denotes premium content | Dec 4 2008 Feature
posted 1 Mar 1999 in Volume 2 Issue 6
In the Know: Knowledge Management
using an information management framework
The information framework
implemented at Barclays Bank integrates space to capture both factual
information and the less defined unstated knowledge travelling through the
specialist vocabulary of a vast and complex institution. Here, Roger Evernden
and Neil Phillips explain the logistics and benefits of moving from a
one-dimensional document based focus to a multi-dimensional model which promotes
knowledge sharing.
By tradition financial institutions have invested heavily in information
technology for transaction processing which has led to a strong
data-orientation. Barclays Bank, along with other leading banks, recognised the
need to shift from this narrow focus on data to a broader focus on information
and knowledge. To help make this transition the corporate banking division
invested in Information FrameWork (IFW) - a project organised by IBM that has
involved more than 150 financial institutions worldwide. The Enterprise
Modelling team at Barclays used Information FrameWork as the foundation for
modelling information and capturing knowledge. Information FrameWork provided a
'starter-kit' that included an information architecture, generic definitions of
information items and information types, models of business rules and management
theory, and a methodology for managing corporate knowledge. This article
describes some of the experiences of Barclays Bank in using an information
management framework approach to knowledge management.
Information FrameWork
(IFW)
Information FrameWork (IFW) consists of four components - an information
architecture, a set of information models that populate the architectural
framework, a methodology that describes how these models can be used in
different types of projects, and consulting services to support the approach.
The information architecture is based on a wide variety of sources, the most
important being an Information Systems Architecture (ISA) developed by John
Zachman and often referred to as the Zachman Framework, and some innovatory work
by Westpac Banking Corporation during the 1980s. IFW was first defined in 1992
in a project that involved IBM Ireland, four leading European banks and a
consulting group specialising in information models. Originally populated by a
data model, the content of the framework has been extended year-by-year by
adding further models covering information about functions, workflow, and
organisational strategies, structures and skills. One of the fundamental
principles behind the IFW is the recognition that about 80% of the information
is very similar across organisations. It is the 20% that is distinct and
differentiates one company from another. The information models are therefore
designed to cover the 80% that is generic or common, allowing an organisation to
rapidly customise the models to include their individual needs.
Information is a
critical resource for many contemporary organisations. Information management is
a practical tool for managing certain types of corporate knowledge in an
explicit form. An information management framework such as IFW includes an
explicit statement of the principles and policies that are used to manage
information in an organisation.
Data and
information
As with many financial institutions, the focus at Barclays Bank was
initially on data. Their requirement was to construct a data model that
represented Business Banking information. Business Banking therefore purchased
the IFW models in 1997. The initial objectives of the data model included
representing the overall needs of Business Banking, providing a consistent
definition of business operational information and providing the internal
structure of a data routing system. The data model and other IFW models provided
the foundation for creating a valuable intellectual asset for the bank. In
business terms the data modelling efforts were focused on improving access to
operational information across the various customer service and product support
areas.
Seven
workshops, each lasting between 8 and 10 days, covered representatives from many
of the main business areas. At the workshops the IFW data model was modified and
extended to become a Business Banking specific model rather than a generic
industry model. The result was a clear picture of the operational data needs of
Business Banking that served as the business requirement for a data routing
system.
Without
IFW as a starting point, the development of the Business Banking data model
would have taken considerably longer to produce and the resulting structure
would have been predicated entirely on the current environment rather than also
including a view of future data needs. The ability to use a generic language
structure, whilst initially foreign to business representatives, provided a
means to overcome those language barriers that develop in any large
organisation. The finished model has since been used by a number of other IT
projects, each time saving valuable time when defining business
requirements.
Using the data model proved the value in defining information needs and
information structures. The Enterprise Modelling Team recognised that the
information needs and personal knowledge within the bank extended far beyond the
types of information that were covered in the data model. A data model generally
covers information that is factual and relatively easy to define and structure.
The EMT realised that a lot of knowledge in the bank was less precise, more
often unstated or tacit, and therefore much harder to model. There was a need
both for a simplified view of the business that could support planning and a
dynamic multi-dimensional decision support tool that took into account both
well-structured, precise business rules and less well-defined, emergent theories
or mental models. There are many ways to depict an information framework. Some
ways emphasise the information architecture and principles for managing
information and knowledge, while others emphasise a business user or systems
implementation perspective. Barclays are developing an Information Framework
that emphasises the business user perspective. The resulting information
management framework (figure 1) is based on 7 conceptual categories, each of
which is inter-linked with the other six types of information. Customer
information is at the top of the diagram to reflect the importance of this type
of information at the bank. Customers access products through various channels,
and communicate with people and business units supporting functions in different
parts of the organisation structure. Information systems provide essential
support for all of the other information categories. Information in all
categories is used to define the strategy of the bank, while the strategy of the
bank also informs activities in each of the conceptual areas.
Figure
1
Barclays are currently focusing on the Function component of the
Framework, developing an understanding of the processes and data that underpin
the functions and the associated roles and skills for those processes. The model
enables the components to be viewed in both their established terms (i.e. as
Barclays currently describes the processes, skills, etc.) and also in the
generic terminology of the IFW language. Having the components described in a
generic language is facilitating re-engineering opportunities and enabling a
greater understanding of the critical aspects of those components.
Organisational
change
An example of the broader use of an information management framework is
the way in which the Enterprise Modelling Team used information modelling to
manage knowledge during the restructuring of the corporate division of Barclays
during 1997-1998. The restructuring followed the sale of the equity and
corporate finance divisions of BZW in 1997. The Enterprise Modelling Team
gathered information about the strategies and structure of both old and new
business units. Traditionally this type of information is stored in documents
and charts; often using word processing or graphics software to control changes
or manage printing and distribution. It is difficult to compare the strategies
of one group with those of another when the information is kept in document
format because of differences in the structure of documents, inconsistent use of
vocabulary and the absence of contextual and explanatory notes that often remain
as tacit knowledge.
The alternative approach was to define a lexicon of terms used
in the bank, and then use these defined words and phrases as a standard
grammar for capturing information about goals, objectives, action plans,
projects, and organisation structures. High-level concepts are used to
analyse information about organisational strategies using the m1 software from Modelware.
If there is any disagreement over terminology, then the word or phase in
dispute is deconstructed into smaller elements that can then be used to
distinguish one meaning from another. For example, the word 'customer' may conjure up a
picture of a person, an organisation, an account, a relationship or a
role. Deconstructing 'customer' would break the concept into each of the
constituent parts. Whenever anyone uses the term 'customer' the model can now be used
to clarify the sense in which the term is used or to compare one meaning
with another. Note that the model is not used to force standardisation but to
clarify meaning and manage divergent opinions. If one interpretation of 'customer' is
better than another, then the reasoning behind this judgement can also be
captured in the model. By capturing opinions, experience, theory and practice
the model becomes a learning tool to help an organisation identify the most
effective ways to structure and manage information.
During the restructuring of the bank
the model was used to clarify the goals and objectives of new business units and
to compare the functions of one unit with another. The model became a reference
point and mechanism for improving communication and discussion between teams as
the new organisation structure emerged.
| Dimension 1. Types of information It is easier to manage information if it is grouped into conceptual categories, such as information about people, about places and locations, or about strategies. There are links or associations between each of the conceptual categories. Inappropriate conceptual categories lead to information structures that are inflexible. 2. Levels of understanding People understand information: By recognising words and language and the meaning that they convey By applying mental models and theories to explain links between one piece of information and another By applying their knowledge and experience in practice to test current theories and discover new mental models and learning 3. Representations The same information content can be represented in different ways. For example, trend information might be shown as a table of values or as a graph or chart. Some representations are more appropriate than others in different contexts. Sometimes an inappropriate representation is deliberately used to mislead or to misinform! 4. Transitions over time Information changes over time. Information can become out-of-date, information can be 'revealed' over time. 5. Tacit / Explicit Some information is explicit, but much knowledge is actually unstated or tacit. Knowledge management is partly what you know, but also what you don t know or don't know you don't know. This dimension includes the management of ignorance! 6. Processes of using information Information is only valuable when it is applied and used. For example, information is used in business or project processes. Sometimes the use of information is explicitly defined in a process, such as the use of information in a data warehouse development methodology. |
Using an information management framework
The two major techniques used by Barclays are the adoption of an information management framework and the use of a model-based approach to information management.
The information resource of a financial institution is complex and vast. The information needs of various involved parties are varied and subject to change. An information management framework is designed to keep track of such a complex resource by providing a high-level structure and a disciplined approach. Information exists in multiple dimensions shown in the table.
The information management framework at Barclays explores some aspects of each of these dimensions. The Business Banking structure shown in Figure 1 has gradually evolved over time as more information and knowledge populate the framework. The conceptual categories used in this structure represent different types of information. Different types of representation have been used, including representations borrowed from information engineering and object oriented methodologies as well as documents, diagrams and intranet formats. There is an awareness of the balance between tacit and explicit knowledge, including a use of the four Nonaka and Takeuchi modes of knowledge transformation to ensure a balance between socialisation, externalisation, combination and internalisation.
One of the big overheads in using an information management framework is creating an awareness of the approach within the organisation. The most effective way to do this is through the practical application of the framework. Information is used in nearly every project that takes place at Barclays. The more frequently the framework is used on projects the easier it becomes to promote the principles of information management.
Model-based information management
A model-based approach for information management means that each information item is described in a consistent and rigorous manner. The discipline and structure of a model is designed to aid understanding and communication. There are two major activities in using a model-based approach - (1) managing the model itself, and (2) publishing information from the model. If the model has been well-designed then it should be possible to create published information of a quality that is significantly better than that created without the use of a model.
There were a number of problems with the information models that Barclays had used in the past. Problems included how to gain agreement on a consistent definition of business needs while simultaneously managing divergent opinions and viewpoints; a lack of definitional clarity within the model; little provision for decision support; the constraints of using paper-based models rather than software-supported models; and poor communication of the value and benefits of using a model-based approach. The new approach at Barclays provides integration between an information model to define and understand diverse needs and the use of decision support tools such as simulation models and systems thinking models that are used to explore the underlying theories and practice.
There is a strong document-based culture in many organisations, where the document is the prime vehicle for publishing information. Moving to model-based information storage and processes involves a number of changes including the need for model management processes to develop information models and distribution mechanisms for publishing information from the model. Figure 2 summarises some of the changes from the document-based to a model-based approach. It also describes some of the changes that result when transferring information ownership from individuals to the corporation that occur when an information management framework is introduced to an organisation.
Figure
2
Modelling information in the context of an information management framework has brought a number of significant benefits to Barclays Bank. It allows information to be used and reused in multiple contexts. Publishing information in a document tends to promote single-use information; the document serves a specific purpose and it is difficult to reuse the information in the document in a new context except through the creation of another document. By modelling the information first and then creating published extracts from the model, the modelled information is always available for subsequent projects or future needs. The information model and framework improve communication and understanding, especially across organisational groups and functional boundaries. Analysing the information needs of business units or projects makes it easier to identify overlaps and improve collaboration.In the long run, the information and knowledge that are captured can become a significant corporate asset. As more Barclays knowledge is added to the models (a process that one manager called 'eagle-ising' the model after the Barclays eagle logo), the model becomes the source for achieving sustainable advantage and distinguishing Barclays from its competitors. The 'eagle-ising' process provides the basis for a realistic information re-use strategy. As an example, the outputs from Barclays current modelling work (described earlier) is already being targeted for use by a number of business and IT projects. The use of the generic IFW language within the models will enable business solutions to be implemented consistently across different platforms, channels and business areas.
Conclusions
The Barclays approach has a strong focus on information (not information technology or information systems) as a vehicle for explicitly capturing and reusing corporate knowledge, so much so that the Enterprise Modelling Team in Barclays is part of the business change function, not a part of the Information Services business area. The information assets of the bank represent personal knowledge that has been gathered from many different projects and business units. The focus on information was important for keeping discussion about knowledge rather than becoming side-tracked by technology issues. Because of the different types of information that the bank wanted to manage and the large number of information items, there was a need for a clear framework or structure for capturing and managing information and knowledge. Although the bank was able to purchase a licence for the Information FrameWork, this structure had to be customised to suit the context, language and culture of Barclays Bank. The framework presented within the bank had to be based on principles that can be explained and understood by everyone who comes into contact with or uses the information management framework. The generic nature of Information FrameWork provided a good foundation that could then be populated with specific knowledge about Barclays Bank. Populating the framework with information and knowledge required a team dedicated to enterprise modelling, clearly defined information management principles and policies, and senior management commitment to the value of information as a key knowledge resource.
Copyright (c) 1998 Roger Evernden
Roger Evernden is a specialist in information and knowledge management. He is a founding director of WorkSpace International supplying information management frameworks and information models. He can be contacted at:WorkSpace@compuserve.com
