.
Figure 5. Information maintenance process for managing expert maps and directories.
Figure 5 shows the activity chain  for the information maintenance process that has been extended to meet the specific requirements of maintaining expert maps and directories. The process is triggered when (1) an expert or the author of the map or directory notices the necessity for an update, (2) a regular verification is scheduled, or (3) a user suggests an update. The author verifies the existence of all listed experts and the completeness. The author asks each expert to verify his or her knowledge profile. The experts forward any changes to the author, who updates the document adequately.
In a similar way, the information creation process has been adapted. While the author creates the document (i.e., the directory or the knowledge map), the knowledge profiles have to be created by the experts themselves. The processes for information distribution and removal do not have to be changed, because the individual experts are not involved. The whole document is being distributed or removed in these processes.
The results of the research project are the methodology described above and an intranet prototype that is continuously being improved.
The main benefit of the intranet solution according to test users is the structured access to relevant information. Much of the information was available in an earlier intranet solution, but it was hard to find the right information as needed due to a confusing structure. The users expect the quality of the consultations to be increased due to process-oriented navigation structure. This enables all advisers to base their work on the same information and to carry out the advisory process in a more structured way. The facilitated access to experts for certain products or customer segments will help the advisers to solve problems more quickly.
Whereas the analysis and structuring of explicit knowledge is a straightforward process, incorporating implicit knowledge in the intranet solution is much more challenging. To ensure the usefulness of the intranet, contents must always be up-to-date. This is especially difficult for the explicit representation of implicit knowledge as only the knowledge owner knows when the contents change, and frequently he or she is not aware of the necessity for an update. The support processes described above are designed to ensure actuality of the contents. In these processes, the experts are regularly forced to verify and update the contents for which they are responsible. Deployment of the solution in practice will show if these mechanisms are sufficient.
In this paper we described a methodology for supporting business processes with corporate knowledge. The individual steps were illustrated with examples from an intranet-based implementation to support the advisory and sales process in retail banking. One of the main challenges was to make available the implicit knowledge of experts to all employees.
Although first experience has been gained, there are still various unanswered questions. To start with, it is unknown if the knowledge structures are sufficiently scaleable for deployment in a large organization. For example, expert maps for one process or one department can be managed comparatively easy, but only the future will show how this works when the maps grow. Another unanswered question is, what type of incentives have to be introduced to motivate the experts to share their knowledge. Furthermore, it will probably have to be part of their job description to be available to other employees as experts. Without profiting themselves, they will not be ready to "waste" their time helping out some strangers.
The intranet platform is continuously being improved and extended to support other business processes. In the future, questions about knowledge leadership, incentives for knowledge sharing, and so forth will be addressed in depth.
On the more technical side, much effort is needed to make existing knowledge sources accessible from the intranet, even though Web technology facilitates integration a lot. Therefore, another research topic will be to develop a reference architecture that facilitates the integration of existing knowledge sources into an intranet solution.