Information Systems – Fundamentals and Issues
©1990-2006 John Lindsay, Kingston University, School of Information Systems
To be a designer of information systems it helps to start first of all with a sense of fury about the world. A sense of fury when you stand in a queue waiting to buy a television licence. A sense of fury as you go round and round an organisation trying to answer a simple question. A sense of fury at the waste of people unemployed, when stopped in the street for a few pennies for a cup of tea; a sense of fury at expenditure on useless waste of advertising or weapons when basic needs for clean air or good housing cannot be satisfied. (But to warn of future complexity we must allow that the sense of fury might revolve around "there are too many blacks in Britain", or "the little yellow blighters are out to get us", or "homosexuality will destroy the family and christian values"!). Second, It also helps to have a sense of one's own ridiculousness, otherwise the tension produced by the fury has nowhere to go; for ranting and raving at end of day results only in bad temper and drinking too much coffee. Third it helps to have a curiosity, to desire to explain why things are so, which of them need be so and which of them can be changed by the conscious intervention of human beings?.
But this is not a text on general philosophy or general problem solving. It deals with information systems, their design and building and their impact on the organisation within which they exist. It grows out of a period when technical change - our capacity to master nature - is accelerating, yet our ability to consciously control that change seems no better than the middle ages understood the Black Death. It deals with a period when technical change has affected the way we process information. (As substantially different from what went before as movable type changed the world of Gutenberg or Copernicus). The effect on organisations is as great as that of movable type on the Holy Roman Empire or the theatres of London.
This technical change is exciting. Organisations are becoming more complex, problems more tractable. Boundaries evaporate as radio signals fail to recognise customs control posts. New boundaries are created as copy protection devices are fitted to software. Some technologies flower as others wither. Some organisations bloom, some fruit, others stunt despite all the fertiliser. Some stay infertile, some barely begin then die.
Every project and every system has some beginning, though it grows out of its own past. An initial text for an undergraduate course presumes a world from which the student comes, with a grounding possibly in a language such as english and a language such as mathematics (so that expected starting points for recruits would be "must have A levels in english and maths"!). These we might called the novices and the structure the noviciate. An initial text for a postgraduate course, or what are increasingly called conversion courses, must have a different starting point. One of two goals is intended: initiation into an established order, or synthesis of different orders. The alternative is the detailed exposition of the narrow domain of the specialist.
Similarly, embarking on designing an information system has some beginning. It is either you, the agent, which is having a beginning, or the institution, which is having a beginning. You might have started a new job, or a new course. Your organisation might have been bought, opened a new factory, signed a new contract, installed a new generation of computer (if we are starting in a post hoc world) or decided to plan to buy a company, open a new factory, tender for a new contract, install a new generation of computer. This might give us our first design parameter: are we concerned principally with proaction or reaction?
It is not often we start with a tabula rasa or a greenfield site. More often we say, "If I was going to that place I wouldn't start from here." But it is from here that we have to start. So it is of "here" that we have to know the location. The definition of here is the process of giving form and meaning to the problem, of problem definition. This is partly also the bounding of the solution space. The problem space and the solution space are the same place. It is either a place we know, well mapped and supplied with visuals, or a place which we don't know - a terra incognito, where be dragons. This might be the second design parameter: are we dealing with the well known - the established - or the unknown - the giving form to? In one sense every project is both, as is every day; in another every project is the synthesis of the two: mapping the known onto the unknown in order to give shape and meaning to it. People make history in conditions not of their own making.
If here has location so has that place. The concept of here and there and journey gives us a dimension of scale of time and distance. Already we have a third design parameter: scale. Modelling some order of magnitude of the time and distance involved in the project gives us an idea of the size of the sheet of paper or the granularity of the map. This book is a one year project, that course is a two year project, this budget is for one year, that plan is for five. This involves one person, me, that involves one hundred people or ten thousand pounds. Information in an organisation has a life as old as the organisation. Land registration in Britain probably relies on information a thousand years old. Yet the life cycle of an information processing technology might be only five. Let us produce a scale for time and distance: information systems involving a period of five years are strategic, one to two years are tactical, less than one year are operational. This will not serve for all purposes, but will for now. Strategic involves all the staff of the organisation. It involves investment or expenditure approaching 10% of turnover. Tactical involves 10% of staff and 1% of turnover. Operational involves 1% of staff or 0.1% of turnover. Staff means staff time, rather than absolute headcount. It does not matter whether these figures serve for all purposes. They are here simply to give marks on the page. Scaling by orders of magnitude is a technique to which we will return in Chapter 5.
Similarly, we have to give a scale to the part of the organisation in which we find ourselves a part. This defines the boundaries of our ability to act. The organisation as system or sub system gives us a location for our actions and defines the limitations of our power. The model of the organisation reveals the complexity of the decision taking process of which we are a part. The range of players who have to be part of that decision- taking process gives us a scale for the likelyhood of the place we arrive at bearing any relation to where we thought we were going. The number of orders in the hierarchy above us and below us, the number of departments alongside us which have to be part of the process or which will be affected by the decisions we take: the quantities of these marks on the sheet of paper will be a measure of the scale of the complexity of the issues we are confronting.
The lines of communication and the effectiveness of the transmission of will and desire have scale too: time and distance, but also reliability. Just as an electronic signal travelling differs at its origin from its destination, so too does a human communication. The strength of the signal, the power differential between the parties, the resistance, other signals, and so on - the metaphor can be stretched almost boundlessly - but again here all we want is a sense of scale.
And scale will also involve the number of originators and number of recipients of messages to and from the organisation, and the number of messages. But to these matters too we will return later.
Let us now try to put this into a more formal method. Let us also presume that we are starting on a fairly substantial strategy exercise. This might be drawing up a five year plan or part of a systematic planning exercise. It might be initiated by a new product release, by a buy- out or by some set of events which can be modelled. It might be undertaken by one person doing a consultancy or by a group of people working together. It might be a very small and short assignment, it might be of considerable complexity.
But the point of the exercise remains the same - to identify the political, economic, social and technical factors which are significant to the problem under consideration. At one level it is simply a mnemonic. At another it contains a theory. The theory is that technical changes, changes in our ability to produce and reproduce our material conditions of existence, produce forces on our political, economic and social relationships which are resisted, to a greater or lesser extent by different groups of people in society as a whole and in organisations affected. These changes in the forces of production might be analysed either in the long term - wind mills, gun powder, electricity, the internal combustion engine, or in the short term - microchips, optical fibres, satellites.
In the short term organisations which grasp correctly the changes and organise accordingly, gain an advantage. A central theme of this book is that technical changes in the field of information processing are occurring such that they present organisations with new opportunities. This technical change is one of the key ingredients in the process of adjustment. When this adjustment is resisted it is at cost to the organisation concerned. This cost can at times be critical. But there is nothing inevitable about the form the adjustment should take or whether the adjustment can be made. The point of the PES- T analysis is to attempt to identify what is needed. Whether it can be delivered or achieved is another matter.
How the PES- T analysis is performed may vary. Techniques such as Delphi, brain storming, soft systems analysis might all be used. The point is to produce an agreed list of technical factors which are significant for the organisation. The thoroughness, reliability and time span will depend on factors already identified in 1.2. This stage might require research, a study of competitors, markets, patents, it might require industrial spying and all sorts of nefarious practices. The effectiveness with which it can be undertaken will depend on the thoroughness of the information gathering activities and competitor analysis of the organisation. It will also throw up a problem of defining what sort of business you are in. The drive of technology is towards disintermediation. The consequence seems to be that organisations lose a central sense of direction or market then have to recover or return to a core activity. But undertaking this process might enable you to grasp that this is what has been happening and take corrective action.
Next identify the political factors at play. Sometimes they might be a determining primacy. Changes in legislation for the single European economic community, 1992, might mean the opening up of markets which will change the competitor analysis you have relied on in the past. Changes in legislation such as the Poll Tax will affect you differently if you are a local government department, an employer or a tax payer. The House of Commons Select Committee into the airline ticketing systems is possibly the most detailed study so far into the politics of an information system
These factors might be subdivided as institutions, laws and regulations, agreements and protocols. The institutions might be sub- national, national or international. They might have compelling authority such as the ability to enforce imprisonment for selling old computers to the USSR, or the ability to exert pressure such as Toshibas burning on the lawn of the White House. Included in the list might be intellectual property rights, standards, data protection, import controls * and add more.
In their study of an information map for the United States McLaughlin & Birinyi, produce a chart of legislation affecting information systems the work of John Morris provides a similar attempt in Britain. The rate of change of these forces, and how they shape up for any particular exercise, makes for the building of a picture which should not be too complicated to be unmanageable. The danger is to be too narrow, not too wide.
The meaning of political might be wide: that involving the distribution of power in and among groups of people, or narrow: the process of government. An example of a political map is shown in fig 1. The temptation to make the map too wide and all embracing at the early stages should not be resisted - it is part of the process of giving form to the issues you are confronting and can always be narrowed down later.
Another danger is that the people forming the map don't have the power or influence to do anything about it - they are just doing what they are told.
It is also worth commenting at this stage on the political nature of systems, and the systemic nature of politics. While resisting the temptation to digress into political science, it might be worth drawing attention to government attempts to "deregulate" television, "privatise" telecommunication, water and power supply. To these points we'll return later in Chapter 5.
Next go on and try to identify the economic factors affecting your organisation. Some of these should be part of the systematic central decision support system of the organisation. Interest rates, capital costs, currency fluctuations, inflation rates, wages, raw material replacement prices, product market prices and, margins should all figure in the decision support system. Here it is more useful to try to identify in a non-systematic and non- quantitative way factors which you are trying to identify.
The social factors which you need are of two sorts, those external to your organisation which are part of the society within which you function, and those internal to the organisation.
It is also worth indicating at this stage that it is necessary to show that there are a number of different goals for each of the players. To presume that an organisation has goals, that it is goal seeking or goal satisfying is simplistic and romantic. To fail to recognise that an organisation consists of people, and that these people have goals of their own, and that these goals differ, should not however be taken to the length that Checkland for example takes it. Though many players can design many games for themselves at work, organisations are hierarchical power structures where some goals, such as maximising rate of return on capital employed, winning market share or such forth dominate, and people do get sacked or made redundant. Their games get broken up. (See Chapter 1 - notes)
The process of articulating the hidden agenda of the organisation might be implicit nature of real goals.
The scale of the operation determines the scale of investigation which is necessary at this stage. It might be essential to have opinion or organisational "fact" from large sections of the organisation. Interviews, surveys and questionnaires might have to become part of a formal process of determining the social map of the set of groups with which the study is concerned.. This strategic level should be separated from the level of drawing up the requirements specification if viewed in a classical "life cycle method" project cycle. Yet at this stage the very process of conducting an investigation is already an intervention into the life of the organisation. Something is going on. As soon as this gets talked about informally the process of design has started.
Another mnemonic. A method for analysing an organisation developed out of common sense is to mark in four boxes the strengths, weaknesses, opportunities and threats of the organisation. This might need to be done at three stages if a check is to be given on the identification of the social groups indicated above. First, you as the strategist undertaking the exercise have ideas and interests of your own. Second, the group you are immediately a part of has interests of its own. Third, according to each level of the hierarchy there will be interests not necessarily conforming with one another. According to the form of the organisation these might be in conflict with one another. Mechanisms of management such as profit centre, cost centre, reporting unit, decentralised decision taking, centralised decision taking, will all influence whether groups work in coherence or opposition. This level of the study might build on the point I made earlier about the relation between organisational form and performance. At this stage you might want to draw on the formal organisational chart or organogram of the organisation, and it formal statement of mission, goals and objectives, to the extent that these documents exist. We are not at this stage setting goals, or yet evaluating ones that exist; merely trying to establish that they exist and what they are. These judgements from already existing documents and from people, either in the group of interviewed, should give an estimation of how the organisation sees itself at the moment, how the groups see themselves, and how you see yourself in the group and organisation.
A fairly obvious design exercise is to attempt to do this to yourself in your current position now. Do not be surprised if you have difficulty deciding whether something is a strength or a weakness. This is a view on reality and depending on how you view it, the object under investigation will be moving - no study of any organisation is static - so it will be either a strength becoming a weakness or vice versa. In the process of this stage of the study check against the PES-T study you undertook earlier and see what points you have left out. Every significant point from the PES-T study should appear in the SWOT. Every item identified in the SWOT should appear in the PES- T as having an agent as an owner.
So far you have been describing things as they are and as you see them as being significant. Now the time has come to try to see whether you have correctly identified the significant, and put them in the right place. You must now be able to ask the question: "What game am I in?" "Who are my competitors?" In part you'll have begun to do this by the process of PES- T and SWOT, but now you want to formalise this into a model. The idea of the five forces comes from the work of Michael Porter, based on several hundred case studies from the Harvard Business School, but it has become so widely used that it now approaches the common sense.
His argument is simple: that every market can be studied as a relation between itself, its buyers and suppliers, the possibility of new entrants and that of substitutes. See fig n. for a diagrammatic representation. Porter in this earlier work makes mention of the role of information systems, but gives no detailed worked out indication of how this is to be done. He return later with Millar to discuss the issue of information, but in no detail.
Porter's analysis gives insufficient attention to the role of information systems in changing the boundaries of intra- industry competition. This is most elegantly shown in the famous Ernst & Whitney advert: If the retailers are into banking, what should the bankers be into? A substantial piece of work has been undertaken on what he calls disintermediation by Alex Stewart on Japan, but studies in Britain appear still too narrowly focused. The argument though is simple: changes in technology are changing the nature of industries. Telecommunications and broadcasting are an obvious example. Both these industries have recently been involved in political battles over the relationship between the private sector and market regulatory mechanisms, so using them as case studies provides a lot more meat for PES- T and SWOT analysis. Banking and insurance, estate agencies and banking, airways and travel agencies.. the list is almost endless.
Similarly moving beyond intra- industry competition to the relation between the supplier, the industry and the buyer shows that there is both a shift in the bargaining power of the players, in terms of how trading conditions may be determined, and in the patterns of ownership and preferred agency relationships. Information systems involved in Tradernet, EDI, airline booking such as Amadeus and Gallileo, EFTPOS such as Connect show the particular influence that information plays in this process
Fairly obviously in turn, the changes in information processing are allowing for new entrants into a market, or increasing the cost of erecting barriers to exit. A property developer could now set up a telephone network using Centrex exchanges, if he could build enough sites in line of sight. The banking network in Spain and Portugal which none could afford not to join, most could not afford to join. The consequence will be a process of amalgamation and centalisation. Building societies in Britain offering a range of financial services wider than the banks don't have the distribution of branches. A network of intelligent terminals in retail outlets fills the gap. Substitutes become available as the technologies of information storage and communication change. CD-ROM provides and alternative to on-line access. Videotapes with VCRs provide an alternative to broadcast programmes. EFTPOS provides an alternative to AmEx..
So defining your five forces model will provide a useful check on your earlier analysis. Have you identified the technological changes which are capable of opening up the organisational changes? Have you identified the dynamics of change correctly right from the beginning? Does your political analysis identify regulatory instruments, spoilers, lobbying groups that you have to be related to? Does your SWOT realistically identify the opportunities and threats which arise from an analysis of the industrial chain in which you are involved? Do you want to change any of the earlier parts of the analysis at this stage?
What will be beginning to emerge is the difference between where you are, as a snapshot, and where you thought you were, as a snapshot. Because change is a continual process, this analysis will have identified the trajectory on which you are moving. The SWOT analysis might be changed a from point on a chart to vectors. But in order to do this you need to know more about the competitors in your industry than you will be able to draw on from common sense.
Competitor analysis needs to be a formal process in the building of the five forces model if it is to be more than a snapshot. Channon gives the most complete listing I know of. He does not however deal with how this might be handled in a systematic manner. To this I'll return later. (In Chapter 6) If your organisation is incapable of providing this information systematically then you have no alternative but to reply on the opinions of the people engaged in the exercise impressionistically, and hope to be able to return to this later.
It hardly needs mentioning that collecting information of this sort might become an end in itself. To have too much information, or to be paralysed for want of it are two sides of the same coin. Information becomes the object of debate because the organisation is incapable of taking the decisions the organisation requires.
There are of course a huge number of consultants who will do all this for you. Market reports and strategic studies may be bought, commissioned or purloined. Knowing which of your competitors are engaged in such an exercise is an elementary part of corporate intelligence. Getting hold of their strategies a simple process in espionage.
This technique too draws on Porter's work and has been widely sited and studied. Less frequent are examples of it being used. There is a general problem in this area of work, that where companies think they've got their strategic level work right they are unlikely to talk about it in public or make themselves open to research. I've had specific rebuffs from Logica and Saatchi and Saatchi for example who say that their use of information systems is their source of strategic advantage so they aren't going to talk about it.
The argument is quite simple: those operations which add value to the product under consideration should be managed to maximise the value being added. Operations which are not adding value should be turned around so that they do, or disposed of. At one level this follows from the five forces model. Within the company there will be activates for which other parts of the company are buyers or suppliers, and for which there will be external but intra- industry competition, and outside the company there will be buyers and suppliers who will be studying their relationship with you in the same way. From the organogram of the organisation it should be possible to produce a value chain for every process through which a product passes in the organisation.
The famous model is reproduced as figure 3.
The most obvious problem, to which Porter passingly refers, is that information isn't available to build the value chain. Organisations don't know the unit cost of their operations. They can't relate undertaking one operation and determining whether it adds value or not. As a simple design exercise try undertaking this analysis on your own organisation.
Further, the very nature of the organisation makes it difficult. Because inputs and outputs from the macrosystem are spatially and temporally disjunct from each part of the process, carrying figures through which have any useful meaning requires the definition of protocols across an organisation Because each part of the organisation has a goal and set of definitions different from other parts of the organisation the very seat of where this study is taking place can be seen as an attempt to change the power balance or bargaining balance of this part of the organisation to that of the whole.
Attempts to formalise this process by the building of an internal market serve only to make the situation more difficult. Organisations function on the basis of an informal market economy phrased as "You scratch my back and I'll scratch yours." Because organisations don't exist to perform only the goal set by the owners or managers of the organisation, but to fulfil all sorts of other internal goals, identified during the S part of the PES- T analysis. Undertaking this part of the analysis might, however, establish units of currency and conversion protocols or exchange rates which would be recognised and carry the weight of meaning.
Nevertheless the tool remains a useful one. It becomes more useful if added to it is the idea of direct product profitability. Rather than just measuring inbound, process and outbound, overheads and margins, this tool allows for the adding of a temporal dimension to the analysis. In manufacturing this has been recognised by the JIT principle and the reduction of stock
The area where these issues have caused the most difficulty is in the public sector. In some sense or other the basic goal of the management of most private sector organisation revolves around a return to shareholders (dividends), rate of return on capital employed, rate of return per employee, or some such other unit of measurement. The release of the figures for the major banks, oil companies and other major industries in Britain once a year provide both an aggregate measure of their performance and a source of curiosity about current levels of wage settlement. Within the public sector however no such obvious unit of measurement exists. In addition, because there is a much larger employer and wider range of bodies responsible, divisions make analysis and allocation of information map positions more difficult.
Recently there have been a number of attempts to produce financial units of control and rewards to achievement. The function of the market and the private sector in the market is a topic rather wider than the scope of this work will allow, but some points must be made. Even if absolutes, such as "bottom lines" cannot be formulated, disciplines such as constructing value chains may still allow for comparing strategies. If there is more than one way of doing something, and one way is cheaper, then it needs a convincing argument why another should be chosen. If something is being done in the public sector, then the argument will be made that it should be done in the private sector. Telecommunications, water, electricity, health, education are currently newsworthy, rail, roads, coal, and soon one expects the environment and air will be. Again it is not the purpose of this work to engage in polemic for the point of it, but to look at the systematic formulation of strategy for an organisation. It is not the purview of the systems approach to dwell on morality or ethics: whether this or that is in and of itself a good thing or worth doing, but it must be within the purview to discuss whether something is more or less systematic, or the organisational implications, and therefore the information processing implications of favouring one path or another. The direct decisions which will be taken following the decision to adopt a particular strategy will inform the design functions of the system. It is difficult to do a stupid thing well. If some or many of the participants believe an activity to be stupid, then it will be difficult to design it to be well done. The traditional literature on economics is not particularly helpful n this domain, discussing a thing called a natural monopoly, but without a discussion of why that particular thing should be natural and not another.
The common sense of competing water pipes or competing electricity cables scampering up the garden path towards one's front door can be grasped. That rainwater falls equally across small spatial areas and therefore cannot be collected individually in densely populated areas can be grasped. Yet the logic does not translate to the movement of bulk goods, as we see by investment patterns of successive governments in roads and railways. The logic does not follow to the regulation of newspapers or the regulation of broadcasting companies.
The systems designer should desire a logic and systematicity which can be translated from microsystem to microsystem,. Disintermediation will build a desire (or an opportunity) to link different systems. But the absence of common units of measurement makes for the building of a value chain in the public sector difficult. To this point I'll return later.
This technique, developed by the Boston Consulting Group, requires you firstly to specify your product range. It then argues that according to your position in the market, using the sort of evidence which Channon would draw on, your products fall into one of four sorts. The method usually requires the drawing of a diagram in the form of a square with axes marked for profitability against time or market share against market growth rate. See figure 4 for an example.
This method is identified with the Sloan Management School It works from first recognising the strategic business units of the organisation. This recognition might come from the organogram, or from the sort of analysis suggested by Wilson, or from the PES- T analysis. Each of these SBUs will have goals or objectives and the units of measurement by which it will be indicated whether they have been achieved or not. The critical factors are established by which these objectives will be achieved, or will fail. However Rockhart makes the point that the method is based on the perceptions and information needs of individual managers.
The analysis is a process of consensus building with the agreement of all participants being needed to establish that a factor is critical. The extent to which resources must be made available to satisfy the criticality helps to distinguish between a CSF and a wish list of niceness or motherhood.
The methods (1.2-1.8) above come predominantly from the management literature. As I've indicated very seldom do they either indicate the implications for information systems design, or recognise that changes in the capacity for handling information are changing the roles of business. In fairness most of them date from the sixties or early seventies when the dynamics which we're considering here were apparent.
There are in turn a series of methods which come out of the tradition of designing information systems, taking the perspective that the writing of "tight code" ought not to be the measure of a good information system. The dynamic of these traditions is a combination of the dissatisfaction of users with historically developed systems and the complexity which follows changes in the economics of information processing.
Presume Osrey is key text ? hunt something down. or Rockhart Harvard Business Review 1977. (See Avison & Fitzgerald )
A management consultancy, which took as its benchmark a very specific insistency on measuring information technology investment against corporate return on investment. They proposed a mapping of investment against value added See fig 5.
Their work for the British Government in 1984 on actually trying to develop a method for why some organisations actually made better use of information technology started the process whereby it was the use of "information technology" which was investigated rather than the use of "information".
The work of James Martin in Information Engineering has produced a powerful argument on the relationship between business requirements and information systems design. There has been a certain feeling that chiliastic enthusiasm makes the method a religion. The production of the Information Engineering Facility and the work of various JMA's (if that is the name for an associate!) have been influential in for example IFIP WG8.1 CRIS and ESPRIT RUBRIC. It appears to remain true however that the argument for rigorous engineering following business analysis still doesn't allow you to determine what you should be doing rather than automating what you are doing. The insistence of James Martin on the primacy of data driven modelling seems a bit of a fight with a ghost, for the work of Chen and Backman, exemplified in Learmonth and Burchett's successful tender for the British Government's SSADM rather predate's Martin's work.
It is not the intention here to develop a history of how ideas of structured systems analysis and design progressed, simply to indicate that from the tradition of data analysis or systems analysis a dynamic was produced that forced the question" what are you doing and why". In addition attempts to automate the process of developing code led to attempts to automate the analysis stage to the design stage. We'll return to this issue later. The logic of the argument is shown for example in fig 6
Strategic value analysis suggests the recognition of information systems in changing the business or changing the methods, but other than suggesting that a team of highly experienced and valuable staff should be doing it, and listing what they should be doing, gives no indication of how it is to be done, or how you might have a sense that it has been done correctly. see fig n(p13).
This summary of methods is rather too obviously Anglo-Saxon in approach. It needs mentioning that there is a vast literature in other parts of Europe. The work of Nijssen in developing the NAIM methodology, much taught in a number of Dutch universities might be given as an example. My argument would be though that the emphasis on mathematical rigour is a return to the technical domain, removing information systems from the social. If part of your system is irreducible to formulae, then you lose the proof benefit of the rigour in another part. There is no value in having nonsense to n decimal points if it remains rubbish.
Some methods limit themselves to the evaluation of current systems without making any pretentions to given tools for guidance on how to move forward. Are you doing what you are doing with maximum economy is all they attempt to answer. Compass is such a method, and suggests that by means of a set of measurements against an industry norm, you can determine whether you are on target. See fig 7.
All these methods have in common that they are attempting to make systematic a process which evidence suggests involves the black box of design. The role of the irrational and un systematised in human affairs comes out of each of them when the issue of evaluation is raised. How can you tell whether the analysis is right?
All these methods have in common that they are top down. They start from the presupposition that there is an organisation, that it has a level of coherence, hierarchy, structure and purpose (though I have attempted to indicate the points at which the subjectivity of individual players in the planning process will interfere, and that (...)
They also have in common that they are Non- evolutionary. Though it should be possible to put in place a process whereby the setting of objectives and the evaluation both of these objectives and performance against them is continuous, in practice the planning process goes through cycles. A common classification is the strategic, on a five year time frame, the tactical on a one to two year frame, and the operational on a less than one year frame . The point must be made that these time frames are too long for fast changing markets. It used to be that the fitting out of a motor car factory would take 10 years. Now it will take n months.. It appears that the nature of large bureaucratic organisations is precisely the limiting factor, so many companies have set up relationships with small entrepreneurial organisations which are capable of fast adaptation, or allowed (or encouraged) management buyouts. In turn though the small companies either grow and become bureaucratic, die, or get bought out.
Worked out case study on the School of Information Systems?
It is only reasonable that I give a worked out example of using a variety of these tools to formulate an information strategy. Given that I am also suggesting the centrality of the individual actor in the process it is only reasonable that I give my own School as a case study. It must be recognised that this is not however a fully worked out case study - just a summary. In some ways it is also more difficult than you might encounter. A semi- public authority organisation at the heart of a political and managerial dispute will test the methods severely. This study was written in August 1989, so reflects the analysis and concerns of that time.
The School of Information Systems is one of four schools within the Faculty of Technology at Kingston Polytechnic. The organogram of the Polytechnic is shown in fig 8.
Here is the first design issue: do you construct an information strategy for the Polytechnic, the Faculty or the School? There are two answers to this question. The fact that "Polytechnic" will be developing a system of its "own" indicates that there is a dissonance between "School" and "Polytechnic". Secondly there is the scaling problem for which a group of people might be said to "own" and "information system". The performing of a PES - T analysis helps to resolve these questions. Because the political level of the PES- T starts in the environment outside the organisation, but which forms or shapes the organisation, we can start either with the Polytechnic or the School.
The Polytechnic is a Higher Education Corporation, set up by the Government under the Education Act of 1988. It exists "in order to support the economy through provision of career- oriented higher education and advanced training". It exists as a consequence of the Conservative Government removing higher education from local authority control, or giving freedom to higher education institutions according to your position. The trajectory is towards privatisation and the placing of higher education in the private sector or towards improved public access to a public service through improving management and market relatedness according to your position. Both these positions might be true at the same time.
Along with the changing status of the institution comes an attempt by the employers to change the working conditions of staff in higher education. In the past one taught a number of hours per week for the academic year. Under the new proposals you will work for a number of hours a week, for the whole year, with leave. This is either to give more incentive and to reward the more dynamic, or to reduce costs overall by increasing throughput.
Government policy appears to be not to favour the expenditure of public money on education but to desire to increase the throughput of students into higher education.
Role of PCFC as co- ordinator of government policy?
Role of the employers in attempts to change national negotiating procedures for staff
It is on the junction approximately of the M3, M4 and M25, half way between Heathrow and Gatwick airports.
Role of PCAS and UCCA as central processing facility for placing students
Impact on costs of national shifts in wage bargaining
Kingston is an attractive riverside area with an historic centre and stunning traffic jams. Polytechnics do not have the "status" of universities (quite correctly for we are trying to do something different) but students and employers might have an impression of this.
Social position of higher education and preparedness of people to take themselves out of the employment market for three or four years in the short term for perceived longer term advantage
Role of CNAA legitimating degrees
Role of professional associations legitimating qualifications
Developments in computer assisted learning and distance learning offer opportunities for changing the spatial and temporal nature of education. Developments in information technology are changing what is taught and the way it is taught in almost all disciplines, and changing the boundaries among disciplines.
Summary: there is an interesting contradiction between the Polytechnics offering a "cheaper" form of education in a more "expensive" area.
There are therefore some points on which the School will be dependant on the Polytechnic, some factors which cannot be changed.
One can in turn do a SWOT analysis on the Polytechnic.
For whom? How is goodness to be measured? Are we looking for income enhancement? Academic reputation?
To whom and from whom? Control of staff over their working environment and conditions? Security of jobs?
But already one sees that there are distinctly different positions available according to whether you are the Director, or someone working in the place. The value and reward systems of these practitioners will differ widely.
Now let us move to the School. A second design point emerges here: the Faculty appears to be irrelevant. Were we to do this analysis it would appear to have no real organisational role. It appears to fulfil a management function only. It has no source of loyalty among staff. It appears that a layer of administration is put in place only for administrative convenience or because information systems are insufficiently robust. This in turn reopens the question of the management relationship between the directorate and the teaching staff. This question will not go away. The argument will be put that you cannot have a Director relating to thirty something Heads of Departments. But what does this director and these heads of departments do? Teach, Research or Generate income?
The political, economic, social and technical for the School are the same as for the Polytechnic, except that as this school is at the heart of information systems the technical which I chose to mention favour us more than others. The PES however are not quite the same because each of us individually, and as an effective group - about fifteen who personally know one another and have, presumably some loyalty to one another, have our political relationship with the employer tied up in our economic ability - our wages, and our social - the fact that we like working with one another, all sharply intertwined. A change in working load affects the personal relationships.
It is also worth mentioning that the origins of the School reside firmly in one person, now the Lucas Professor of Engineering Design at the Open University, and that since December 1988 when he ascended to his new Chair, we have been without a Head, and have had for that period also an acting Dean of the Faculty.
Good reputation and shared coherent view of the discipline. Colleagues get on well together
Working in a little known discipline where we stand or fall between a powerful business school and "computing" which everyone knows and understands. What is information systems?
Absence of powerful player in politics of organisation as a whole.
Opportunity to generate income through consultancy, to get research grants
Attitude of the Polytechnic administration to our friendly working environment;
Differing attitudes to income and abilities to generate them leading to tensions as some people get paid more and others have to do less rewarding work.
Here another design point has emerged: the SWOT analysis for a component organisation might well have to indicate that internal competitive pressures are greater than external.
In order to understand the SWOT analysis a chart has to be drawn up indicating what these swots are being measured against. For the Polytechnic it might be other institutions of similar standing or in the same area - Surrey University, Kingston College of Higher Education and so forth. For the School, it is our attractiveness as a School in a discipline, so we can't compete with Templeton, LSE, LBS, Henley, Ashridge, Warwick, or can we? In an emerging discipline like this, our activity is as much a matter of making the discipline and forming the professional institutions as it is teaching. In turn the ability of our students to get prime jobs will depend on our success for we are not Oxford or Cambridge and cannot survive on social credit.
In order to be able to effectively swot the problem we also have to be able to outline the product line, for all products might not be easily located if the swot is understood as dynamic rather than static.
Research and consultany
Here is a design question: is the unit of organisation we are examining actually capable of being examined separate from the parent organisation? Is it realistic to say anyone comes to the Poly because of this School - no I suspect. Is it realistic to say we work here because of the School, not the Poly? In part yes. The subjective and the objective comes together at this point.
Another design point: is there an optimum size for an organisation such as ours where success depends so much on effective bonds? Is there an organisational form for an institution involved in higher education which would make it more effective than another form?
The VCA for the School and the Poly is at one level fairly similar. For the teaching courses there are students who are selected, join courses, attend lectures, seminars and tutorials in groups of various sizes, read books, use computers, take examinations, pass and fail, attend graduation ceremonies and get jobs. For teaching courses and short courses there are staff who prepare and give lectures, set and mark examinations, organise, plan and develop courses. For research and consultancy there are staff who prepare projects, win or lose them, undertake them and write reports. All this activity operates at the level of the School. However it is difficult to value or cost these activities because there is no unit of currency which converts the sources of income from the PCFC, via the unit of staffing, the full time equivalent or FTE and the staff- student ratio (SSR), into a unit of time a member of staff spends on an activity.
It is therefore impossible to value giving a lecture against preparing a research contract. The significance of this problem will become clear later.
At the level of the Polytechnic administration there is no apparent mechanism at all for deciding what level of organisation is necessary to support a particular level of income generating activity. Allocation of resources to student welcome, careers, carpets in the Director's suite, the Chaplain and the gardens are all drawn off resources which could be otherwise utilised, but without protocols which justify expenditure. Library budgets, computing resources similarly appear to have a level of randomness to them.
There is a further point which arises from the particular nature of a semi- public sector activity. The product of the student does not have a market value or price. There is a fixed income per student therefore the expenditure on producing that student is a variable without an upper limit. Extension is through the elasticity of staff time. Nor is it proven that there is a relationship between "value" of teaching and attraction of students. Similarly there is a problem of quality control, because unlike a factory where you specify production requirements and measure conformance, with education if you improve standards the norm simply shifts. There is a perception that within a cohort a certain percentage will be 1st, 2nd and 3rd class. If you increase the number in each division upwards then the perception is that standards have declined and the currency has been diluted.
Some appear fairly absolute:
So what, by now the determined reader will be saying, has all this to do with an information system?
From all this it should be clear that what is to be measured and the units of measurement have to be drawn from this analysis. Without this analysis we would have had no basis for deciding what to measure or on what units to do it. Without the analysis leading to strategic decisions we would have no basis for making decisions about the allocation of resources or how staff should decide to spend their time.
There are clearly a series of conflicts of interest. There is the impression that political decision taking by central government with an unelected and unrepresentative governing body will change the ethos of public service which will in turn change the need for units of measurement and reporting forms. It is also likely to produce an unreality of management figures at each level of the process. Either this will have to be taken into account and much flexibility will have to be given to the operating levels or a regime of controls will have to be introduced which is likely to be ineffective and bureaucratic (though that wont stop the attempt being made).
There is also a wide difference among staff in their ability to generate income, either for themselves or the organisation, with the possible consequence that others will have to carry an unrewarding workload. This too will determine the sorts of units of measurement and reporting forms.
This is not the place to move on to actually constructing the data flow diagrams or specifying the entities and relationships to be modelled. First we need to understand a lot more about the nature of information. I hope though this walkthrough has given some ideas on how to do it, what to look for, and why, without doing it, you would have been measuring things which wouldn't have been asking the right questions.
As this work is aimed at people who are studying the process of designing information systems it is worth making some comments on learning materials and the learning process. At the outset it is worth re-emphasising that information systems are social processes so studying them is most usefully done socially. It might be an academic course, a management team, or some such semi- structured or structured occasion. The material suggested should be viewed or read as the case may be, and often the best path to follow is to break the course into groups which go away for a while, study the issues, then come back and make presentations on their conclusions. Sometimes the exercise can be set up to make this role playing decision making.
It isn't possible to give more than a brief description of each of these and an idea of how to get hold of them. Hopefully the description isn't so epigramatic that it isn't clear why they are here at all.
This is an Open University Programme * get full title* which describes a food market information system implemented by the Brittany farmers agricultural union. Its starts in the 1950s when the farmers were very poor and the market dominated was by the trucking companies. If the farmers were efficient and productive, the excess of vegetables produced would lead to a fall in price with the farmers destroying vegetables to hold up the price. The key was that when any particular farmer set out for market, he had no way of knowing how many others were doing likewise, what the state of the market would be, and therefore the likely price.
The Union provided a computerised marketing system in three towns. The farmer phones in, says what he'll be bringing and what quality. He is given a guaranteed minimum price. The buyers are able to make a market on what will be available so they can still affect the price. The consequence of increased efficiency led to the need to increase the size of the catchment area, which led to the recognition that London was more important than Paris, the consequence of which was the formation of Brittany Ferries.
The advantage of this case study is that it shows the past and the present, with the case study worked out and solved. Watching the film, you should be able to:
This follows on well after the Brittany farmers as it presents a problem very clearly, but leaves unclear how it will be solved, or whether it can be solved at all. In addition it demonstrates an information system without computers which appears to have worked for about a thousand years.
The television programme was broadcast by * and made by *. It is available from * (find all this out). It deals with the work of a couple of US anthropologists looking at the Bali water supply system and the impact of pressure from UN agencies and the Indonesian government to increase output. The system of water supply has depended on the link between the water temples, the dams and the villages. The anthropologists model this on a Macintosh to show the various participants that what they are arguing is true.
You should be able to perform a PES- T at any time during the past thousand years, and one for the period of the making of the programme;
a SWOT for each of the players you have identified;
a VCA for the rice cycle and for the aid cycle (if you aren't aware of the aid debates and literature then Cousens' book will be helpful. This area of work will raise lots of interesting issues about the role of information systems and the development process.);
construct an information map;
you could try constructing a VCA, but this will need information on the food chain which you might have difficulty getting hold of - it's worth doing though;
a five forces model is unlikely to be fruitful when at the primary production end of the food chain, but by all means try it and see;
outline a scenario for each of the players you have identified;
you could go as far as constructing an information strategy for them, including most of chapter n* after you have worked through it.
This is a broadcast television programme, available from (*? find out). It follows on well from the previous two as the problem is much more complicated and unsolved. The programme deals with road and rail planning in the south- east of England, focusing on the M25. It shows how various information systems are being applied and discusses the consequences.
The best way to handle a case study as complicated as this is to work with a group of about twenty people, say on a course or in a management team, break up into four groups of five, and take on the roles of four major players right from the beginning. It should be fairly obvious which these are. There can be no good role in a scenario as complicated as this one.
You should be able to do a PES- T for each group and construct information maps. Presenting these at a round table should bring out the differences of opinion. If this is played in a series of rounds then groups could form alliances to try to win position.
You should be able to do a SWOT for the player you have chosen. This can be made more realistic by actually tracking down the necessary information the managers would require if they were engaged in this - it gives a chance to actually make use of your information sources in practice.
You will find difficulty with the VCA, but the five forces model should be clear. You could add to this case study the write- up done on a real study done for British Rail and critise it as part of the exercise. It seems to me to have missed all the opportunities after having used the tools at length. It does give you an idea of the timescales and resources which were used for a major exercise.
You could go on and develop a strategy for each of the players after you've worked though chapter n.
This deals with the unfriendly take-over bid of Allied Lyons by Ex*
It had the advantage that a Financial Times journalist was allowed to sit in on a number of meetings and wrote it up fully, published in the FT then as a booklet.
Open University M352
Open University Fundamentals of computing
Henley Management series
Airline ticketing systems
Library files of cuttings
Seminars and group work
Documents on which you might draw
Doing design exercises on your own situation
Number of books dealing with corporate strategy but not giving form to ISD
Notes to teachers on students having difficulty applying these techniques - being mechanical
This idea of advantage, and of competition, seems central to the current flavour of ideology. It is likely to prove a passing fashion, but it involves a metaphor and a view of the world which people understand so for the moment I'll use it uncritically, but to this matter I'll return.
See for example Anderla, Georges and Dunning, Anthony. Computer strategies 1990-9: technologies - costs - markets. Chichester: John Wiley, 1987 for a summary of their expectations, based on an OECD funded research project
See for example Baring Securities detailed study of the process in Japanese finance, manufacturing and retailing by Alex Stewart, Automating distribution: revolution in distribution, retailing and financial services. London: Baring, 1987. I know of no British study which is as clear.
Frequently I'll be referring to events which might be known only to readers in a narrow national market or of a significance which seemed greater at the time of writing than they have proven to have. The details aren't important. Replace by one of your own favourite flavour.
Johnson and Scholes give a summary of how to apply the technique and refer to Tilles, S Making strategy explicit (1966) reprinted in Ansoff, I. Business strategy. Harmondsworth: Penguin, 1968 as an early account in their Exploring corporate strategy. London: Prentice Hall, 1988.
I don't want to replicate here a substantial literature. See for example Wilson, Brian. Systems concepts: methodologies and applications. Chichester: John Wiley, 1984 for a method to undertake this analysis.
His two historic works are Competitive strategy: techniques for analysing industries and competitors. New York: Free Press, 1980 and Competitive advantage:creating and sustaining superior performance. New York: Free Press, 1985 Johnson and Scholes draw heavily on his work, as does Remenyi, D.S.J. Increase profits with strategic information systems. Manchester, NCC, 1988, a cheap populariser of the ideas for computer managers. Earl. Michael J. Management strategies for information technology. London: Prentice Hall, 1989 provides a more substantial framework for attempting to fit these methods into designing strategies for information systems.
There are two arguments here: use someone who is close to you such as your bank or auditors to help in this process, because they you know well, or don't use them because they know you well. Building a long term relationship with a consultancy group allows for an evolutionary planning process. This appears to be the method favoured by Bain for example, but consider the result over the Guinness takeover bid. How this is handled will be a matter of corporate policy outside the information strategy study. The existence or otherwise of a special relationship will have been one of the identified factors in the early political level of the PES- T study.
This has been developed most fully in the retailing trade. Norton, Damien. Evolving merchanise profitability measures in Retail performance analysis for profit: improving sales and profits through analysis of key retail performance criteria. London: Management Horizons, 1987.
For arguments about the role of the privatisation programme see *get citation. For the social role of the market see Nove * get citation. For the function of the market in "state capitalist" societies see Harman, Chris. and for arguments on the limit of the function of the market see Harman or Callinicos.
This discussion cannot be separated from changes in higher education. See for example the Times Higher Education Supplement editorial (get date* 1988) "academic work has to be decentralised for professional and intellectual reasons that are easily understood"
I've not gone into these issues in any detail in this book, but if you are intersted in this rather specialised area of work contact the British Computer Society Developing Countries Group or see my Information *get citation London: British Library, 198?
get citation * from Long Range Planning