|Clearing the Mind
What is 'Information'?
Knowledge is Power
People and Management
Direction versus Management
Authority and Delegation
Decisions and Events
Return to:World Views
Ardue Site Plan
See also:Information and technology, Part 2
Information System Design
In such a climate, it is little wonder that far too many business and administrative information systems projects fail to deliver the promised benefits. Technical complexity and malfunction are usually cited as the culprits but, more often than not, such disappointments arise directly and predictably because the expected benefits and the constraints upon their delivery are not defined clearly enough.
So let's begin by going back to first principles.
A book can contain much information, but nobody would pretend that it 'knows' anything. Only people can be 'informed'. Without final conversion into knowledge in the minds of human beings whose subsequent actions may be influenced thereby, the products of the information industry have no practical value.
Thus any 'information project' worth the name is primarily concerned with people and their relationships and inter-actions as they participate in a common endeavour to achieve some end or ends that one or more of them deems desirable. The more clearly the end or ends can be defined, the greater the likelihood that the project will have a successful outcome.
The starting point for any information is an item of data. Data is the plural of datum, a Latin word meaning something given, a reference point, a 'fact'. Information is normally taken to mean any ordered collection of data. Thus data processing is the collection and arrangement of representations of given facts in accordance with sets of rules in such a way as to highlight any 'significance' they may have from the points of view the rules are devised to serve. The points of view in question should be relevant to the interests of the individual or group which owns or uses the collection and the rules should be drawn up by someone who understands these interests.
Whether or not an ordered collection of data constitutes information depends largely on its relationship to other information with which it may be associated in the mind of the individual who is to be informed thereby. In some cases a single datum (fact) may be information; in others, a report compiled from processing many ordered data may be needed. So whether or not an ordered collection of data constitutes information depends on the purpose for which the information is needed and on who needs it. There is no point in generating ordered collections of data for their own sake.
If information is food for thought, data processing is like food preparation. The object is to convert data into nourishing mental 'meals' and serve them up in ways which make them 'appetising' and easy to 'swallow' and 'digest'. Thinking about information converts it into knowledge in the mind of the thinker. And because every individual possesses a unique set of facts, ideas and thinking habits, the same item of information can inform different individuals in quite different ways.
It sometimes happens that, for whatever reason, an individual or group becomes dissatisfied with its current state. Circumstances may arise which make it necessary or desirable to adopt new aims and abandon old ones. At times like these, we may become aware that we lack some of the knowledge we need to make well-informed decisions and must extend our information-gathering activities to fill gaps in our knowledge or extend it to include some hitherto unfamiliar subjects. We cannot afford to stop learning as long as we intend to remain on the 'active list' of human beings.
We normally expect the information in an invoice to confirm what we already know about the transaction in question. If it doesn't, we must re-check our existing information and effect a reconciliation which may result in a modification of our previous knowledge, or an amended invoice, or both. It may even be necessary to modify the system in order to eliminate any malfunction that may have caused the discrepancy. This is a simple illustration of the profound truth that we learn from our own mistakes. A well-designed information system should also be able to incorporate what we could learn from the mistakes of others!
An integrated system is one which combines several simpler sub-systems into one, avoiding duplication of inputs and thereby reducing the potential for error. For example, an integrated sales order processing system would accept data about customers, suppliers, orders, sales, purchases, requisitions, deliveries, returns, etc. and process all these separate facts to produce delivery notes, invoices, statements of account, receipts, stock records, and other reports designed to highlight matters of interest such as stock shortages, overdue deliveries or payments, changes in customer buying patterns, and other matters which might affect the well-being of the firm.
Business knowledge is of two kinds: objective and intuitive. Objective knowledge is factual knowledge about things which already exist or events which have already taken place. Knowledge of what is possible in the future can only be intuitive. It is intuition that enables the businessman to envisage what satisfactions he may be able to supply economically tomorrow as a result of pursuing a particular course of action today.
Intuition may, of course, be informed by knowledge of regular patterns of relationships between things and events which have held good in the past and may be assumed to hold good for the future. Observation and recording of events over long periods have enabled us to discern cyclical variations in human values, behaviour and fashions that induce corresponding variations in business or political activity. We can sometimes make 'sense' of complex and seemingly chaotic sequences of events by analysing them into cycles of differing periodicities superimposed on one another.
However, it is dangerous to extrapolate the past into the future without taking precautions against the possibility of a sudden deviation from our expectations. Hence every information system should include provision for comparison of its actual outputs with these expectations. This will not be possible unless the expectations are clearly defined in the first place.
An information system cannot itself convert information into knowledge. Every organisation has its own special 'body of knowledge' which is created in the minds of the people within the organisation who think about the information available to them. Because knowledge exists only in the mind, the individual's knowledge must continually be re-created to take account of new information or to resolve inconsistencies as they become apparent. Likewise, an organisation's stock of knowledge must continually be re-created in fresh minds if the organisation is to flourish and evolve while retaining its continuity. This re-creation is what enables knowledge to be expanded and adapted to meet new circumstances and challenges.
We can neither force people to think nor prevent them from thinking; but a comprehensive and responsive information system can make it easier for people to obtain fresh supplies of food for thought as and when they feel 'hungry'. Thought generates new ideas and concepts which can be stored in, and disseminated by, the information system. If it is properly integrated, the same system can also store and disseminate the instructions by which the ideas and concepts may be converted into effective action and by which the results of these actions can be measured against prescribed targets.
People tend to be volatile. No two people ever share exactly the same set of personal attributes or priorities. It is extremely unlikely that any two people will act in precisely the same way in any moderately complex set of circumstances. So if two or more people are to co-operate successfully in pursuit of a common objective, they must not only be able to share information; they must also have some frame of reference within which their actions can be co-ordinated. For this reason, an information system works best if it also includes the functions of a control system; it then becomes a management information system.
For much of the twentieth century, management in all its aspects has been a widely discussed and debated topic. Many management pundits have expounded their ideas, many schools of management have sprung up, and many expensive management seminars have been held to extol the benefits of this management system and that. During the latter half of the century, there has been a cultural and legal emphasis on 'job security'. In large organisations, the personal foibles of individuals have been catered for by the introduction of elaborate hierarchies and job titles, all in the name of 'good' management.
In recent years, it has become apparent that this sort of 'good' management is too expensive to be affordable in the long run and many long-established traditions of management have become subject to critical scrutiny. This gives us an opportunity to re-build from the bottom up.
To direct is to define a result to be achieved and to specify any qualitative or quantitative constraints upon the actions required to achieve it.
To manage is to achieve a pre-determined result working within the prescribed constraints. Unless the result to be achieved and the relevant constraints are clearly defined, there can be no way of telling whether or when efforts to achieve it have been successful, or how much still remains to be done.
It often happens in practice that the director and the manager are one and the same person: but the two roles must not be confused. Direction comes first. Without direction, there can be no progress. There is no means of measuring a manager's performance except by reference to a set of goals to be scored. The goals are defined by directors.
Setting high-level, long-term strategic goals calls in large measure for intuitive knowledge: but objective knowledge is also needed, especially to identify some of the constraints which must be imposed upon actions required to pursue the goals and particularly those actions requiring the application of scarce resources. Long-term business success requires a careful balance between what is perceived as desirable and what can actually be achieved with the available means.
Assessment of what is possible more often than not entails a detailed quantitative analysis. It may be necessary to translate strategic objectives into a hierarchy of 'tactical' or 'enabling' objectives and to cost each of these before the likely overall cost of pursuing the strategy can be established with any confidence. Definition of each of these subordinate objectives may itself call for some degree of direction and some managerial assessment of its feasibility.
Thus we see that, in practice, direction is not the exclusive prerogative of members of the Board; everybody who makes a business decision is acting as a director. Similarly, everybody who achieves a business result at any level is acting as a manager. Neither direction nor management is a respecter of persons and there is no logical ground for the long-established management practice of tying each individual person to a particular level in a hierarchy of objectives and establishing chains of 'responsibility' and 'accountability'. Although they look nice and tidy in organisational diagrams, such artificial links merely tend to strangle initiative and substitute buck-passing for constructive activity. The concept of authority offers a viable alternative.
Thus the functions of the Board of Directors are not merely to direct but also to constitute the top level of the authority structure of the organisation. At Board level, the functions over which each Board member has authority should be defined with as little overlap as possible to reduce the risk of conflict. Where conflict arises, it must be resolved by appeal to the ultimate source of authority in the field — whoever that is perceived to be. Definition of who has what authority is one of the most important requirements for a successful control system.
Each Board member may delegate all or part of his or her authority to any other person deemed worthy and competent to exercise the authority in the best interests of the organisation, and that person may in turn delegate to someone else. But it is important in every instance of delegation to define the function or functions which fall within the scope of the authority and any constraints which apply to the exercise of the authority in respect of each function. Every delegation of authority constitutes a decision.
Delegating authority in no way diminishes the responsibility of the delegator. If the delegatee turns out to be unworthy of the trust, the responsibility rests with whoever delegated the authority in the first place. This probably explains why many people are reluctant to delegate and insist on doing things themselves which could be done as well or better by others. Consideration should be given to reducing the level of authority such people are authorised to exercise. (It may be easier to do this in an environment that is untrammelled with job titles and other status symbols.)
Delegation of authority is by far the most effective way of harnessing the creative talents of individuals because it fosters a climate of trust and creates new responsibilities which trustworthy people are normally happy to accept. It has the effect of reducing the number of irresponsible people in the organisation — ideally to zero.
It is important to note that once authority has been delegated, the delegator no longer has it. Nothing but confusion can result from two or more people trying to exercise the same authority in different ways at the same time. Business flexibility may give rise to quite complex authority structures in which a person who has authority over a particular individual for one function may be subordinate to the same individual for another function. For this reason, new delegations of authority or withdrawal of previous delegations must be notified to all affected parties as quickly as possible: the most important thing about authority is knowing who has it.
Thus a facility to define who holds what authority and to keep abreast of changes is an essential component of an effective management information system.
We have seen how a management information system should be able to keep track of objectives, of progress towards objectives, and of authority relationships, as well as record details of the myriad transactions and other significant events that take place both within the organisation and between the organisation and its environment.
Until about 1980, nearly all these functions used to take place in 'the office'. Around 1980, every other job was an 'office' job. The telephone and typewriter were the tools of the 'white collar worker'; the desk and the filing cabinet were the twin foci of millions of working lives. A little variety was added to this limited ambit by the telex, the photocopier and the facsimile machine, all of which not only speeded up the processing of external transactions but also contributed their own flow of paper into the filing cabinet and the waste paper basket. All this frantic paper-shuffling meant that an ever-increasing proportion of administrative effort had to be devoted to internal management. Yet despite the best efforts of this devoted army of paper-shufflers, they proved unable to keep up with the rapidly increasing speed and capacity of external communications.
This introverted office came about largely as a result of treating it as a mere tidying-up operation. It incurred overheads but made no contribution to revenue-earning. The balance had to shift. And the only thing that could shift the balance was a device that could perform all the functions of a paper filing system at a speed that could not only match the rate at which communications flowed in and out of the organisation but also provide spare capacity to analyse the contents of communications and present the results to the organisation's directors and managers in readily digestible packets of 'food for thought'. This device is the electronic computer discussed in Part 2.