Georgetown University

Georgetown School of Business

MGMT550 Information Technology and Business Strategy

Organisational Learning

Ewan Sutherland

Amongst the many methodologies for planning information systems the following are representative of the approaches taken.

Business Systems Planning (BSP) is a well-established methodology developed by IBM for both private and public sector organizations. It is a structured approach to planning both short- and long-term information systems requirements, derived from experience within IBM and based on attempts to create an all-encompassing corporate database.

Critical Success Factors (CSFs) were originally intended for use by information systems managers in dealing with senior executives by Rockart. It enabled senior managers to express their needs in terms of the few factors which were absolutely critical to the performance of their job. The IS manager could then use these to propose information systems which would meet the needs of the senior manager. The methodology was subsequently extended to cover planning for all information systems for a company.

Portfolio theory was adopted from financial management—the use of a ‘portfolio’ of projects information systems to balance the risks.

In addition to the more formally planned areas of information systems, there is a wide range of unplanned and sometimes anarchic uses of information systems, such as the adoption and use of personal computers, notebook computers, personal digital assistants, cellular telephones and facsimile machines. These operate under few if any organizational controls, being funded from many different budgets and by many different individual managers.

Beneath the level of information systems planning and resource allocation, there is a need for the detailed work of systems analysis, design, construction, delivery and implementation. Information engineering is the term used increasingly to describe the activities of building information systems, in more formalized ways than in the past. Although information engineering ought to be based on the application of information science, it relies instead on the codification of empirical knowledge, putting it firmly in the craft tradition. Examples of the move towards engineering practices have been the adoption of prototyping and re-working of software already in use.

Information engineering methodologies can be grouped under the following headings:

Structured programming
From the original ‘hand-crafted’ or do-it-yourself style of programming, more systematic approaches have been developed. In order to improve the productivity and the reliability of the ‘code’ generated by programmers, more rigorous approaches were developed for the construction, layout and documentation of computer programs.
Systems analysis and design
Following the success of structured approaches to programming, more rigorous approaches have been developed for systems analysis. The systems development cycle is now well covered by reliable and tested methodologies for investigations, analysis, design, implementation and review. These methodologies address issues such as overall planning, quality management, project management and security, using well-established techniques such as data flow diagrams and entity-relationship models.
Software engineering
A major factor in the drive for improved productivity was the shortage of staff and especially skilled staff. Considerable successes have been achieved through these of software engineering, especially through Computer-aided Software Engineering (CASE).
Human computer interaction
Again ideas for human computer interaction (HCI) started from relatively ad hoc beginnings but have gone on to a more formal basis. Part of the formality has been achieved through market forces which have driven companies towards the use of Graphical User Interfaces (GUIs), such as Microsoft Windows in imitation of the highly successful Apple Macintosh, itself based on work at Xerox PARC.
Formal specifications
There has been a growth in methodologies intended to specify systems in ways which are either provably correct or which can be compiled into runnable code.
Project management
Information systems have come to rival civil engineering projects in terms of their reputed failure to deliver on time and on budget. This is often aggravated by delivering something markedly different from what the customer expected, with the problem of the expectations of the customer changing during the production phase. Therefore, there have been considerable efforts to develop systems and supporting software which will help in managing large software projects. Systems such as Hoskyns’ Project Manager Workbench have been developed to support project management. Project management software has also been integrated into Computer Aided Software Engineering (CASE) tools.

Stages of Growth

The scope and complexity of the tasks being handled by IS managers has grown enormously over the last forty years. Following the example of Nolan and Dearden [1973] it is conventional to show an IS management learning curve with stages of growth. These authors took a somewhat Marxian view with the changes being brought about by crises in the management of information systems, giving the model an apparently solid theoretical feel. The model was later revised to take into account subsequent crises, their solutions and criticisms of the model (see table).

The original model comprised four stages:

This was drawn against the background of an S-curve. The basis for this was "organisational learning":

Growth of specialisation of personnel

computer efficiency
operators, programmers and systems analysts
development of applications
systems programmers, systems analysts and applications specialists
control and effectiveness
maintenance programmers, functional systems analysts (e.g., finance)
database and on-line
database and telcommunications experts

Management techniques

lax management
no priorities, FIFO, no chargeout
pro-active sales orientation
few priorities or control, informal project control
control by centralisation and a steering group, growth of importance of maintenance, imposition of budgetary controls, chargeout, quality programme
EDP as a functional area, growing specialisation in EDP department, refinement of cntrols and planning
The revised model (see table) was again an s-curve. One important difference was the y-axis was labelled data processing expenditure.

Table Nolan's revised stages of growth

StageApplications portfolioDP organizationDP planning and controlUser awareness
InitiationFunctional cost-reduction applicationsSpecialization for technological learningLaxHands-off
ContagionProliferationUser-oriented programmersLaxerSuperficially enthusiastic
ControlUpgrade documentation & restructuring of applicationsMiddle managementFormalized planning and controlArbitrary accountability
IntegrationRetro-fitting applications with data base technologyEstablish computer utility and user account teamsTailored planning and control systemsLearning about accountability
Data administrationIntegration of applicationsData administrationShared data and common systemsEffectively accountable
MaturityMirroring of information flows in systemsData resource managementData resources strategic planningAcceptance of joint accountability with DP
[Source: Adapted from Nolan (1979)]

Sometimes in stage 3, therefore, one can observe a basic shift in orientation from management of the computer to management of the company's data resources. This shift in orientation is a direct result of analyses about how to put more emphasis in expanding DP activities, on the needs of management control and planning as opposed to the needs of consolidation and coordination in the DP activities themselves. [Richard Nolan]
Guidelines for action from Nolan et al. were:

An overhaul of the stages of growth model has recently been attempted.

Too many people have fallen into the trap of sophisticated copying of this linear model, including this author. Experience of IS and of the management of IS gives rise to a whole matrix of learning curves, with different technologies and in different parts of the organization. The idea of ‘development’ ceased to be useful when the applications of information systems became too diverse to be mapped onto a single curve; there is no easy way to combine, say, experience in managing mainframe-based on-line transaction processing with end-user experience in electronic publishing. The issue is further confused by the different aims and strategies of subsidiaries and strategic business units which, with different business goals, may require different uses of information systems.

Critical Success Factors

Table Differences between management and IT/DP

ManagementInformation Systems
Little knowledge of information systemsLittle knowledge of business and management
Limited abilty to express equirementsNeed for detailed specifications
Business orientationTechnical orientation

Problems in the late 1970s for determing CEO data needs:

supply of data as a by-product of OLTP
cannot predict needs (therefore why bother?)
Total study
comprehensive study of informations (e.g., IBM BSP)
Key indicators
infoprmation on 'health' indicators of the business
OLTP = On-Line Transaction Processing

John Rockart of MIT proposed Critical Success Factors (CSFs):

... those few key areas of activity in which favorable results are absolutely necessary for a particular manager to reach his or her goal ...
CSFs need special and continuous attention. They provide critical areas without providing detailed development plans.

Sources of CSfs

In summary, CSFs appear valuable as a means ofbuilding a conceptual model of the key facets of an organisation or a of a manager's role in an organisation. Such a model can then be used to drive the requirements analysis process. However, because certain managers may experience some difficulty indealing with conceptual thought processes, the CSF methodology may not be universally appropriate. Because rather concrete thought processes are required to arrive at a detailed specification of information requirements, the CSF method might not, by itself, be a effective requirements analysis tool.

[Boynton and Zmud]

Guidelines for use CSFs

A disturbing observation

Second, the CSF exercise provided senior management with important information for strategic planning efforts. The explicit identification of CSFs sharpened management's understanding of those facotrs central to the firm's success. Corporate reorganisation efforts were accelerated by the CSF exercise, and they are likely to be more successful because of it.

Boynton and Zmud

CSFs for MIS Managers

How does IS performance affect overall business performance?

Background: problems and massive growth and spending:

Portfolio approach

Based on the work of F Warren McFarlan (Harvard Business School)

Failure by IS managers to:


Dimensions of project risk

Risk declines as the project proceeds.


Portfolio Risk Profile

In addition to determining relative risk for single projects, a company should develop an aggregate risk profile of the portfolio in the abstract, there are appropriate risk profiles for different types of companies and strategies.

F Warren McFarlan

Project management

external integration
communication links between the project team and the user management and community
internal integration
ensuring that the team works as a team
formal planning
to give a structure and sequence to the tasks including allocations of money, time and technical resources
formal control
means to help project managers evaluate progress and failure to progress

External integration

Internal integration

Formal planning

Formal control

The right approach flows from the project rather than the other way around.

F Warren McFarlan


Benbasat et al. (1984)
Boynton & Zmud (1984)
Britcher (1990)
Brooks (1975)
Cash et al. (1992)Cohen et al. (1986)
Davis (1979)
Dijkstra (1976)
Earl (1989)
Galliers (1991)
Galliers & Sutherland (1991)
Gibson & Nolan (1974)
Harrison & Thimbleby (1990)
Humphrey & Kellner (1989)
IBM (1984)
McFarlan (1981)
Martin (1984)
Martin (1985)
Munro & Wheeler (1980)
Nolan (1979)
Nolan and Dearden [1973]
Olle et al. (1991)
Rockart (1979)
Rockart (1987)
Rockart & Crescenzi (1984)
Sommerville (1992)
Steinholz et al. (1990)
Sutherland (1991)
Thimbleby (1990)
Yeates (1991)
Zmud & Lind (1985)